Elsevier

Physics Reports

Volume 886, 25 November 2020, Pages 1-84
Physics Reports

The electromagnetic counterparts of compact binary mergers

https://doi.org/10.1016/j.physrep.2020.08.008Get rights and content

Abstract

Mergers of binaries consisting of two neutron stars, or a black hole and a neutron star, offer a unique opportunity to study a range of physical and astrophysical processes using two different and almost orthogonal probes — gravitational waves (GW) and electromagnetic (EM) emission. The GW signal probes the binary and the physical processes that take place during the last stages of the merger, while the EM emission provides clues to the material that is thrown out following the merger. The accurate localization, which only the EM emission can provide, also indicates the astrophysical setting in which the merger took place. In addition, the combination of the two signals provides constraints on the nature of gravity and on the expansion rate of the Universe. The first detection of a binary neutron star merger by the LIGO–Virgo collaboration, GW170817, initiated the era of multi-messenger GW–EM astrophysics and demonstrated the great promise it holds. The event produced an unprecedented data set, and although it was only a single event, it provided remarkable results that revolutionized our knowledge of neutron star mergers. GW170817 is especially exciting since we know that it is not one of a kind and that many more events will be detected during the next decade. In this review, I summarize, first, the theory of EM emission from compact binary mergers, highlighting the unique information that the combined GW–EM detection provides. I then describe the entire set of GW and EM observations of GW170817, and summarize the range of insights that it offers. This includes clues about the role that similar events play in the r-process elements budget of the Universe, the neutron star equation of state, the properties of the relativistic outflow that followed the merger, and the connection between neutron star binary mergers and short gamma-ray bursts. I conclude by discussing some of the future prospects of this new window that has been opened.

Introduction

For many decades, the mergers of two neutron starts or of a black hole and a neutron star, were considered the most promising candidates for a joint gravitational wave (GW) and electromagnetic (EM) detection. This promise was realized in August 2017 when the gravitational waves from a binary neutron star merger were detected by the Laser Interferometer Gravitational Wave Observatory (LIGO) and Virgo (Abbott et al., 2017c). A short flash of γ-rays was detected independently by the Fermi Gamma-ray Burst Monitor (Goldstein et al., 2017) and by INTEGRAL gamma-ray observatory (Savchenko et al., 2017) only 1.7 s after the end of the chirping GW signal marked the merger of the two neutron stars. Later, EM emission was detected across the whole spectrum and GW170817 became one of the most studied astrophysical transients.

The combination of GW and EM signals provides unique information. The GWs carry information about the binary at the last moments before it merges. This includes constraints on the masses, spins and tidal deformability of the binary members, as well as the orientation and eccentricity of the binary orbit, the exact time of the merger, and its luminosity distance from earth. The EM emission has the potential to provide a precise localization, which puts the merger in its astrophysical context, including a measure of its redshift as well as identification of the host galaxy and the specific environment in which the merger took place. In addition, the EM emission probes the outflow that the merger ejects into the circum-merger medium. By modeling this emission, we can learn about the various outflow components and their composition, geometry, and velocity.

Further reward results from the combination of the two signals, which holds the key to a range of fundamental questions in physics and astrophysics. The difference in the arrival times of the GWs and the EM emission teaches us about the nature of gravity. The independent constraints on the luminosity distance and redshift provide a measure of the expansion rate of the Universe, a method which is independent of the assumptions used in all other methods. The precise measurement of the local merger rate, which can be achieved with a well-defined sample of merger events detected by their GW signal, together with the EM measurement of the mass and the composition of the ejecta, shed light on the origins of the heaviest elements. GW information on the binary inspiral before and during the merger, together with EM mapping of the various outflow components, put unprecedented constraints on the equation of state of matter at supranuclear densities. Finally, the identification of the binary members and the emission of the outflow, may reveal the long-sought progenitors of short gamma-ray bursts (sGRBs).

