Geometrical inequalities bounding angular momentum and charges in General Relativity
Abstract
Geometrical inequalities show how certain parameters of a physical system set restrictions on other parameters. For instance, a black hole of given mass can not rotate too fast, or an ordinary object of given size can not have too much electric charge. In this article, we are interested in bounds on the angular momentum and electromagnetic charges, in terms of total mass and size. We are mainly concerned with inequalities for black holes and ordinary objects. The former are the most studied systems in this context in General Relativity, and where most results have been found. Ordinary objects, on the other hand, present numerous challenges and many basic questions concerning geometrical estimates for them are still unanswered. We present the many results in these areas. We make emphasis in identifying the mathematical conditions that lead to such estimates, both for black holes and ordinary objects.
Keywords
Geometrical inequalities Black holes Ordinary objects1 Introduction
Geometrical inequalities in General Relativity, that is, relations play an important role in understanding some physical systems. The basic questions behind these inequalities are the following. What are the reasons that such inequalities do exist at all? Another, more humble but more practical and hopefully illuminating question is about the elements in Einstein theory, that produce such inequalities. In other words: Why should we expect such inequalities, and where do they come from? The straightforward answer is that there are a number of different objects predicted by the theory that live in different regimes.
This is clear when one thinks about one of the most important class of solutions to Einstein equations: Black holes. In the evolution of stars, there are two natural limits to consider. One of them is the maximum mass a star can have, beyond which it collapses into a black hole. This problem was addressed in the 1930s by Chandrasekhar (see Wali 1982). The other is the maximum charge and/or angular momentum a black hole can have, beyond which it becomes a naked singularity. This problem arose after Reissner (1916) and Nordström (1918) found the solution describing a static, spherically symmetric, electrically charged object.
These thresholds in physical parameters values can be identified as limit cases of appropriate geometrical inequalities. In more general terms, we can naively imagine a function f depending on the physical parameters of the system, like the mass M, size R, angular momentum J, electromagnetic charge Q, etc., denoted by \(f:=f(M,R,J,Q, \ldots )\), such that when f takes values in \([f_{obj}^,f_{obj}^+]\), the system describes a nonblackhole ordinary material object, like a star or soccer ball. When f takes values in \( [f_{bh}^,f_{bh}^+]\), it describes a black hole, and when f takes values in \( [f_{nak}^,f_{nak}^+]\), a naked singularity.
We emphasize that this is a very rough and overly simplified picture of what the geometrical inequalities found so far actually say, and of what to expect for more general systems. But, as we see in Sects. 4.2 and 5.2, the above division showing the different regimes in which the system can exist, is what one actually finds in some cases.
In this article, we address systems containing ordinary material objects and/or black holes. The latter, and the frontier between black holes and naked singularities, is the original and main motivation for the geometrical inequalities presented here. Therefore, we explore it in what follows in the case of a paradigmatic blackhole solution.
It is important to recall that the Kerr–Newman metric is a solution of Einstein electrovacuum equations for any choice of the parameters (M, J, Q). However, it represents a black hole (and hence the area A of the horizon is well defined) if and only the inequality (3) holds. Otherwise the spacetime contains a naked singularity.
Above, we have derived inequalities (2)–(4) from a very particular exact solution of Einstein equations: the Kerr–Newman stationary black hole. However, remarkably, these inequalities remain valid (under appropriate assumptions) for fully dynamical black holes. Moreover, they are deeply connected with the expected properties of the global evolution of Einstein equations, in particular with the cosmic censorship conjecture.
 1.
\(\sqrt{\frac{A}{16\pi }}\le M\): the area appears as lower bound.
 2.
\(\frac{Q^2+\sqrt{Q^4+4J^2}}{2} \le M^2\) and \(4\pi \sqrt{Q^4+4J^2} \le A\): the angular momentum and the charge appear as lower bounds.
As the title of this article suggests, we will focus on the second group. For dynamical black holes, the inequality (2) in the first group is the Penrose inequality. There exists already an excellent and up to date review on this subject by Mars (2009).
Furthermore, the inequalities can also be classified as global or quasilocal. We explain this distinction in more detail in Sect. 2. Roughly speaking, the total ADM mass M is a global quantity, in contrast the area A is quasilocal, it depends on a bounded region of the spacetime. In contrast, there are both global and quasilocal definitions of charge Q and angular momentum J. With this in mind, (2) and (3) are global inequalities, and (4) is a purely quasilocal inequality. Global inequalities can be interpreted as refinements of the positive mass theorem in the presence of a black hole.

Geometrical inequalities involving quasilocal mass This is a very broad subject as there are many different notions of quasilocal mass and energy. The problem of determining a unique appropriate notion that will give general useful and representative geometrical inequalities is open. There is a beautiful review by Szabados (2004), on quasilocal quantities, and in particular, quasilocal mass, that discusses this issue

Geometrical inequalities for black holes in higher dimensions and within alternative theories of gravity This topic has been growing during the last years and many very interesting results have been obtained. See the work of Gibbons and Holzegel (2006), Gibbons (1999), Hollands (2012), Yazadjiev (2013a, b), Fajman and Simon (2014), Alaee et al. (2016, 2017a, b), Alaee and Kunduri (2014, 2016), Rogatko (2014, 2017), and references therein.
There are also a number of articles reviewing the subject of geometrical inequalities that include some of the results presented here. They were written by Dain (2011, 2012, 2014a), and Jaramillo (2013) with slightly different approaches and focuses.
1.1 Motivation from stationary black holes
Before discussing the general setting, it is important to analyze, in a heuristic way, the physical meaning of the inequalities (2)–(4) for stationary black holes.
Finally, consider the last inequality (4). In the following, using thermodynamics arguments for stationary black holes (Wald 2001), we will argue that this inequality is a consequence of the inequality (3) and the existence of extreme black holes [i.e., black holes with nonzero area that saturates (3)].
Given that the function A(M, J, Q) ends at the value \(A_0\), one can ask what happens at that point. For the extreme black hole the temperature is zero, hence the derivative (11) is infinite and then the function A(M, J, Q) can not be extended in any smooth way beyond the point \(A_0\). This can be explicitly checked in the Kerr–Newman case given by (1).
The implications (16) also put the Penrose inequality in a very especial place as being, in a sense, more basic than the other inequalities. We come back to this issue in Sect. 1.2 where it is deduced from standard arguments in collapse scenarios, and in Sect. 1.3 where it is a result of Newtonian considerations with the only condition that the speed of any particle should be smaller than or equal to the speed of light.
1.2 Heuristic arguments in dynamical blackhole regimes
The extension of the inequality (2) for dynamical black holes was done by Penrose (1973) using a remarkable physical argument that connects global properties of the gravitational collapse with geometric inequalities on the initial conditions. We briefly review this argument below (see also Mars 2009; Dain 2012, 2014a and references therein)
 (i)
Gravitational collapse results in a black hole (weak cosmic censorship).
 (ii)
The spacetime settles down to a stationary final state. Furthermore, at some finite time after the collapse, all the non electromagnetic matter fields have fallen into the black hole and hence the exterior region is electrovacuum.
The blackhole uniqueness theorem implies that the final state postulated in (ii) is given by the Kerr–Newman black hole (we emphasize however that many important aspects of blackhole uniqueness still remain open, see Chruściel et al. (2012) for a recent review on this problem). Let us denote by \(M_{{\textit{KN}}}\), \(A_{{\textit{KN}}}\), \(J_{{\textit{KN}}}\) and \(Q_{{\textit{KN}}}\) the mass, area, angular momentum and charge of the remainder Kerr–Newman black hole. These quantities will, of course, satisfy the three inequalities (2)–(4).
Let us consider an initial data set for a gravitational collapse such that the collapse has already occurred on the data. That means that the initial spacelike surface \(\varSigma \) intersects the event horizon of the black hole. The intersection is a spacelike, closed, 2surface denoted by \(\mathcal {S}\) with area \(A(\mathcal {S})\). Let M be the total mass of the spacetime defined by (69).
The key point in the previous argument is that there exist simple inequalities that relate the quantities M and A on the initial conditions with the quantities \(M_{{\textit{KN}}}\) and \(A_{{\textit{KN}}}\) on the remainder black hole, where the geometrical inequalities are satisfied. For the second inequality (3) we need to consider the electric charge and the angular momentum. The problem is that there is no simple inequality, like (18), that relates the total electric charge and angular momentum of the initial data \((Q_\infty ,J_\infty )\) with the corresponding quantities of the final, Kerr–Newman black hole \((Q_{{\textit{KN}}},J_{{\textit{KN}}})\). We need additional assumptions.
We have seen that the two inequalities (2) and (3) extend to the dynamical regime in the forms (21) and (27). These inequalities are global because the mass M is the total mass of the spacetime. Whether the quantities in these inequalities can be replaced by an appropriate defined black hole quasilocal mass and angular momentum is unknown.
Penrose argument is remarkable because starting from conjectures (i) and (ii) one is able to deduce inequalities that can be written purely in terms of the initial conditions. That is, the inequalities do not involve the unknown parameters \(M_{{\textit{KN}}}\), \(Q_{{\textit{KN}}}\), \(J_{{\textit{KN}}}\), \(A_{{\textit{KN}}}\) of the remainder black hole. A counter example to inequalities (21) or (27) in axial symmetry, will be a counter example of the conjectures (i) or (ii). On the other hand, the proof of such inequalities gives indirect evidences of the validity of the conjectures (i) and (ii). In that sense, the physical heuristic argument is quite strong: in either direction (i.e. if the inequalities are valid or not) it provides highly nontrivial new insight. In contrast, the physical heuristic arguments for the validity of the quasilocal inequality (4) in the dynamical regime are less conclusive.
The argument we present in Sect. 1.1 uses thermodynamics, and hence its validity outside equilibrium is not clear. Nevertheless, the quantities involved in (4) are, as we have seen in Sect. 2, well defined quasilocal quantities in the full dynamical regime (in the case of angular momentum we need the additional assumption of axial symmetry). Consider a stationary black hole that satisfies inequality (4) (we can assume instead that it satisfies the more general version (13), the following argument will be identical). We make a perturbation to this stationary black hole that preserves the charge and angular momentum of the black hole. For example, a vacuum axially symmetric perturbation will have this property. Physically, the stationary black hole will absorb axially symmetric gravitational waves without changing its charge and angular momentum. The area, however, will increase. That means that the same inequality (4) will be satisfied for this dynamical black hole in the future. Slightly more general, if we have a system of multiple black holes such that in the past they can be approximated by isolated stationary black holes and such that the whole spacetime is axially symmetric and electrovacuum, then, by the same argument we expect that the inequality (4) will be satisfied for each individual black hole. Head on collision of black holes is an example of such situation. Hence, by this simple argument, we expect a large class of dynamical black holes for which the inequality (4) is satisfied. However, it is not obvious how to rule out black holes that can not be treated as continuous deformation of stationary black holes (although perhaps such situation does not occur). For those cases, we can argue as follows. If the inequality is not satisfied for a dynamical black hole, then it should be possible to perturb it in the same way as above, increasing the area and preserving the angular momentum and charge. There is in principle no physical restriction to how much the area changes, as long as it increases. Hence, it should be possible to increase the area until the equality in (4) is reached. The arguments presented in Sect. 1.1 suggest that the equality is reached only for extreme black holes. And there are well known physical arguments which suggest that extremal black holes can not be produced in a finite process (Bardeen et al. 1973; Carroll et al. 2009).
By the area theorem, we know that the horizon area will increase. If we assume axial symmetry and electrovacuum, then the total angular momentum and the charge will be conserved at the quasilocal level as we see in Sect. 2. On physical grounds, one would expect that in this situation the quasilocal mass of the black hole increases with the area, since there is no mechanism at the classical level to extract mass from the black hole. In effect, the only way to extract mass from a black hole is by extracting angular momentum through a Penrose process (Penrose and Floyd 1971; Christodoulou 1970). But angular momentum transfer is forbidden in electrovacuum axial symmetry. Then, one would expect both the area A and the quasilocal mass \(M_{bh}\) to monotonically increase with time.
It is important to emphasize that the physical arguments presented above in support of (35) are certainly weaker in comparison with the ones behind the Penrose inequalities (21), (8) and (14). A counter example of any of these inequalities in axial symmetry would prove that the standard picture of the gravitational collapse is wrong. On the other hand, a counter example of (35) would only prove that the quasilocal mass (27) is not appropriate to describe the evolution of a nonstationary black hole. One can imagine other expressions for quasilocal mass, maybe more involved, in axial symmetry. On the contrary, reversing the argument, a proof of (35) will certainly suggest that the mass (27) has physical meaning for nonstationary black holes as a natural quasilocal mass (at least in axial symmetry). Also, the inequality (35) provides a nontrivial control of the size of a black hole valid at any time.
1.3 Motivation from Newtonian objects
Geometrical inequalities for ordinary matter fields have gained much interest in the recent years. These inequalities are not expected to arise in Newtonian theory, unless some specific systems or matter fields are considered, which have intrinsic restrictions. With the state of the subject at present, we can not say that geometrical inequalities for objects are produced solely by Einstein equations. Other ingredients must be considered. In particular, one needs an analog of the variational characterization of extreme black holes, that we showed in the previous sections.
What is interesting is that one obtains geometrical inequalities from Newtonian considerations if they are supplemented with a key ingredient from General Relativity, namely, that any speed in the system is not greater than the speed of light. We discuss these inspirational inequalities in what follows.
As opposed to (39), and (42), which are quasilocal inequalities, (45) is global in nature as it contains the total energy of the system.
What is remarkable is that these informal and naive arguments lead to similar inequalities that can be formally obtained from purely relativistic considerations about ordinary, nonblackhole objects. We review them in Sect. 5.
2 Basic definitions
In order to extend inequalities (2)–(4) to dynamical black holes and even to more general situations like ordinary objects we need to introduce these elements with detail, and properly define the physical quantities associated to them \((M,Q, J,A, \text{ etc })\) in the fully dynamical regime.
Black holes are the main type of object we present in this article. That is the setting that originally motivated the study of the geometrical inequalities given in this article and where the most important results have been found so far. However, there are other quasilocal objects that are relevant, namely isoperimetric surfaces and bounded regions representing ordinary objects. We present these objects in Sect. 2.2
On the other hand, the physical quantities studied in this review can be divided in three groups: local quantities, global quantities and quasilocal quantities. Local quantities are tensor fields, global quantities are associated to the whole spacetime and quasilocal quantities are associated with finite regions. We define them in Sects. 2.3, 2.4 and 2.5 respectively.
Before proceeding further with these concepts, we fix the basic notation we use throughout the article.
2.1 Asymptotically flat and cylindrical ends
The falloff conditions (49) are far from being the minimal requirements for the validity of the theorems presented in this article. We have chosen these particular falloff conditions because they are simple to present and they encompass a rich family of physical models. For more refined assumptions we will refer to the original references.
See however, Sect. 2.4 where the stronger falloff condition (71) on the second fundamental form \(K_{ij}\) is imposed. This stronger requirement is necessary to make the integral in the definition of angular momentum converge.
An initial data may have more than one asymptotically flat end, and the asymptotic conditions (49) should hold at each one of these ends.
2.2 Black holes and other objects
2.2.1 Black holes
Black holes are global concepts referring to the causal structure of the whole spacetime and therefore, can not be defined in terms of local or quasilocal quantities. This property makes practical applications difficult to study and has led to the development of quasilocal meaningful characterizations of black holes. The intuitive idea that a black hole is a region of spacetime from which no signal can escape is captured by the notion of trapped surface described below. See the articles by Beig and O’Murchadha (1996), Dain (2004), Booth (2005), Jaramillo et al. (2008), Mars (2009), Hayward (2011) and Senovilla (2011), for further references, details and discussions on these quasilocal characterizations.