There has been significant progress in the theoretical study of compact binary mergers during the past decade. This progress was motivated by the continuous improvements in the sensitivity of GW detectors, and facilitated by growing computational power, which is needed for simulations of various phases of the merger. The breakthrough that came with the detection of GW170817 brought our understanding of compact binary mergers to a new level. It provided an exceptional observational data set, and put the study of binary mergers at the focus of numerous observational and theoretical efforts. In the past two years, about a thousand papers have been written on GW170817 and on various issues that its detection raised. However, this is still a young field where most of the progress is yet ahead of us, anticipating the discovery of many more mergers during the next decade. This is, therefore, a good time to summarize the knowledge that has been accumulated to date and which will serve as a basis for future advancements.

Naturally, the entire topic of compact binary mergers cannot be covered by a single review. Here I focus on the main sources of the EM emission from compact binary mergers and on the information that we can extract from their detection jointly with GWs. I pay special attention to the theory of the various components of the outflow that is ejected from the system following the merger and the physical processes that shape their EM emission. I then discuss the observations of GW170817 and the theoretical models of the event. Most of the review is a summary of published studies, but it includes also some novel derivations, synthesis and conclusions.

Even within the narrow scope of this review, one cannot cover the entire relevant literature, and I apologize for all the important studies of the EM emission from binary mergers that are not discussed here. Below, I provide, as an introduction, a very brief summary of the outflow from compact binary mergers and its EM emission, with references to the sections in the review that elaborate on each topic, followed by a description of the review layout.

The outflows from a binary neutron star (BNS) merger and from a black hole–neutron​ star (BH–NS) merger are expected to have, broadly speaking, two components - a sub-relativistic and a relativistic one (see Fig. 1 for example). The sub-relativistic outflow has three major sources (Section 2.2 and Section 2.3): (i) tidal forces that operate during the final stages of the inspiral and the initial stages immediately after the merger; (ii) shocks driven by the collision between the two binary members in the case of a BNS; and (iii) winds from the accretion disk that forms following the merger. All of the sources involve decompression of highly dense neutron rich material and therefore they are all prime sites for the nucleosynthesis of r-process elements (Section 2.1). In fact, BH–NS and BNS mergers were suggested already four decades ago as major sources of r-process elements in the Universe (Lattimer and Schramm, 1974, Symbalisty and Schramm, 1982). The heavy nuclei are formed very far from the valley of stability and therefore they go through a chain of beta-decay, alpha-decay and nuclear fission on their way to stability. This radioactive decay provides a continuous source of heat, similarly to the case of Type Ia supernovae, which eventually escapes as a ultraviolet (UV), optical and infrared (IR) radiation that can be detectable for weeks and even months (Li and Paczyński, 1998) (Section 3). In the case of compact binary mergers, this radiation has been called by various names, with kilonova and macronova being the most common ones. The properties of the sub-relativistic ejecta depend on the nature of the binary (i.e., BNS or BH–NS), the masses and spins of the binary members, and the NS equation of state (EOS). Observationally, the macronova/kilonova emission is sensitive to outflow properties such as its mass, velocity, composition, and geometry. Therefore, when the observations of the macronova/kilonova are combined with the information obtained from the GW signal, they provide a powerful tool to study the physical processes that take place during the merger. The sub-relativistic ejecta is expected to continue and radiate due to its interaction with the circum-merger medium. This interaction is expected to produce a radio and possibly X-ray remnant, which may be detectable on a timescale of months to years, depending on the ejecta and the medium properties (Nakar and Piran, 2011).

Within several dynamical times after the merger, some of the bound material settles into an accretion disk that surrounds a rapidly rotating central object. The central object may be a NS (most likely highly magnetized, i.e., a magnetar, see Section 3.6.1) or a BH. In BNS merger, the nature of the central object depends on the total binary mass and on the NS EOS. If the merged central object is a NS which is supported by rotation, then it can collapse to a BH at any time after the merger. In BH–NS mergers the merged object is, naturally, a BH. Both types of systems (a rapidly rotating magnetar and a BH with a disk) are promising sources of ultra-relativistic jets, such as those that are present in gamma-ray bursts (GRBs). The jets are the second, relativistic, outflow component expected in a compact binary merger. And indeed, the idea that BNS mergers are the progenitors of GRBs was suggested already three decades ago (Eichler et al., 1989). Following the realization that GRBs are separated into at least two sub-classes, short and long, the common expectation was that the short gamma-ray bursts (sGRBs) are generated by BNS and/or BH–NS mergers. The detection of sGRB afterglows in 2005 (Gehrels et al., 2005, Fox et al., 2005, Berger et al., 2005) provided strong support for this model, but the evidence was only circumstantial (Nakar, 2007 and references therein).