marginally outer trapped surfaces (MOTSs), for which \(\theta _+=0\) on \(\mathcal {S}\),

minimal surfaces for which \(\theta _\pm =0\).
Minimal surfaces have played a key role in the study of geometrical inequalities for black holes in General Relativity since the early days and especially since the proof of the positive mass theorem (Huisken and Ilmanen 2001) . Without mention of null expansions, minimal surfaces are characterized by the vanishing extrinsic curvature when seen as surfaces \(\mathcal {S}\) embedded in a 3dimensional slice \(\varSigma \). Note also that if the slice is part of a time symmetric initial data \((\varSigma , h_{ij}, K_{ij}\equiv 0)\), then \(\mathcal {S}\) is also a MOTS. Moreover, as we show in Sect. 4.2.5, the variational characterization of stable minimal surfaces is closely related to the one for stable MOTSs. Roughly speaking, stability for minimal surfaces and MOTS (and also for isoperimetric surfaces, see Sects. 2.2.2, 5.2.3) implies a minimization of the area function under certain surface deformations (see Sect. 4.1.1 for precise definitions). The relation between stability of MOTS, black holes and black hole inequalities is discussed by Mars (2014).
With the concept of trapped regions, one can study a (globally hyperbolic) black hole spacetime as the manifold \((M=R\times \varSigma , \,g_{\mu \nu })\) where \(\varSigma \) is a spacelike Cauchy surface with trapped inner boundary \(\mathcal {S}\). This is the approach taken to study quasilocal Lorentzian inequalities in Sect. 4. Other alternative, mainly used in the study of global and quasilocal Riemannian inequalities is to consider \(\varSigma \) as a surface with one asymptotically flat end (representing the region far away from the black holes), and as many extra ends as black holes one wishes to consider (Chruściel 2008). The extra ends will be asymptotically flat if the black holes are subextremal and asymptotically cylindrical if they are extremal black holes. These extra ends are usually called punctures in the numerical community and this type of topology seems to be more appropriate to numerical simulations than 3surfaces with inner boundaries (Immerman and Baumgarte 2009). So, in a sense, the black holes are represented by nontrivial topology on the initial surface \(\varSigma \) (see Gannon 1975; Lee 1976; Meeks et al. 1982; Andersson and Metzger 2009; Eichmair 2007; Andersson et al. 2011; Chruściel et al. 2011; Eichmair et al. 2013).
2.2.2 Isoperimetric surfaces and ordinary objects
As we mention in the introduction, Sect. 1.3, geometrical inequalities for nonblack hole objects have gained impetus in recent years.
The first difficulty when studying these systems is the characterization of such ordinary objects. This problem does not appear in the black hole case where there is a well identified surface (the trapped surface) locating the black hole, to which one can associate convenient stability properties.
Away from black holes, one can consider isoperimetric surfaces. These surfaces have been studied within the context of geometrical inequalities in General Relativity, mainly in the context of Penrose inequality (Gibbons 1984, 1997; Malec 1992; Gibbons and Holzegel 2006; Corvino et al. 2007). Isoperimetric surfaces are such that its area is a critical point with respect to nearby surfaces enclosing a given volume. This variational characterization what makes them potentially useful for the study of inequalities.
Ordinary objects, on the other hand, are connected, open, bounded sets with smooth boundary in an initial spacelike hypersurface. Intuitively, an ordinary material object would be the bounded region where smooth matter fields have support. The main difficulty in this case is the lack of a variational characterization, which makes the obtention of geometrical inequalities hard to achieve. In Sect. 5.2 we describe what conditions have been imposed on the objects in order to produce the desired physicalgeometrical estimate.
2.3 Local physical quantities
The local physical quantities relevant for General Relativity are shown in Einstein equations (46) and, in particular, in Einstein constraints (47), (48). We focus in this section on the energy momentum tensor \(T_{\mu \nu }\) describing matter fields.
No specific matter model will be used, the only equation that \(T_{\mu \nu }\) is required to satisfy is the local conservation law (74)
It is important to emphasize that, unless otherwise stated, the tensors \(T^{{\textit{EM}}}_{\mu \nu }\) and \(T^M_{\mu \nu }\) are not, individually, divergence free.
Summarizing, the relevant local quantity is the energy momentum tensor \(T_{\mu \nu }\), which satisfies the local conditions (74) and (59). Two important particular cases are vacuum \(T_{\mu \nu }=0\) and electrovacuum \(T^{M}_{\mu \nu }=0\).
2.4 Global physical quantities
Global quantities are associated to isolated systems. An isolated system is an idealization in physics that assumes that the sources are confined to a finite region and the fields are weak far away from the sources. This kind of systems are expected to have finite total energy, linear momentum, angular momentum and charge. In General Relativity there are several ways of defining isolated systems. For our purpose the most appropriate definition is through initial conditions for Einstein equations. Most results concerning global inequalities discussed in this article has been proved, so far, in terms of initial conditions.
We emphasize that for every asymptotically flat end \(U_k\) we have the corresponding, in principle different, quantities \(E_{(k)}\), \( P^i_{(k)}\), \(J^i_{ (k) \infty }\), \(Q_{(k) \infty }\).
We use a subscript \(\infty \) in the notation for \(J_\infty \) and \(Q_\infty \), to distinguish them from the quasilocal quantities presented in Sect. 2.5. However, since we will not discuss geometrical inequalities involving quasilocal mass or linear momentum we will not use a subscript \(\infty \) in E and \(P^i\). We expect future extensions of inequalities (2)–(4) in this direction, giving purely quasilocal geometrical inequalities.
2.5 Quasilocal physical quantities
Quasilocal quantities depend on a bounded spacelike 3dimensional region \(\varOmega \), which can be thought as a subset of some initial data \(\varOmega \subset \varSigma \). There are two kinds of quasilocal quantities, the first ones depend only on the boundary of the region \(\varOmega \), that is a 2dimensional spacelike closed surface that we will denote by \(\mathcal {S}\). The second ones depend also on the interior of \(\varOmega \) (for more details on this classification, that is both subtle and important, see Szabados 2004). See also Wieland (2017), Chen et al. (2016), Epp et al. (2013), Tung (2009), Yoon (2004) and Nester et al. (2004). It turns out that for black holes only quasilocal quantities of the first kind are relevant. For objects, quantities of the second kind are also needed. For example, in spherical symmetry the geodesic distance to the center is a relevant quasilocal measure of size. We will present some of these measures with more detail in Sect. 5. Below, we concentrate on quasilocal quantities that depend only on 2dimensional closed spacelike surfaces \(\mathcal {S}\).
On \(\mathcal {S}\) we define intrinsic and extrinsic quantities. The former depend only on the induced Riemannian 2dimensional metric on the surface that we denote by \(q_{ij}\). The extrinsic quantities depend also on the extrinsic curvature of the surface and possible additional fields like the electromagnetic fields. The most important intrinsic quantity is the area \(A(\mathcal {S})\) of the surface. For black holes, this is certainly the most relevant intrinsic quantity. But, even for black holes, there exists also other intrinsic quantities that measure the shape of the surface and satisfy geometrical inequalities. We will present them in Sect. 4.2.
2.5.1 Conserved quantities
For the discussion of conserved quasilocal quantities, we essentially follow Sect. 2 in Szabados (2004) and Weinstein (1996), see also Dain (2014a).
In Minkowski this construction provides, for each Killing vector field \(\eta ^\mu \), the conservation law for all relevant physical quantities associated with the matter field \(T^{\mu \nu }\) (i.e. energy, linear momentum, angular momentum).
2.5.2 Electromagnetic charge
In electromagnetism, we start with the field Eqs. (90)–(91) and then we deduce (92) and hence the conservation of \(Q_E\).
In order to make contact between quantities written in terms of differential forms and other equivalent expressions used in the literature written in terms of tensors it is convenient to choose a tetrad adapted to a closed, oriented, spacelike 2surface \(\mathcal {S}\). Consider null vectors \(\ell ^\mu \) and \(k^\nu \) spanning the normal plane to \(\mathcal {S}\) and normalized as \(\ell ^\mu k_\mu = 1\), leaving a (boost) rescaling freedom \(\ell '^\mu =f \ell ^\mu \), \(k'^\mu = f^{1} k^\mu \). The induced metric and the volume element on \(\mathcal {S}\) (written as spacetime projectors) are given by \(q_{\mu \nu }=g_{\mu \nu }+\ell _\mu k_\nu +\ell _\nu k_\mu \) and \(\epsilon _{\mu \nu }=2^{1}\epsilon _{\lambda \gamma \mu \nu }\ell ^\lambda k^\gamma \) respectively. The area measure on \(\mathcal {S}\) is denoted by \(ds\).
2.5.3 Angular momentum
To define quasilocal angular momentum in general is a difficult problem (see the review by Szabados 2004). However, for axially symmetric spacetimes there exists a simple and well defined notion of quasilocal angular momentum which was introduced by Komar (1959) (see also Wald 1984). In the literature this definition is usually discussed in vacuum settings, where the angular momentum is conserved. Remarkably, it turns out that the quasilocal inequalities of the form (4) are still valid in the nonvacuum case. The inclusion of matter fields (in particular, electromagnetic fields) presents some peculiarities in the definitions and also in the discussion of the conservation (and nonconservation) of angular momentum.
We have seen in Sect. 2.5.2 that the superpotentials \(\varvec{V}\) for electric and magnetic charges are given by \(\varvec{V}_E=\varvec{F}/4\pi \) and \(\varvec{V}_B={}^*\varvec{F}/4\pi \).
Note that Eq. (119) is valid for nonzero sources \(\varvec{j}^{{\textit{EM}}}\). We discuss the case \(\varvec{j}^{{\textit{EM}}}=0\), and more generally, the case \(\varvec{j}^{{\textit{EM}}}(\bar{\eta })=0\) for axially symmetric initial data in Sects. 3.2.5 and 4.2.5.
3 Global inequalities for black holes
In this section we present results involving the total mass of a black hole and some of its quasilocal physical parameters like the electromagnetic charge and angular momentum. We group the results into two sections. The first group, in Sect. 3.1 refers to inequalities involving total mass and electromagnetic charge, with zero angular momentum. The second group incorporates angular momentum. As we discuss in Sect. 2, in order to have a well defined conserved angular momentum, axial symmetry is required, whereas the pure charge case needs no symmetry at all to be well formulated. This difference makes the techniques used to solve the problems very different as we discuss below. Before going into the details, we want to make two remarks.
Settings The geometrical inequalities presented in this section are proven for a set of initial data \((\varSigma , h_{ij}, K_{ij})\) for Einstein equations. Evolution arguments are not considered. Moreover, Einstein constraint equations are used in a crucial way. The initial surface has an asymptotically flat end, where the ADM mass is computed. Also, it may have an inner boundary, connected or not, or be complete, with at least one more end. These features capture the presence of the black hole (see Sect. 2.2). In the introduction, Sect. 1.2, we give arguments indicating that the global black hole inequalities are valid for all times if they are valid at some initial time.
Inequality producer We wish to mention here the features in the systems considered that ultimately produce the inequalities. Note that we are not thinking about the general qualities of systems in the theory that allow such inequalities (this was discussed at the beginning of the introduction). But we wonder about the underlying mathematical hypothesis or condition on the initial data \((\varSigma , h_{ij}, K_{ij})\) that translate into the desired relation between the physical and geometrical parameters. Interestingly, in order to derive the global inequalities presented in this section, the only requirement are certain energy conditions and the presence of at least two ends on \(\varSigma \) if \(\varSigma \) has no inner boundary (one of which needs to be asymptotically flat to have a well defined ADM mass). However, as we present in Sect. 3.1, the mass–charge inequality is also proven when \(\varSigma \) is an asymptotically flat manifold with inner trapped boundary. This is in contrast with the Penrose inequality (Mars 2009), which is also global, but where the area of a closed 2surface is included explicitly into the estimate and therefore, extra assumptions on such 2surface must be imposed.