If compact binary mergers are sGRBs, then the jet that they launch must drill through the sub-relativistic ejecta that covers the polar region at that time, break out of the ejecta and release the intense burst of γ-rays. The following interaction of the jet with the circum-merger medium drives a blast wave that generates the long lasting X-ray, optical and radio afterglows. The emission from the jet is extremely bright and can be seen to high redshifts, but only if the observer is within the opening angle of the jet. Relativistic beaming renders the emission at large angles from the jet too faint for detection even at the distance where GWs from mergers are detectable. Since sGRB jets are most likely narrow, and the GW detection is not very sensitive to the inclination of the binary orbit, the expectation is that only rarely we will be able to see the jet directly in mergers that are detected by their GW signal. Moreover, the jet needs a significant power in order to cross the entire sub-relativistic ejecta successfully, at least in BNS mergers. Therefore, it is expected that in some mergers the jets will fail to do so and get choked while still within the ejecta, in which case the jets cannot be observed directly by any observer.

Luckily, the jet can leave an observable imprint also when it points away from us or when it is choked (Section 4). When a relativistic jet propagates through the sub-relativistic material it inflates a highly pressurize bubble known as the “cocoon” (Section 4.1). The jet and the cocoon interact with each other and evolve together within the ejecta. When (and if) the jet breaks out of the ejecta successfully, it is accompanied by the cocoon and together they form an outflow, which is spread over an opening angle that is much wider than the opening angle of the jet alone. The entire outflow is known as the “jet-cocoon” and it has a structure which is dictated by two factors, the structure with which the jet is launched from the central compact object and the interaction of the jet with the sub-relativistic ejecta. The expectation is that the jet-cocoon outflow will have mostly an angular structure where, along the jet axis, there is a narrowly collimated core with high isotropic equivalent energy and high Lorentz factor, while outside of the core the energy and Lorentz factor decrease with the angle. This type of outflow is often called a “structured jet”. In this picture, a sGRB is produced by the narrow jet core and can be seen only by observers that look directly into its cone. Fainter, yet potentially observable emission (at GW detectable distances) is generated over a wider angle by the jet-cocoon via several processes. First, the jet and the cocoon drive a mildly relativistic or even a fully relativistic shock into the ejecta. When such a shock breaks out it produces a short flash of γ-rays  (Nakar and Sari, 2012) (Section 4.2.2). γ-rays may be also emitted over a wider angle than the opening angle of the jet core, due to internal dissipation within the jet-cocoon outflow at regions that are outside of the core (Section 4.2.4). Second, the energy deposited by the shock in the cocoon diffuses as it expands and escapes to the observer, producing an X-ray, UV and optical cooling emission (Nakar and Piran, 2017, Lazzati et al., 2017a) (Section 3.6.3). Third, radioactive decay of the r-process elements in the cocoon can dominate the early macronova/kilonova emission (Gottlieb et al., 2018a) (Section 3.6.4). And, finally, the interaction of the jet-cocoon with the circum-binary medium produces an X-ray, optical and radio afterglow that can possibly be seen over a wide observing angle (Nakar and Piran, 2017, Lazzati et al., 2017a) (Section 4.3).

A choked jet may also produce observable emission . If the jet deposits enough energy in the cocoon before it is choked, then the cocoon can break out of the sub-relativistic ejecta. The cocoon outflow in that case is mildly relativistic and no sGRB is seen (by any observer). However, the cocoon produces emission that may be observable at a distance where GWs are detectable via most of the same processes as those of the successful jet-cocoon, i.e., shock breakout, cooling emission, radioactively powered cocoon emission and interaction with the circum-merger medium (Nakar and Piran, 2017). Some of the properties of the emission in the case of a choked jet are different than those of a successful jet-cocoon, allowing us to potentially distinguish between the two cases.

The goal of this review is to describe the current state of the theory and observations of the EM signal from compact binary mergers. It discusses separately the different components of the ejecta, where for each component I consider first the dynamics of the outflow, followed by a description of the resulting emission. Each of the sections and many of the subsections, are as self contained as possible, so that one can read only the section(s) of interest. The review is focused more on theory, but for each of the topics I describe the relevant observations and how the observations can constrain the theoretical models. I also try, to the degree possible, to explain what is the basis for each of the ingredients of our current theoretical views, paying special attention to how robust is each conclusion and how it can be tested in the future. Below I describe, briefly, the content of each of the sections.

In Section 2 I review the properties of the sub-relativistic ejecta and the various physical processes that drive this outflow. Since the sub-relativistic ejecta is expected to be rich with r-process elements, I start (Section 2.1) with a brief review on the r-process and discuss observational constraints on r-process sources in the Universe. I then discuss the various processes of mass ejection and the properties of the outflows that they drive from BNS mergers (Section 2.2) and from BH–NS mergers (Section 2.3).

The UV/optical/IR emission from the sub-relativistic ejecta is discussed in Section 3. As already noted, different names are used in the literature for this emission, where the most common are kilonova and macronova. In this review I use the term kilonova and I discuss shortly its etymology in Section 3.1. I describe a simple model of the kilonova emission (Section 3.2), which depends on the radioactive heating rate (Section 3.3) and on the opacity of r-process material (Section 3.4). I then describe a robust method to estimate the mass of the ejecta (Section 3.5). I conclude this section by discussing the first-day kilonova emission and non-radioactive energy sources (Section 3.6).

I start the discussion of the relativistic outflow with a description of the interaction between the jet and the sub-relativistic ejecta, and the structure that this interaction induces on the relativistic outflow (Section 4.1). The γ-ray emission from the relativistic jet depends strongly on the observer viewing angle, where observers within the opening angle of the jet see, presumably, a GRB. Since GRB emission has been discussed extensively in a large number of reviews and it is most likely that the jet of a GW-detected merger will point away from us, I restrict the discussion here to the γ-ray emission seen by observers that are away from the jet (Section 4.2). The discussion includes a generalization of the compactness limits to off-axis observers (Section 4.2.1), shock breakout theory (Section 4.2.2), jet off-axis emission (Section 4.2.3) and high latitude emission from the wings of a structured jet (Section 4.2.4). The interaction of the relativistic outflow with the external medium and the afterglow has been also discussed extensively in the literature. Here I provide only a partial description of this theory, which focuses on aspects and regimes that are important for the interpretation of the afterglow of GW170817, and are expected to be useful also for future events (Section 4.3). The discussion is separated according to the various phases of the afterglow (rise, peak and decay), where for each of these phases I focus on the information that we can extract from the observations on the structure of the outflow.

In Section 5 I cover aspects of the GW emission that are either directly related to the physical interpretation of the EM emission, or that can provide unique information when combined with the EM signal. The former includes the GW-based measurements of properties such as the masses, spins and tidal deformability of the binary members as well the inclination of the orbital plane (Section 5.1). The latter aspects of the GW emission that I discuss include the constraints that the arrival times of the GW and EM signals pose on the propagation of GWs (Section 5.2), and the usage of compact binary mergers as standard sirens to measure the Hubble constant (Section 5.3).

Section 6 reviews GW170817. First, I summarize the observations without any attempt to provide theoretical interpretations (Section 6.1). Then, I describe what we can learn from the observations on the properties of the binary (Section 6.2), the Hubble constant (Section 6.3), the sub-relativistic ejecta (Section 6.4), the structure of the relativistic outflow (Section 6.5) and on the source of the observed γ-rays (Section 6.5.2). I then discuss the implications of these results for the fate of the remnant (Section 6.6.1), the NS equation of state (Section 6.6.2), and the origin of r-process elements in the Universe (Section 6.7).