3.1 Mass–charge
The mass–charge inequality arises as a way to refine the positive mass theorem (Schoen and Yau 1979, 2017, 1981; Witten 1981) and to give a strictly positive lower bound for the total mass in the Einstein–Maxwell theory.
3.1.1 Results
Below is the main theorem showing this result. Its proof is greatly due to Gibbons and Hull (1982), and to Gibbons et al. (1983), but many authors have contributed to the final version. We describe their particular input after the statement.
Theorem 1
 (i)
if the equality in (143) is attained then the associated spacetime is, locally, an Israel–Wilson–Perjés metric.
 (ii)
if the initial data are maximal (i.e. \(K=0\)) and electrovacuum, then the equality in (143) holds if and only if the data set arises from the Majumdar–Papapetrou spacetime.
 (iii)
if the Dirac–Jang equations have an appropriate solution then the equality in (143) holds if and only if the data set arises from the Majumdar–Papapetrou spacetime.
As we mention above, (143) is proven in Gibbons and Hull (1982) and Gibbons et al. (1983). They use spinorial arguments similar to ones used in Witten’s proof of the positive mass theorem (Witten 1981). Gibbons and Hull (1982) shows that the equality in (143) holds if and only if there exists a super covariantly constant spinor, and the Israel–Wilson–Perjés and Majumdar–Papapetrou metrics are discussed in this context.
Tod (1983) (see also Herzlich 1998; Horowitz 1984) addresses the rigidity statement (i) by finding all smooth spacetimes admitting such super covariantly constant spinors. They are gravitational and electromagnetic plane waves possibly with dust, and metrics describing charged rotating dust. Particular cases of the latter are the Israel–Wilson–Perjés metrics, some of the Bonnor metrics and the Majumdar–Papapetrou metric.
Chruściel et al. (2006b) prove the rigidity statement (ii) by showing that under certain conditions, the Israel–Wilson–Perjés metrics are of Majumdar–Papapetrou class.
Bartnik and Chruściel (2005) generalize the proof of (143) to include low differentiable metrics, namely \(h_{ij}\in H^2_{loc}\) and \(K_{ij}, E^i, B^i \in H^1_{loc}\). However, the equality is left open, as the classification of metrics admitting super covariantly constant spinors of Tod (1983) does not apply in the rigidity case.
The rigidity statement (iii) is proven by Khuri and Weinstein (2013) assuming that a system of equations, namely the Dirac–Jang equations, has appropriate solutions. They also assume what they call the charged dominant condition, which reads \(\rho _{{\textit{EM}}}\ge \mu _Mj_M\) and is stronger than condition (142).
3.1.2 Discussion
A few comments about the hypotheses and applications follow.
The manifold \(\varSigma \) can have multiple asymptotic ends, on different ends the value of the quantities \(E,P, Q_\infty \) are different and the inequality (143) holds on every end. Also, the theorem admits a manifold with an inner boundary given by the 2surface \(\mathcal {S}\). Consider the simplest case, where the manifold \(\varSigma \) is \(\mathbb {R}^3\) (which, of course, means that it has only one asymptotic end and no inner boundary). For that case, in order to have nonzero \(Q_\infty \) we need to have charged matter in the interior, that is \(\rho _E \ne 0\). The value of \(Q_\infty \) represents the total charge of the spacetime, it will be in general different than the value of the charge calculated for an arbitrary surface \(Q(\mathcal {S})\) in the interior. An important spacetime that satisfies these conditions is the electrically counterpoised dust studied by Bonnor (see Bonnor 1980, 1998 and reference therein). These are explicit solutions that describe static configurations of charged dust with \(\mu _M=\rho _E\), and hence the gravitational attraction is exactly balanced by the electric repulsion. The shape of the configuration can be arbitrary and does not need to have any spatial symmetry. These spacetimes achieve the equality \(M=Q_\infty \) in (143).
If we assume electrovacuum, in order to have a nonzero \(Q_\infty \) we need to allow either a nontrivial topology with multiple asymptotic ends or a nontrivial inner boundary \(\mathcal {S}\). The Reissner–Nordström black hole initial data are the model example of both cases: either we consider them as a complete manifold with two asymptotically flat ends or as a manifold with inner boundary \(\mathcal {S}\) at the black hole horizon which is weakly outer trapped. The equality \(M=Q_\infty \) is achieved by the extreme Reissner–Nordström black hole, which has one asymptotically flat end and one cylindrical end. On the other hand, initial data of the super extreme Reissner–Nordström metric does not satisfy the hypotheses since the Riemannian manifold \((\varSigma ,h_{ij})\) is not complete and does not have any weakly outer trapped 2surface.
The Majumdar–Papapetrou spacetime also satisfies the hypotheses of Theorem 1. The initial data has only one asymptotically flat end and an arbitrary number of extra cylindrical ends. This spacetime describes the equilibrium configuration of multiple extreme black holes for which the gravitational attraction is balanced by the electric repulsion. It satisfies \(M=Q_\infty \) (see Hartle and Hawking 1972 for a discussion of these spacetimes). The extreme Reissner–Nordström black hole is a particular case of Majumdar–Papapetrou with only one cylindrical end.
The Israel–Wilson–Perjés metrics are characterized by the existence of a ‘supercovariantly constant’ spinor field (for details about this metrics see Tod 1983; Chruściel et al. 2006b and references therein). Example of this class of metrics are the above mentioned electrically counterpoised dust and the Majumdar–Papapetrou metrics. We emphasize that the equality in (143) can be achieved by a nonelectrovacuum solution. However, for the electrovacuum case a stronger result is given in statement (i).
3.2 Mass–angular momentum–charge
The inclusion of the angular momentum in the inequality (3) involves completely different techniques as the one used in Theorem 1. In particular no spinorial proof of this inequality with angular momentum is available so far (see however Zhang 1999, where a related inequality is proven using spinors).
3.2.1 Results
Below is the mass–angular momentum–charge theorem, the proof in vacuum is mainly due to Dain (2008). Some of the later generalizations and refinements by other authors are included in the statement and discussed after it.
Theorem 2
Let \((\varSigma , h_{ij}, K_{ij}, E^i, B^i)\) be an electrovacuum, axially symmetric, maximal, initial data set with two asymptotic ends. One end is asymptotically flat, where the falloff condition (25) is assumed for the second fundamental form. The other end is asymptotically flat or cylindrical.
The proof of the inequality (146) was provided by Dain in a series of articles (Dain 2006a, b, c), which end up in the global proof given in Dain (2008). There it is shown that (146) holds in vacuum and for a class of axially symmetric black hole initial data known as Brill data. The argument exploits the relation between a certain mass functional \(\mathcal {M}(\eta ,\omega )\) and the energy for harmonic maps from \(\mathbb {R}^3\) to the Hyperbolic plane (Carter 1973; Ernst 1968) (\(\eta \) and \(\omega \) are the norm and twist potential of the axial Killing vector). See Sect. 3.2.5 for the definition and properties of \(\mathcal {M}\) and details about the proof of Theorem 2.
Along the same lines, Chruściel (2008) and Chruściel et al. (2008) give a simpler proof of the theorem and avoid some technical assumptions on Dain’s proof but consider a class of axially symmetric initial data that do not contain the limit case in vacuum, namely, extreme Kerr. They allow positive matter density that does not enter explicitly into the inequality. Moreover, they assume the existence of a twist potential \(\omega \) which is known to exist in vacuum.
Electromagnetic charge is brought into the inequality by Chruściel and Lopes Costa (2009) and Costa (2010). In particular, in Costa (2010) an appropriate potential, related to the twist potential for the axial Killing vector is shown to exist. This allows the definition of a new mass function \(\mathcal {M}(\eta ,\omega ,\psi ,\chi )\), where \(\eta \) and \(\omega \) are as before and \(\psi \), \(\chi \) are the electromagnetic potentials (see Sect. 3.2.5). As in Chruściel et al. (2008), the initial data considered does not include extreme Kerr–Newman and therefore, the rigidity statement can not be analyzed.
Schoen and Zhou (2013) make several improvements on the assumptions, they give a relevant lower bound for the difference between general data and the extreme Kerr–Newman data and they prove the rigidity statement in the charged case.
In Theorem 2 the data are assumed to be maximal. For the case with no charge, the maximal condition is relaxed to a small trace case assumption by Zhou (2012).
The maximality condition is replaced in Cha and Khuri (2014, 2015) by the assumption that a system of equations has appropriate solutions. Also, nonelectromagnetic matter fields \(\mu _M,j_M\) that satisfy the charged dominant energy condition \(\mu _M\ge j_M\) are included in the hypotheses. In these articles a deformation procedure of the initial data is constructed that provides a natural and clean way to automatically generalize geometrical inequalities proved in the maximal case.
The key role that extreme Kerr–Newman black holes play in giving the limit values for inequality (146) (and, as we will see later, in the quasiliocal inequalities for black holes) is confirmed and reinforced by the gedanken experiments performed by Sorce and Wald in Sorce and Wald (2017). There it is proven that an extremal Kerr–Newman black hole can not be overcharged or overspun.
3.2.2 Discussion
Before presenting related results to Theorem 2, a few remarks are in order.
We begin by comparing Theorem 2 and the pure charge Theorem 1. In Theorem 2 axial symmetry is assumed globally on the initial data, in contrast with Theorem 1 where there are no symmetry assumptions. As we have seen in Sect. 1.2, on physical grounds, the inequality (146) is not expected to hold for nonaxially symmetric data. General families of nonaxially symmetric counter examples of the inequality (146) have been constructed by Huang et al. (2011) for pure vacuum and complete manifolds.
In Theorem 2 electrovacuum is assumed. It is conceivable that, using similar techniques as in the current proof of this theorem, the electrovacuum assumption can be slightly relaxed by assuming, in analogy with the Assumption (142) in Theorem 1, that the matter fields have small angular momentum and charge. We expect that this assumption will be rather unphysical, since ordinary rotating matter can easily violate the inequality (146). It is however interesting that in Theorem 1 a rigidity statement is obtained even in the nonelectrovacuum case. It is not known whether an analogous rigidity result holds for axially symmetric matter fields with nontrivial angular momentum in which the equality in (146) is achieved. Note however, that the rigidity statement in Theorem 1 depends strongly on the spinorial proof. It appears to be unlikely that a generalization of this rigidity result holds for the case of angular momentum.
On the other hand, we expect that the dominant energy condition should be required, as there are numerical examples by Bode et al. (2011) of spinning black holes with matter violating the null (and dominant) energy condition, with \(J\ge M^2\).
No inner boundary is allowed in Theorem 2. The inclusion of an inner boundary (which presumably should be a weakly trapped surface as in Theorem 1) appears to be a difficult and a relevant problem. An inner boundary requires appropriate boundary conditions for the variational problem used in the proof of Theorem 2. The results presented by Gibbons and Holzegel (2006) and Chruściel and Nguyen (2011) contribute in this direction, but so far the problem remains open.
Since electrovacuum is assumed, in order to have nonzero charge and angular momentum the manifold should have a nontrivial topology. In Theorem 2 a particular geometry is assumed: manifolds with two ends. This is certainly the stronger restriction of this theorem. Let us discuss this point in detail.
The model initial data set that satisfies all the hypotheses of Theorem 2 is a slice \(t=constant\) in the standard Boyer–Lindquist coordinates of the Kerr–Newman black hole. In the nonextreme case these initial data have two asymptotically flat ends, where the standard falloff conditions (49) are satisfied, plus the stronger falloff condition (71) of the second fundamental form. However, in the extreme case, the geometry of the initial data changes: one end is cylindrical and the other is asymptotically flat. That is, in order to include the extreme case more general falloff conditions need to be allowed on one of the ends.
3.2.3 Multiple black holes
For multiple ends the problem is open. There exist however the following very interesting results. In order to describe them, we need to highlight some properties of the mass functional \(\mathcal {M}\) (see Sect. 3.2.5 for more details). This functional represents a lower bound for the mass. Moreover, the global minimum of this functional (under appropriate boundary conditions which preserve the angular momentum) is achieved by a harmonic map with prescribed singularities. This is the main strategy in the proofs of all the previous theorems which are valid for two asymptotic ends. Remarkably enough, Chruściel et al. (2008) prove the existence and uniqueness of this singular harmonic map for manifolds with an arbitrary number of asymptotic ends, and then, as a corollary they prove the following result
Theorem 3
This theorem reduces the proof of the inequality with multiples ends to compute the value of the mass functional on the corresponding harmonic map.
Khuri and Weinstein (2016) extend the result by Chruściel et al. (2008) in that electromagnetic charge is included and weaker falloff conditions are assumed on the extrinsic curvature. Also, the rigidity statement is proven.
Some numerical calculations have been made to get insight about the value \(\mathcal {M}(J_2,\ldots , J_N)\). Dain and Ortiz (2009) perform numerical calculations of the mass functional mentioned above (see Sect. 3.2.5) and find evidence for the validity of an intermediate inequality of the form \(\mathcal {M}\ge J_1\), where \(J_1\) is the angular momentum of a system of two Kerr black holes with positive individual masses, computed at the end 1 in the notation of Theorem 3. Following this result, CabreraMunguia et al. (2010) work on the Tomimatsu and Dietz–Hoenselaers solution describing two Kerr black holes, one of which has negative mass. They find that there is a rank in the parameters of the individual black holes such that the total mass is smaller than the total angular momentum, that is \(M^2{<}J_1\) and therefore \(\mathcal {M}^2{<}J_1\), which opposes to Dain and Ortiz result. These are intesting results that open up the questions of what the reasonable hypotheses on the multiple black hole system should be and of what the precise form the inequality should have. Since the Tomimatsu and Dietz–Hoenselaers solution has a naked singularity, one might expect that inequality \(M^2\ge J_1\) holds for regular multiple black hole solutions. But on the other hand, there is the possibility of conjecturing and proving a different inequality involving not only ADM mass and angular momentum (whether total or individual), but also other phyisical parameters like separation distance between black holes or some other properties of the system. This is certainly an interesting open problem.
3.2.4 Nonasymptotically flat manifolds
Recently, global inequalities for asymptotically hyperboloidal initial data have started to be explored. Cha et al. (2016) generalize a procedure used by Schoen and Yau (1981) which consists in a deformation that transforms an asymptotically hyperboloidal structure into an asymptotically flat one. By doing this, they are able to use the geometrical inequalities known to hold in the asymptotically flat case, to prove them in the hyperboloidal case. More precisely, they consider a smooth, simply connected, axially symmetric initial data satisfying the charged dominant energy condition and the matter condition \(\mu _{M}\ge j_{M}\) and the matter condition \(j_M^i\eta _i=0\). The data is assumed to have two ends, one asymptotically hyperboloidal and the other either asymptotically flat or asymptotically cylindrical. If certain system of equations, consisting of a Janglike equation, and two extra equations on the deformed data, admits a smooth solution with prescribed asymptotics, then inequality (146) holds. Moreover, if equality is attained, then the initial data arise from an embedding into the extreme Kerr–Newman spacetime. This result effectively reduces the proof of the geometrical inequality to proving existence of solutions to certain equations with appropriate falloff conditions.
Cha and Khuri (2017) apply the same arguments to obtain an identical global inequality to (146) in the case that the initial data has two ends, one AdS hyperbolic and the other either asymptotically AdS hyperbolic or asymptotically cylindrical. The cosmological constant is not, however, explicitly included into the inequality. See Sect. 3.3 for a conjectured global inequality including cosmological constant.
3.2.5 The mass functional \(\mathcal {M}\)
We present in this section the main arguments behind the mass–angular momentum inequalities (146) and (147). They are heavily based on a certain mass functional \(\mathcal {M}\) and its relation with the energy of harmonic maps. For simplicity we assume electrovacuum and maximality to present the basic properties of \(\mathcal {M}\) that are crucial for proving (146). We also assume only two ends on the initial data.
The proof consists of two steps. The first one is to prove \(m\ge \mathcal {M}\) for the given black hole initial data, using the Hamiltonian equation and some energy conditions. The second step is to prove that extreme Kerr–Newman black hole with the same angular momentum as the given data, is a minimizer for \(\mathcal {M}\), that is \(\mathcal {M}\ge \mathcal {M}_{\text{ extr } \text{ KerrNewman }}\) using known results on harmonic maps with prescribed boundary conditions. Let us see this in more detail.
Step 1. \(m\ge \mathcal {M}\)
Step 2. \(\mathcal {M}\ge \mathcal {M}_{{\text{ extr } \text{ KerrNewman }}}\)
We can apply the results of Hildebrandt et al. (1977) (see also Chruściel 2008) stating that when \(\varOmega \) is compact and does not contain the axis \(\varGamma \), and the target manifold has negative curvature (as \(\mathbb {H}^2_{\mathbb {C}}\) in our case), the minimizers of \(\tilde{\mathcal {M}}_\varOmega \) with Dirichlet boundary conditions exist, are unique, smooth and satisfy the Euler Lagrange equations.
Since the difference between \(\mathcal {M}_\varOmega \) and the harmonic energy \(\tilde{\mathcal {M}}_\varOmega \) is the boundary term \(B_{\partial \varOmega }\), the minimizer of \(\tilde{\mathcal {M}}_\varOmega \) is the minimizer of \(\mathcal {M}_\varOmega \) as well. This minimizer is the extreme Kerr–Newman solution, and we obtain \(\mathcal {M}_\varOmega \ge \mathcal {M}_\varOmega _{\text{ ext } \text{ KerrNewman }}\).
After a subtle limit procedure that allows to extend the inequality valid in \(\varOmega \) to all \({\mathbb {R}}^3\), one arrives at the desired inequality \(\mathcal {M}\ge \mathcal {M}_{\text{ extr } \text{ KerrNewman }}\).
3.3 Cosmological constant
Related inequalities are given by Chruściel et al. (2006a), which is an extension of a previous work by Maerten (2006).
4 Quasilocal inequalities for black holes
The geometrical inequalities presented in this section relate purely quasilocal quantities defined on a closed 2surface. By this we mean that no bulk quantities defined on the 3dimension region inside the surface are considered.
As is the case with global inequalities, here we divide the subject in three parts, the pure charge case, in Sect. 4.1, the inequalities involving angular momentum, in Sect. 4.2 and the inequalities involving (explicitly) a cosmological constant in Sect. 4.4. We do this for two reasons. First, the pure charge case problem does not need axial symmetry to be formulated, whereas it is needed in order to have a well defined quasilocal angular momentum (see Sect. 2). Also, even when the electric charge gives a flavor and a hint of what may happen with angular momentum, the techniques employed in the three treatments are usually different.
Let us analyze the settings and the factors that produce these quasilocal inequalities.
Settings As opposed to the global inequalities, which can be proven for complete initial surfaces, quasilocal inequalities need a well identified surface representing the black hole, where things are computed. Even then, the problem can be formulated from two different perspectives: a Riemannian and a Lorentzian points of view. These two approaches have to do with the different surfaces used in the derivation of the inequality. In the Riemannian setting, a minimal surface in an initial data set is studied. In the Lorentzian case, it is a MOTS in spacetime
Inequality producer In the quasilocal inequalities some form of a stability condition for the 2surface considered is the main factor that produces the estimate. The Riemannian treatment requires the positivity of the second variation of the area function. In the Lorentzian setting the inequalities arise from stability of the MOTS.
4.1 Area–charge
The relation between the area and the electromagnetic charges of a closed 2surface was the first result obtained in the form of a quasilocal geometrical inequality.
It was first studied in a spacetime setting where trapping properties of 2surfaces embedded in a Cauchy surface were assumed. Later, the inequality was proved purely from an initial data view point. And finally, from a purely Lorentzian one. We state a precise version of the result in the latter form below, and then discuss the different contributions, the hypotheses and conclusions.
4.1.1 Results
We extract the following theorem from the work of Dain et al. (2012).
Theorem 4
We define the stability condition and address the other hypotheses below. Let us first review the different previous results on the subject.
Gannon (1976) obtains an inequality in the spirit of (169) when analyzing properties of electrovacuum black hole spacetimes that are strongly future asymptotically predictable from a partial Cauchy surface \(\varSigma \) regular near infinity. This means that \(\varSigma =\cup _{i=1}^\infty \varOmega _i\), with \(\varOmega _i\subset \varOmega _{i+1}\) such that \(\partial \varOmega _i:=\mathcal {S}_i\) are future inner trapped surfaces (i.e. \(\theta _<0\)). These 2surfaces are expected to be found in most isolated gravitating systems after moving sufficiently far from the center of the system and are the alternative to stable MOTS and minimal surfaces used by later authors. In this setting, Gannon proves that if the boundary of the black hole can be foliated by spacelike twosurfaces whose surface area is bounded above by \(A_{\max }\), then \(A_{\max }\ge 4\pi Q^2\), provided the electric charge Q is not zero (this restriction guarantees that the 2surfaces are topological spheres). Strictly speaking, Gannon’s result is not quasilocal as it makes assumptions about all \(\varSigma \). Nevertheless, we present the result here because of the great role it plays in the subject as a model and inspiration to subsequent work.
Later, Gibbons (1999), using the positivity of the second variation of the area function, obtains the inequality (169) as a particular case of results in higher dimensions. For our purposes here, he considers maximal, electrovacuum initial data with a stable minimal surface \(\mathcal {S}\). Recall that given a maximal initial data \((\varSigma , h_{ij}, K_{ij}, \mu , j^i)\), and an embedded surface \(\mathcal {S}\) in \(\varSigma \), we say it is a minimal surface if the mean curvature of \(\mathcal {S}\) vanishes and it is stable if the second variation of the area function is nonnegative, namely \(\delta ^2_{\alpha s}A\ge 0\) for all functions \(\alpha \), where \(s^i\) is a normal vector to \(\mathcal {S}\) in \(\varSigma \) (see Gibbons 1999, also Andersson et al. 2008a; Mars 2014). The stability condition gives the desired inequality (169) between the area A and the charge Q of \(\mathcal {S}\). This result is a refinement of a previous work of Gibbons et al. (1996) in the spherically symmetric case.
In the Lorentzian settings, Dain et al. (2012) study the inequality between area and charge of a MOTS in a spacetime, satisfying certain stability property. Let us see it in some detail. Andersson et al. (2005, 2008b) introduce a notion of stability for a closed marginally trapped surface \(\mathcal {S}\) which motivates the notion used in Theorem 4. \(\mathcal {S}\) is said to be spacetime stably outermost if there exists an outgoing vector \(X^\mu =f_\ell \ell ^\mu f_k k^\mu \) with functions \(f_\ell \ge 0\) and \(f_k>0\) such that the variation \(\delta _{X}\) of \(\theta _+\) with respect to \(X^\mu \) satisfies \(\delta _{X}\theta _+\ge 0\), where \(\delta \) is the variation operator associated with a deformation of \(\mathcal {S}\). Charged matter fields are included, although the nonelectromagnetic matter fields must satisfy the dominant energy condition. Extensions of this inequality are also proven for regions in the spacetime which are not necessarily black hole boundaries, but ordinary objects (see Sect. 5 for more details). They prove Theorem 4. They also prove a similar inequality to (169) for an oriented surface screening an asymptotically flat end. A screening surface of an end is a closed 2surface that encloses an open, connected region \(\varOmega \subset \varSigma \) which contains the mentioned end and no other.