In Section 7, I summarize our understanding of the connection between sGRBs and BNS mergers following GW170817. I explain why, contrary to the popular view that GW 170817 provided the ultimate proof that BNS mergers produce sGRBs, this connection is almost, but not entirely, secure. I review the various kilonova candidates in sGRB afterglows (Section 7.2). I discuss the implications of the jet seen in GW170817 on sGRB jets (Section 7.3.1), and address the question of how many GRB170817A-like events hide in the current sample of sGRB prompt emission (Section 7.3.2). I compare the constraints on the rate of sGRBs to the rate of BNS mergers and discuss the implications of this comparison (Section 7.4). I conclude this section by evaluating the current status of the possibility that BH–NS mergers are progenitors of sGRBs (Section 7.5).

Finally, in section Section 8 I briefly discuss some of the future prospects in this field, focusing on some of the things that we will most likely learn once there will be a large sample of BNS and BH–NS​ mergers.

Section snippets

Sub-relativistic mass ejection from compact binary mergers

A BNS merger ejects a considerable amount of mass at sub-relativistic velocities. Substantial theoretical effort has been invested over the past decades in order to predict the properties of the ejected material and its observational signatures. The general predictions about the mass of the sub-relativistic ejecta, before GW170817, varied by more than an order of magnitude centered around a canonical value of about 103102M of r-process material, that is ejected at a velocity of about 0.10.3

Theory of the UV/optical/IR emission (macronova/kilonova)

A sub-relativistic r-process-rich outflow produces a radioactively powered UV/optical/IR transient in a manner that is very similar to Type Ia SNe. The basic physics is that radioactive heat is deposited into the expanding material, is thermalized to some extent, and leaks out towards the observer. The luminosity, spectrum and duration of the transient therefore depend on the mass, velocity, radioactive heating rate, and opacity of the outflow. The first to suggest that compact binary mergers

The relativistic outflow — propagation, γ-ray emission, and the afterglow

BNS mergers launch relativistic jets. This was a long-standing expectation based on the conjectured association of short GRBs with BNS mergers (Eichler et al., 1989), and was confirmed by the afterglow observations of GW170817. The engine of the jet is presumably the central compact object formed following the merger, either a magnetar or an accreting BH. The launched jet needs to propagate through the sub-relativistic ejecta that covers the polar region at that time. The jet propagation drives

Information carried by the gravitational waves

The combination of GW and EM signals makes BNS and BH–NS mergers unique. The focus of this review is on the EM emission, and here I give only a brief description on what can be learned from the GW signal alone about the properties of the binary, and from the combination of GW and EM emission on some fundamental questions in physics and cosmology. I refer the reader to Sathyaprakash and Schutz (2009) and Baiotti and Rezzolla (2017) for reviews that discuss the GW emission from compact binary

Gravitational waves

GW170817 was the fifth confirmed GW event detected by LIGO and the second joint detection with Virgo. But, unlike all earlier events, the mass of the binary, as deduced from the GW signal, implied that the binary members are most likely neutron stars rather than black holes. The main details of the GW signal detection are given in Abbott et al. (2017c) and references therein. Here I give a brief summary.

The merger was detected on August 17, 2017, towards the end of the advanced LIGO O2 run

A brief historical account

The connection between GRBs and BNS and/or BH–NS mergers was first predicted32 by Eichler et al. (1989). This paper, written at a time when most of the community believed that GRBs were Galactic, already included most of the ingredients that turned out to be relevant three decades later, such as a prediction that a significant

Future prospects

The continuous search for GW emission from compact binary mergers, and the expected improvements in the detectors sensitivity, guaranty that within a decade we will have a large sample (hopefully several dozens) of identified BNS mergers and BH–NS mergers. If the EM emission from GW170817 was representative, then many of the GW detected BNS mergers will have EM counterparts, and in some of those the kilonova emission will be detected within hours of the mergers, or even earlier. It is

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

Special thanks to Dan Maoz for a careful and constructive review of the manuscript. His invaluable comments and help have improved it considerably. I thank Tsvi Piran for many enlightening discussions. I also thank Iair Arcavi, Omer Bromberg, Francois Foucart, Ore Gottlieb, Kenta Hotokezaka, Mansi Kasliwal, Amir Levinson and Dovi Poznanski for valuable discussions and comments. This work was partially supported by an ERC, The European Union grant (JetNS) and by Israel Science Foundation Grant

References (420)

  • BadnellN.R.