As noted by Jaramillo (2013), the area charge inequality relies on the algebraic properties of the electromagnetic energy momentum tensor. This observation provides a straightforward generalization of (169) to other matter fields having similar algebraic properties. In particular, in Jaramillo (2013), the inequality (169) is extended to include the Yang–Mills charges.
4.1.2 Discussion
We wish now to make a few observations about this result.
As in the global case for the inequality between mass and charge (see the beginning of Sect. 3), no axial symmetry is required to obtain (169) due to the area and charge being well defined quasilocal quantities.
On the other hand, as opposed to the global inequalities involving the ADM mass, (169) is a quasilocal relation, it refers to the properties of one 2surface describing the horizon. This means that if one considers a spacetime containing many black holes, that inequality should hold for each one of them. The paradigmatic example of this case is the Majumdar–Papapetrou solution (Majumdar 1947; Papapetrou 1945), which consists of an arbitrary number of extreme Reissner–Nordströmlike black holes, all with charges of the same sign. This is a very special solution as each black hole saturates the bound (169), i.e., \(A_i=4\pi Q_i^2\), where \(A_i\) and \(Q_i\) are the individual areas and charges respectively of each black hole.
The Majumdar–Papapetrou solution is static, nevertheless, inequality (169) holds in completely dynamical scenarios as well, even in the presence of charged matter fields satisfying the dominant energy condition. This is in contrast with the mass–charge inequality presented in Sect. 3.1, where a strong (and in a sense, unnatural) local condition is imposed on matter fields [see the discussion after Eq. (145)].
Note that there is no rigidity statement saying that if equality is attained in (169), then the solution must be the near horizon geometry of Reissner–Nordström.
In fact, the key ingredient used to prove Theorem 4 is the notion of stability for the 2surface. This condition plays an analogous role as the nonnegativity condition on the second variation of area function in Riemannian settings.
4.2 Area–angular momentum–charge
At the beginnings of 2007, two results relating horizon area and angular momentum of black holes were given, one for stationary black holes by Ansorg and Pfister and the other for isolated horizons by Booth and Fairhurst (see the Living Review by Ashtekar and Krishnan 2004 for definitions and general results on isolated horizons). These two works have motivated the more general study of quasilocal inequalities that explicitly include the angular momentum of a given surface in a dynamical system.
The different settings where the desired inequality have been proven are stationary spacetimes, maximal initial data and finally trapped surfaces.
4.2.1 Results
We present here one of the results, taken from the article by Gabach Clément et al. (2013) that is valid for dynamical as well as stationary black holes represented by an appropriate closed 2surface.
Theorem 5
 1.
a smooth spacetime stably outermost axisymmetric marginally outer trapped surface (MOTS) embedded in a spacetime, satisfying the dominant energy condition, or
 2.
a smooth stable axisymmetric minimal surface in a maximal data set, with nonnegative scalar curvature,
Moreover, the equality in (170) is achieved if and only if the surface is the extreme Kerr–Newman sphere.
We review the many results that led to this theorem and discuss the theorem afterwards, in Sect. 4.2.2.
The first, known to us, work on a geometrical inequality in the spirit of (170) is due to Ansorg and Pfister (2008). They treat stationary and degenerate black holes and find that they must satisfy the equality in (170). More precisely, the equality in (170) holds for every element in a parametric sequence of axially and equatorially symmetric, stationary systems consisting of a degenerate black hole surrounded by matter such that the limit system is the Kerr–Newman black hole. Moreover, the authors conjecture that for axially and equatorially symmetric, stationary black holes surrounded by matter, the inequality (170) should hold, with equality at the degenerate case.
The Ansorg and Pfister conjecture is finally proven in vacuum by Hennig et al. (2008, 2010), and Ansorg et al. (2011). They show that every axially symmetric and stationary black hole with surrounding matter satisfies (170) and equality holds if the black hole is extremal. The proof consists in showing that if \(A\le \sqrt{(8\pi J)^2+(4\pi Q^2)^2}\), then the black hole can not be subextremal in the sense of Booth and Fairhurst (2008). Recall that a black hole is subextremal if there exist trapped surfaces in every small interior vicinity of the event horizon. Einstein equations near the horizon are then considered and a variational problem is formulated and solved. As we discuss in Sect. 4.2.2 this stationary variational problem is closely related to the variational problem arising in the dynamical regime.
Dain (2010) conjectures the validity of the vacuum case of (170) for the connected component of the apparent horizon in a dynamical scenario. The proof was given later, as in the pure charge case, from two perspectives, one Riemannian and one Lorentzian.
With a Riemannian approach, Aceña et al. (2011) prove that extreme Kerr initial data is the global minimum for certain mass functional \(\mathbb {M}\) related to the second variation of the area functional and analogous to the mass functional \(\mathcal {M}\) presented in Sect. 3.2.5. They prove the inequality (170) with \(Q=0\) for vacuum axially symmetric initial data containing a minimal surface and such that the metric \(h_{ij}\) satisfies certain technical conditions. However, this class considered includes many known black hole initial data. They extend the validity of the inequality to include a nonnegative cosmological constant, not appearing explicitly into the inequality though. However, this generalization is relevant because there exists a counterexample of the inequality (170) with \(Q=0\) for the case of negative cosmological constant, as it was pointed out in Booth and Fairhurst (2008).
In Gabach Clément (2011), the author relaxes some of the restriction on the type of surfaces studied and Dain and Reiris (2011) prove the inequality (170) with \(Q=0\), in vacuum, replacing the technical conditions of Aceña et al. (2011) by the stability condition on the minimal surface (see Sect. 4.1.1 for definition of stable minimal surface).
With a Lorentzian treatment, Jaramillo et al. (2011) prove (170) for an axially symmetric closed marginally trapped surface \(\mathcal {S}\) satisfying the spacetime stably outermost condition (see Sect. 4.1.1), in a spacetime with nonnegative cosmological constant and matter fields satisfying the dominant energy condition. The rigidity statement with the extreme Kerr sphere, instead of the extreme Kerr–Newman sphere is also proven for the \(Q=0\) case.
The inclusion of electromagnetic charges is done by Gabach Clément and Jaramillo (2012) and by Gabach Clément et al. (2013). Through the introduction of appropriate electromagnetic and angular momentum potentials, they prove Theorem 5 in two different ways and show the connection with the variational problem for the stationary case (see the remarks below). Also a relation between the two mass functionals \(\mathcal {M}\) introduced in Sect. 3.2.5, Eq. (160) and \(\mathbb {M}\) is pointed, which suggests a relation between the global inequality (146) and the quasilocal inequality (170). We explore this in Sect. 4.2.5 with more detail.
In Gabach Clément (2012), an extension of (170) to systems of many black holes with struts are obtained, and in Manko et al. (2013, 2014), CabreraMunguia et al. (2013) and CabreraMunguia (2015), binary systems saturating this inequality are presented.
4.2.2 Discussion
We wish to make a few remarks about the hypotheses and statements of Theorem 5.
Theorem 5 calls for stable closed surfaces. As we mention at the beginning of Sect. 4, it is this stability property what drives the inequalities. Both concepts of stability appeared already in problem with no angular momentum, see Sect. 4.1 for definitions. The only extra assumptions we make in this result is that the functions \(\alpha \) and \(f_\ell , f_k\) entering the stability criteria for minimal surfaces and MOTSs respectively must be axially symmetric. In Sect. 4.2.5 the connection between these two stability conditions is revised. The interesting point made in Gabach Clément et al. (2013) is that the corresponding stability assumptions both for minimal surfaces and MOTS lead to exactly the same integral condition.
The theorem admits a nonnegative cosmological constant, but it does not enter explicitly into the inequality. The treatment of such problem needs different techniques and is reviewed in Sect. 4.4. The relevant property of the cosmological constant in this theorem is that its positivity allows one to disregard it altogether from the Einstein or constraint equations. Clearly, the same can not be made for negative \(\varLambda \) and the problem of determining and proving the appropriate inequality for negative \(\varLambda \) is still open (see Sect. 4.4).
Nonelectromagnetic matter fields are admitted in the hypotheses of Theorem 5 and they are not required to satisfy the dominant energy condition. Only the complete energy momentum tensor \(T_{\mu \nu }^M+T_{\mu \nu }^{{\textit{EM}}}\) must satisfy the dominant energy condition. Also there can be matter surrounding and crossing the surface \(\mathcal {S}\). This in particular extends the results in Ansorg et al. (2011) and Hennig et al. (2010).
A major difference between Theorems 4 and 5 is the rigidity statement. The horizon in extreme Reissner–Nordström clearly saturates inequality (169), but so far it is not proven that it is the only horizon that does. On the other hand, the horizon in extreme Kerr–Newman is the unique solution that satisfies the equality in (170). In the latter case, it is the connection between a certain mass functional \(\mathbb {M}\) and the energy of harmonic maps and the uniqueness of minimizers of that energy what ultimately gives uniqueness in Theorem 5. See Sect. 4.2.5 for details.
The extreme Kerr–Newman sphere mentioned in Theorem 5 has a precise meaning in terms of intrinsic and extrinsic quantities defined on the surface \(\mathcal {S}\) (Gabach Clément et al. 2013). Basically the surface has the geometry of a horizon section in the extremal Kerr–Newman black hole. In Sect. 4.2.5 we give proper definitions, here, however, we want to emphasize that the rigidity statement refers to the extreme Kerr–Newman horizon, not the entire initial data. Interestingly, the fact that the equality in (170) is only attained by the extreme Kerr–Newman sphere has been known since the work of Hájiček (1974), Lewandowski and Pawlowski (2003), and Kunduri and Lucietti (2009, 2013) on isolated horizons and near horizon geometries of extreme black holes. See also the more recent results of Reiris (2014a) and Chruściel et al. (2017).
4.2.3 Area products
4.2.4 Shape of black holes
4.2.5 The mass functional \(\mathbb {M}\)
Analogous to the global case, the proof of the quasilocal inequality (170) is based on some remarkable properties of a quasilocal mass functional \(\mathbb {M}\) and consists of two intermediate inequalities for \(\mathbb {M}\). Again, for simplicity, we assume electrovacuum.
The proof consists of two steps. The first one is to prove a lower bound on the area of the 2surface \(\mathcal {S}\) in terms of a mass functional \(\mathbb {M}\). One starts with the appropriate stability condition (for either type of surface, minimal or marginally outer trapped) and proves \(A\ge 4\pi e^{\frac{\mathbb {M}8}{8}}\). The second step is to prove that extreme Kerr–Newman horizon is the unique minimizer for \(\mathbb {M}\), that is \(\mathbb {M}\ge \mathbb {M}_{\text{ extr } \text{ KerrNewman }}\). We follow Gabach Clément et al. (2013).
Step 1. \(A\ge 4\pi e^{\frac{\mathbb {M}8}{8}}\)
The two inequalities (178) and (179) become identical when the Hamiltonian constraint (149) is inserted in (178) and the relation \(\varUpsilon ^{(\eta )}_\mu \eta ^\mu =K_{\mu \nu }\eta ^\mu s^\nu \) is considered. Here \(s^\mu \) is a spacelike normal to \(\mathcal {S}\).
One of the most remarkable property of the functional \(\mathbb {M}\) is that the boundary conditions for the functions it depends on are the angular momentum and charges (185) and the area (181). This is especially relevant for the formulation and solution of the variational problem in the next step.
Step 2. \(\mathbb {M}\ge \mathbb {M}_{\text{ extr } \text{ Kerr } \text{ Newman }}\)
The proof of the inequality (189) can be approached in several ways, as presented in Gabach Clément and Jaramillo (2012) and Gabach Clément et al. (2013). We already commented the reduction to the variational problem in stationary settings. Here we mention the other two arguments.
Finally a very subtle limit procedure must be performed to arrive at the desired inequality \(\mathbb {M}\ge \mathbb {M}_{\text{ extr } \text{ Kerr } \text{ Newman }}\).
It is worth mentioning that the previous variational problem can be solved without assuming axial symmetry on the functions \(\sigma ,\omega , \psi ,\chi \).
The second approach is restricted to axial symmetry and hence in the minimization problem for \(\mathbb {M}\), the Euler–Lagrange equations reduce to a system of ordinary differential equations. When solving these equations, the boundary conditions J, \(Q_{E}\) and \(Q_{ M}\) determine uniquely the boundary conditions for the remaining potential \(\sigma \). This is the key fact under the sharpness of inequality (170). A constructive explicit proof of existence and uniqueness for the minimizer of \(\mathbb {M}\) is given in Gabach Clément et al. (2013) with prescribed values of \(J, Q_{ E}, Q_{ M}\) and without any reference to the boundary values of \(\sigma \). This is different to what one does in the first approach discussed above, where the boundary values of \(\sigma \) are prescribed from the relation \(A=4\pi e^{\sigma }_{\theta =0}\) valid for the particular coordinate system employed.
4.3 Relation between \(\mathcal {M}\) and \(\pmb {\mathbb {M}}\)
It is remarkable that both, the global (146) and the quasilocal (170) inequalities involving angular momentum are derived from mass functionals \(\mathcal {M}\) and \(\mathbb {M}\) respectively, and that these functionals are minimized by some form of the extreme Kerr–Newman solution. In Gabach Clément et al. (2013), a connection between these two functionals is presented, which in turn, gives a connection between the two inequalities.
4.4 Cosmological constant
The results of Sects. 4.1 and 4.2 admit nonnegative cosmological constants, but do not include them explicitly into the inequalities. These results are presented in this section as, in general, they require different techniques.
By analyzing explicit solutions and collapsing black holes, Hayward et al. (1994) and Shiromizu et al. (1993), prove that a positive cosmological constant sets restrictions on how large a black hole can be (see also Maeda et al. 1998). They study blackhole spacetimes with positive cosmological constant \(\varLambda \), that satisfy the dominant energy condition, and find that the area of the black hole horizon, as described by an outer marginal surface, is bounded as \(A\le 4\pi /\varLambda \). The same inequality holds for the area of a connected section of the event horizon in the case of strongly future asymptotically predictable, asymptotically de Sitter spacetime. The inequality is saturated for the extreme Schwarzschild–de Sitter horizon.
For negative \(\varLambda \), Gibbons (1999) (in the time symmetric settings) and Woolgar (1999) (in the nontimesymmetric case) find the bound \(A>4\pi (g1)/\varLambda \), for the area of an outermost MOTS of genus \(g>1\).
These inequalities show the important role that the cosmological constant plays in determining the size of a black hole. Note in particular, that the positive and negative cosmological constants bound the area in opposite directions. Namely a de Sitterlike black hole can not be too large and an anti de Sitterlike black hole can not be too small. This has interesting implications for studying possible colliding scenarios.
Gibbons (1999) also considers the combined effect of a cosmological constant and matter fields satisfying the dominant energy condition. He finds that the area of a stable minimal surface \(\mathcal {S}\) in a time symmetric 3surface is bounded as \(4\pi (1g)\varLambda A\int _S8\pi T_{00}>0\), where \(T_{00}\) is the energy density of the matter fields on the 3surface.
The inequalities (195)–(196) are saturated in spherical symmetry by the Reissner–Nordström–de Sitter black holes if and only if the surface gravity vanishes. Simon also discusses the time evolution of MOTSs and the application of his inequalities as restrictions on the merging. He deduces an interesting Corollary, nontrivial only when \(\varLambda \ne 0\), which gives lower and upper bounds on the quotient between the initial and final areas of MOTSs (homologous MOTSs). This result is applied to the situation where only a single MOTS is initially present, and to the problem of merging of MOTSs.
Inequality (200) was proven by Bryden and Khuri (2017) following the same ideas as in Gabach Clément et al. (2015), but simplifying the resolution of finding the minimizer of \(\mathbb {M}^\varLambda \). The argument is based on the result of Schoen and Zhou (2013), which states that \(\mathbb {M}^\varLambda \) is convex along geodesic deformations within \(\mathbb {H}^2_{\mathbb {C}}\).
The case of negative cosmological constant is considerably more complicated as \(\varLambda \) appears with the wrong sign in the mass functional \(\mathcal {M}\). In a different context, Kunduri and Lucietti (2009, 2013) prove that the near horizon geometry of axisymmetric and stationary black holes is the one of the extremal Kerr–Newmannantide Sitter horizon and therefore they saturate (200) (see Hennig 2014 for an explicit expression and discussion).
4.5 An application: nonexistence of two black holes in equilibrium
A very interesting application of the area–angular momentum inequality (170), with \(Q=0\), is the result by Neugebauer and Hennig (2009, 2012, 2014) and Hennig and Neugebauer (2011) where they prove that a two black hole configuration in equilibrium does not exist. This problem has been open since the early days of General Relativity (see Neugebauer and Hennig 2014 for further references and Beig and Chruściel 1996; Beig and Schoen 2009; Manko et al. 2011 for different approaches and results on the subject); an alternative proof of nonexistence was given by Manko and Ruiz (2001). The Neugebauer and Hennig argument is the following: Start out with the spacetime metric for an axially symmetric, stationary system containing two disconnected Killing horizons on the symmetry axis. Use the Ernst formulation Ernst (1968) to obtain a system of equations equivalent to Einstein vacuum equations. Then the inverse scattering method is used to build a unique and exact solution to the Ernst equations, known as the double KerrNUT solution. A particular property of this solution is that both black holes can not satisfy the \(A\ge 8\pi J\) inequality simultaneously, which proves the nonexistence of two black holes in equilibrium. This result was generalized by Chruściel et al. (2011) to \(I^+\) regular black hole spacetimes.
5 Inequalities for objects
The inequalities presented in Sects. 3 and 4 are valid for black holes. The presence of such black hole is manifested through the hypothesis of the existence of a trapped surface or of a nontrivial topology in the initial data.
The interest in geometrical inequalities for ordinary objects is twofold. The most basic question is whether Einstein equations set restrictions on the values that physical parameters for objects can attain. This is not the case in Newtonian theory unless some specific matter model with intrinsic restrictions is used. Is this the case in General Relativity? Are there some conditions on the mass, size, rotation, and charge of an object, such that if they are not fulfilled, the object can not exist within the theory? This is related to the second question we want to address in this section. Are there geometrical inequalities for objects such that if violated, the object collapses to form a black hole? Clearly, the formation of a black hole after the collapse of an ordinary object is one possible scenario leading to the nonexistence raised in the first question. However, we emphasize that these two situations, i.e., an object exists and satisfies certain inequalities, and an object does not exist because it forms a black hole are in principle very different and require different treatments.
The problem of finding geometrical inequalities in the nonblack hole setting is wide open. At this point it is not all that clear what kind of inequalities one should look for (some of them have been motivated in the introduction though, see Sect. 1.3), nor what the proper systems and physical quantities are, that will produce such inequalities (we discuss this point below). This makes research in this area look a bit erratic, where new ideas are proposed or applied in almost every article. Because of this, we choose to present the results and discussions in a different manner as we do in previous sections. We discuss the general problems first and then present the results with specific remarks for them.
5.1 Discussion
There are two major differences between nonblack hole objects and black holes. The first is the problem of how to characterize the object in such a way that it produces the desired relation between physical quantities (like mass, electromagnetic charges or angular momentum) and size or shape parameters. That is, we would need some positivity condition to play the role of the stability of MOTS or minimal surfaces used for obtaining black hole inequalities (a nontrivial topology for the underlying initial surface is not considered when studying physically reasonable ordinary objects).
The second problem, maybe less challenging but still open, is how to properly measure the object. When nonblack hole objects are considered, one may want to consider measures of 3dimensional subsets of an initial data, and not just measures of 2surfaces. This raises several difficulties as there does not seem to exist consensus about what the best or more appropriate measure is for the size of a nonblack hole object. Indeed, a proper and suitable measure of size of an object should satisfy certain requirements. Namely, it should give a good, intuitive idea of size, it should be relatively easy to compute, it should be so chosen as to actually appear in the aimed geometrical inequalities.
Note that to explore the rigidity case in (201) means to address the problem of minimizing [Size] for given [Mass] (or given charge or angular momentum). Which in turn is closely related to the isoperimetric problem of minimizing area for given volume.
As a measure of size in the lefthand side of (201), one may attempt, inspired by the quasilocal inequalities for black holes (Sect. 4), to use the area of the surface enclosing the object. However, there are counter examples to an inequality of the form \(A\ge 4\pi Q^2\). The electrically counterpoised spheroids of dust, presented by Bonnor (1998), are regular, static, isolated systems that satisfy the energy conditions and whose enclosing surface can be made arbitrarily small relative to the charge enclosed, namely, \(A<kQ^2\) for any positive, arbitrary number k. Since these objects are highly prolate, it is expected that by assuming some kind of roundness on the enclosing surface, the area may give the desired estimate of charge. This example does not mean that the area should not be considered as a measure of size for the not round enough objects. But it says that in some cases, the area alone is not enough to control the amount of charge the object can carry.
Taking this observation into account, two paths can be taken to arrive at estimates of the form (201). One is to use special surfaces that are round enough. The other possibility is to use a measure that takes into account the deformation away from sphericity. In the first approach we encounter the following surfaces that capture the notion of round enough surface: isoperimetric surfaces (Sect. 5.2.3) coordinate spheres and convex surfaces (Sect. 5.2). Within the second approach, one may, as a first step, seek estimates using combinations of different well known measures, like area, distance to the boundary, etc. We come back to this point in Sect. 5.2.
There is another important issue referred to ordinary objects that is closely related to black holes, and it is the question of the collapse of an object to form a black hole. A few black hole formation criteria were constructed from geometrical inequalities stating that if certain inequality is not satisfied, then a black hole is formed. We review them below.
5.2 Results
As is the case for black holes, we divide the results according to whether angular momentum is considered explicitly into the inequality or not. This has to do mainly with the requirement of axial symmetry needed to define quasilocal angular momentum. As we see below, some of the results for ordinary objects employ similar techniques as the ones used in the treatment of black holes, i.e., harmonic maps theory, inverse mean curature flow, stability conditions, etc. Some results are quasilocal and some are global as they also incorporate the ADM mass. Various approaches have been taken to obtain the estimates. However, no variational problem has been formulated. This, in particular, implies that there is no (nontrivial) rigidity statement on the inequalities.
We mention here the setting where these inequalities are proven, and the main properties that lead to them.
Settings. All inequalities presented in this section are proven for objects in an initial data set \((\varSigma , h_{ij}, K_{ij}, \mu ,j^i)\). The objects themselves are taken to be open, bounded regions \(\varOmega \subset \varSigma \) with smooth boundaries \(\partial \varOmega \).