    Comput. Phys. Comm.

    (2011)
  • ChadwickM.B. et al.

    Nucl. Data Sheets

    (2011)
  • AbbottB.P. et al.

    Nature

    (2017)
  • AbbottB.P. et al.

    Astrophys. J. Lett.

    (2017)
  • AbbottB.P. et al.

    Phys. Rev. Lett.

    (2017)
  • AbbottB.P. et al.

    Astrophys. J. Lett.

    (2017)
  • AbbottB.P. et al.

    Phys. Rev. Lett.

    (2018)
  • AbbottB.P. et al.

    Phys. Rev. X

    (2019)
  • AbbottB.P. et al.

    Phys. Rev. X

    (2019)
  • AbeK. et al.

    ApJL

    (2018)
  • AlbertA. et al.

    Astrophys. J. Lett.

    (2017)
  • AlexanderK.D. et al.

    Astrophys. J. Lett.

    (2017)
  • AlexanderK.D. et al.

    Astrophys. J. Lett.

    (2018)
  • AloyM.A. et al.

    Astrophys. J. Lett.

    (2000)
  • AndreoniI. et al.

    Publ. Astron. Soc. Aust.

    (2017)
  • AnnalaE. et al.

    Phys. Rev. Lett.

    (2018)
  • AntoniadisJ. et al.

    Science

    (2013)
  • ArcaviI.

    Astrophys. J. Lett.

    (2018)
  • ArcaviI. et al.

    Nature

    (2017)
  • ArcaviI. et al.

    Astrophys. J. Lett.

    (2017)
  • ArgastD. et al.

    Astron. Astrophys.

    (2004)
  • ArnettW.D.

    Astrophys. J.

    (1982)
  • AubourgÉ. et al.

    Phys. Rev. D.

    (2015)
  • BaiottiL. et al.

    Rep. Progr. Phys.

    (2017)
  • BarnesJ. et al.

    Astrophys. J.

    (2013)
  • BarnesJ. et al.

    Astrophys. J.

    (2016)
  • BausweinA. et al.

    Astrophys. J.

    (2013)
  • BausweinA. et al.

    Astrophys. J. Lett.

    (2017)
  • BeatonR.L. et al.

    Astrophys. J.

    (2016)
  • BeloborodovA.M. et al.

    Astrophys. J.

    (2020)
  • BeniaminiP. et al.

    Astrophys. J.

    (2016)
  • BeniaminiP. et al.

    Mon. Not. R. Astron. Soc.

    (2019)
  • BeniaminiP. et al.

    Mon. Not. R. Astron. Soc.

    (2019)
  • BeniaminiP. et al.

    Mon. Not. R. Astron. Soc.

    (2019)
  • BergerE.

    Annu. Rev. Astron. Astrophys.

    (2014)
  • BergerE. et al.

    Astrophys. J. Lett.

    (2013)
  • BergerE. et al.

    Nature

    (2005)
  • BernuzziS. et al.

    Mon. Not. R. Astron. Soc.

    (2020)
  • BertottiB. et al.

    Nature

    (2003)
  • BlanchardP.K. et al.

    Astrophys. J. Lett.

    (2017)
  • BlinnikovS.I. et al.

    Sov. Astron. Lett.

    (1984)
  • BonvinV. et al.

    Mon. Not. R. Astron. Soc.

    (2017)
  • BoranS. et al.

    Phys. Rev. D.

    (2018)
  • BovardL. et al.

    Phys. Rev. D.

    (2017)
  • BrombergO. et al.

    Astrophys. J.

    (2011)
  • BrombergO. et al.

    Mon. Not. R. Astron. Soc.

    (2016)
  • BrombergO. et al.

    Mon. Not. R. Astron. Soc.

    (2018)
  • BuckleyD.A.H. et al.

    Mon. Not. R. Astron. Soc.

    (2018)
  • BudnikR. et al.

    Astrophys. J.

    (2010)
  • BurbidgeE.M. et al.

    Rev. Modern Phys.

    (1957)
  • Cited by (163)

    • Neutrinos and nucleosynthesis of elements

      2024, Progress in Particle and Nuclear Physics
    View all citing articles on Scopus
    View full text