Strict positivity of the first eigenvalue of the linear differential operator \(\varDelta +\frac{1}{2}{}^3R\).

Stability of minimal surfaces and MOTSs.

Stability of the quotient space of maximal slices in axial symmetry.

Positivity and monotonicity properties of the Geroch energy.

Stability of isoperimetric surface.
5.2.1 Inequalities for objects without angular momentum
The key point in this result is the fact that the first Dirichlet eigenvalue of the operator \(\varDelta +\frac{1}{2}{}^3R\), if it is strictly positive, sets an upper bound to \({\mathcal {R}}_{SY}\).
Inequality (202) is purely quasilocal, the initial data does not need to be asymptotically flat.
Shoen and Yau also obtain the following black hole formation criterion: If \(\varSigma \) is asymptotically flat and matter fields satisfy the energy condition \(\mu j\ge \lambda >0\) on \(\varOmega \subset \varSigma \), then the opposite inequality to (202) implies that \(\varSigma \) contains an apparent horizon.
We wish to remark that inequalities (203) and (204) use stable minimal or trapped surfaces inside the object under study. In this way, they introduce the positivity condition which ultimately produces the inequality.
This result gives also the following black hole existence criterion: If the energy density of the object satisfies \(\rho >\pi /6M^2\) where M is the ADM mass, then the object lies inside a black hole and is not in static equilibrium.
5.2.2 Inequalities for objects with angular momentum
The key ingredient needed to include the angular momentum into a geometrical inequality is to relate it with the Einstein constraints. This is done via the relation with the current density \(j^i\) or with the extrinsic curvature \(K_{ij}\) of \(\varSigma \).
By assuming \(\varOmega \) to be mean convex, the same inequality is obtained when \({\mathcal {R}}_{OM}\) is used in the definition of \({\mathcal {R}}_D\).
This result gives the following black hole formation criteria: Given an axially symmetric initial data such that the initial surface \(\varSigma \) is asymptotically flat or has a strongly untrapped boundary. Under the energy condition stated above, if there exists a bounded region \(\varOmega \) where (212) does not hold, then \(\varSigma \) contains an apparent horizon.
We finally present an inequality that is global in the sense that in includes the ADM mass.
Inequality (215) also gives information about the shape of the object. It says that if the total mass is fixed, then the angular momentum determines how oblate or prolate the object can be. We also notice that the term \(m_T\) plays the role of a quasilocal mass (see Malec et al. 2002).
It is remarkable that the inequalities obtained in this section, although with different technical conditions, give rise to inequalities similar to the ones discussed in Sect. 1.3, which were informally derived from Newtonian considerations and the condition that nothing travels faster than light.
5.2.3 Isoperimetric surfaces
As we mention in Sect. 2.2, isoperimetric surfaces in initial data are an important system from which relations between physical and geometrical quantities can be obtained. We refer the reader to the articles by Eichmair and Metzger (2013a, b) for a detailed account on the results related to isoperimetric surfaces in Riemannian manifolds with application to General Relativity. See also Sect. 2.2 for references on discussions about the Penrose inequality for isoperimetric surfaces.
In this section we focus on inequalities relating size, angular momentum and charges, so, in this sense, they are quasilocal inequalities.
6 Open problems

Removing axial symmetry Axial symmetry is a requirement in all the geometrical inequalities involving angular momentum presented in this article. However, deviations from axial symmetry are of major importance, especially in astrophysics and numerical simulations. As was previously discussed, axial symmetry is an unremovable condition for the global inequalities (Huang et al. 2011). However, for quasilocal inequalities, the necessity of axial symmetry is not so clear. On one hand, there are heuristic and Newtonianlike considerations that suggest the possibility of nonaxially symmetric quasilocal geometrical inequalities. Also, the fact that the variational problem for the mass functional \(\mathbb {M}\) presented in Sect. 4.2.5 holds for not necessarily axially symmetric functions is somewhat encouraging. In particular, this last point shows the major role that extreme Kerr–Newman black holes plays as a limit solution, among a wider class of solutions. We need to emphasize that the problem of quasilocal geometrical inequalities involving angular momentum outside axial symmetry is not a well defined mathematical problem, because there is no proper quasilocal notion of angular momentum in this setting. Solving this problem, for one way or another is a major open problem.

Maximality and Janglike equations As we have seen in several parts of this article, most inequalities for initial data were originally proved assuming maximality. Whether this condition could be lifted or not was an unsolved question for some time. Later maximality was replaced by the existence of solutions to especific equations including equations of the Jang type. To prove that such solutions do exist is key to regard the maximality condition a removable one.

Mass–angular momentum inequality for data with inner boundary We have seen in Sect. 3.1 that the mass–charge inequality can be formulated in terms of initial data where the initial surface \(\varSigma \) is either complete with nontrivial topology, or has a weakly trapped inner boundary. On the other hand, the mass–angular momentum inequalities presented in Sect. 3.2 are proven only for complete initial surfaces with nontrivial topology. Extending this result to manifolds with boundaries is important for three reasons: first it would complete and unify the results about this type of global inequalities. Second, the proper formulation and resolution of the variational problem needed to derive the desired inequality when an inner boundary is present (analogous to the one used in the proof for the case of complete manifold), would clarify the role that extreme black holes play as borderline solutions. Finally, it seems that if one wants to make a connection between this geometrical inequality and the Penrose inequality, (see next item) a careful understanding of this case may be of use.

Connection with Penrose inequality The connection of the geometrical inequalities presented in this article and the positive mass theorem and the Penrose inequality seems to become deeper as further studies are performed. Not only they involve the same physical and geometrical quantities, i.e., mass, area (or size), angular momentum, charge, but also the techniques used in both problems seem to not be so different (see Sect. 5.2 for an inequality for objects using the inverse mean curvature flow). Exploring this connection may shed light into the problems and possible resolutions. In this respect, see the article by Anglada (2017), where he adapts the results in Anglada et al. (2017) for ordinary objects to study the Penrose inequality.

Minimum of mass functional for multiple rotating black holes The global inequality for multiple black holes, (147) is written in terms of the value of the mass functional on a minimizer solution. This minimizer is not known explicitly although there are a few numerical calculations indicating that for two regular Kerr black holes, the total mass square should be greater than the total angular momentum. Whether this relation holds for more general configurations or some other parameters are needed is unclear. Obtaining an explicit form of this value is of great importance because it would tell us, in particular, whether the total mass controls the individual angular momenta of the black holes, or the total angular momentum of the system, and exactly how it does it.

Global inequality with \(\varLambda \) Global inequalities relating mass, angular momentum and/or charge that also include explicitly a cosmological constant \(\varLambda \) have not been proven yet. A negative cosmological constant, however, has been admitted in the statements of the mass–angular momentum–charge inequalities (Cha and Khuri 2017).

Connection between \(\mathcal {M}\) and \(\mathbb {M}\) Another issue that must be better understood is the connection between the global and quasilocal inequalities for black holes that include the angular momentum. A partial implication was presented in Sect. 4.3 but there are many issues that are not entirely clear yet. This is not an easy problem since it involves relating global and quasilocal settings. Its full resolution might give a hint into the connection with the Penrose inequality.

Quasilocal estimate with negative \(\varLambda \) The way the cosmological constant appear into the mass functional \(\mathbb {M}\) makes the procedure used to prove the area–angular momentum–cosmological constant inequality hard to adapt when \(\varLambda \) is negative. Note that the problem with the negative \(\varLambda \) is not about how it enters into a generalization of the Area–Angular momentum inequality, but about whether such an inequality does exist. The works mentioned in Sect. 4.4 suggest that it does exist and a particular inequality motivated by extreme Kerr–Newman AdS black hole has been proposed. This problem is far from solved and new techniques must be implemented.

Ordinary objects As was discussed in Sect. 5, there are very basic questions that are unanswered with respect to geometrical inequalities for objects. Things like: what inequality we expect to obtain, how we should characterize the object and how we should measure them, are not clear. Concerning the first two issues, it is crucial to understand in what class of ordinary objects one expects to obtain a geometrical inequality. By this we mean that a positivity condition seems to be needed in order to derive the desired estimate. This leads naturally to the following question: Do all objects, say, in axial symmetry for simplicity, have a restriction on the allowed valued of their parameters? In particular, should they be round enough?

Measure of size for ordinary objects This issue was discussed in Sect. 5. There are various alternative notions of size but more work needs to be done. As seen in the results presented in Sect. 5.2.2, in the case of axial symmetry, it may be convenient to study measures constructed from the norm of the axial Killing vector field, \(\sqrt{\eta }\). This is supported by the following observations: The norm \(\sqrt{\eta }\) is bounded by the equatorial radius \({\mathcal {R}}_c\) (defined as the length, divided by \(2\pi \) of the greatest axially symmetric circle. This gives a clear and natural measure of size relevant for rotating objects. Also, for convex surfaces the variation of \(\eta \) along the inverse mean curvature flow is controlled by \(\eta \) itself. Also, the measure should take into account deviations from sphericity in all directions.

Connection with hoop conjecture Some versions of the hoop conjecture suggest to look for geometrical inequalities relating size, angular momentum and some measure of quasilocal mass of a certain region of spacetime (Senovilla 2008). There are several different quasilocal masses in the literature (Szabados 2004), but the problem of identifying the appropriate one(s) that simultaneously capture the matter content of the region, and that give rise to the desired meaningful geometrical inequalities is still open.
Footnotes
 1.
We thanks L. Szabados for clarifying this point.
Notes
Acknowledgements
We wish to thank László Szabados, Helmut Friedrich and José Luis Jaramillo for fruitful comments and discussions. Also, to the anonymous referees for their careful observations. This work was supported by Consejo Nacional de Investigaciones Cieníficas y Técnicas and Secretaría de Ciencia y Técnica  Universidad Nacional de Córdoba.
References
 Aceña AE, Dain S (2013) Stable isoperimetric surfaces in superextreme Reissner–Nordström. Class Quantum Grav 30:045013. https://doi.org/10.1088/02649381/30/4/045013. arXiv:1210.3509 ADSCrossRefzbMATHGoogle Scholar
 Aceña A, Dain S, Gabach Clément ME (2011) Horizon area–angular momentum inequality for a class of axially symmetric black holes. Class Quantum Grav 28:105014 arXiv:1012.2413 ADSMathSciNetCrossRefGoogle Scholar
 Alaee A, Kunduri HK (2014) Mass functional for initial data in 4 + 1dimensional spacetime. Phys Rev D 90:124078. https://doi.org/10.1103/PhysRevD.90.124078. arXiv:1411.0609 ADSCrossRefGoogle Scholar
 Alaee A, Kunduri HK (2016) Remarks on mass and angular momenta for \(U(1)^2\)invariant initial data. J Math Phys 57:032502. https://doi.org/10.1063/1.4944426. arXiv:1508.02337 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Alaee A, Khuri M, Kunduri H (2016) Proof of the mass–angular momentum inequality for biaxisymmetric black holes with spherical topology. Adv Theor Math Phys 20:1397–1441. https://doi.org/10.4310/ATMP.2016.v20.n6.a4. arXiv:1510.06974 MathSciNetCrossRefzbMATHGoogle Scholar
 Alaee A, Khuri M, Kunduri H (2017a) Mass–angularmomentum inequality for black ring spacetimes. Phys Rev Lett 119:071101. https://doi.org/10.1103/PhysRevLett.119.071101. arXiv:1705.08799 ADSMathSciNetCrossRefGoogle Scholar
 Alaee A, Khuri M, Kunduri H (2017b) Relating mass to angular momentum and charge in fivedimensional minimal supergravity. Ann Henri Poincaré 18:1703–1753. https://doi.org/10.1007/s0002301605421. arXiv:1608.06589 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Andersson L, Metzger J (2009) The area of horizons and the trapped region. Commun Math Phys 290:941–972. https://doi.org/10.1007/s002200080723y. arXiv:0708.4252 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Andersson L, Mars M, Simon W (2005) Local existence of dynamical and trapping horizons. Phys Rev Lett 95:111102. https://doi.org/10.1103/PhysRevLett.95.111102. arXiv:grqc/0506013 ADSCrossRefGoogle Scholar
 Andersson L, Mars M, Simon W (2008a) Stability of marginally outer trapped surfaces and existence of marginally outer trapped tubes. Adv Theor Math Phys 12:853–888. https://doi.org/10.4310/ATMP.2008.v12.n4.a5. arXiv:0704.2889 MathSciNetCrossRefzbMATHGoogle Scholar
 Andersson L, Mars M, Simon W (2008b) Stability of marginally outer trapped surfaces and existence of marginally outer trapped tubes. Adv Theor Math Phys 12:853–888. https://doi.org/10.4310/ATMP.2008.v12.n4.a5. http://projecteuclid.org/getRecord?id=euclid.atmp/1216046746
 Andersson L, Eichmair M, Metzger J (2011) Jang’s equation and its applications to marginally trapped surfaces. In: Agranovsky M et al (eds) Complex analysis and dynamical systems IV, part 2. General relativity, geometry, and PDE. Contemporary mathematics, vol 554. American Mathematical Society, Providence, pp 13–46 arXiv:1006.4601
 Anglada P (2017) Penroselike inequality with angular momentum for minimal surfaces. ArXiv eprints arXiv:1708.04646
 Anglada P, Dain S, Ortiz OE (2016) Inequality between size and charge in spherical symmetry. Phys Rev D 93:044055. https://doi.org/10.1103/PhysRevD.93.044055. arXiv:1511.04489 ADSMathSciNetCrossRefGoogle Scholar
 Anglada P, Gabach Clément ME, Ortiz OE (2017) Size, angular momentum and mass for objects. Class Quantum Grav 34:125011. https://doi.org/10.1088/13616382/aa6f3f. arXiv:1612.08658 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Ansorg M, Hennig J (2008) The inner Cauchy horizon of axisymmetric and stationary black holes with surrounding matter. Class Quantum Grav 25:222001. https://doi.org/10.1088/02649381/25/22/222001. arXiv:0810.3998 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Ansorg M, Hennig J (2009) The inner Cauchy horizon of axisymmetric and stationary black holes with surrounding matter in Einstein–Maxwell theory. Phys Rev Lett 102:221102. https://doi.org/10.1103/PhysRevLett.102.221102. arXiv:0903.5405 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Ansorg M, Pfister H (2008) A universal constraint between charge and rotation rate for degenerate black holes surrounded by matter. Class Quantum Grav 25:035009. https://doi.org/10.1088/02649381/25/3/035009. arXiv:0708.4196 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Ansorg M, Hennig J, Cederbaum C (2011) Universal properties of distorted Kerr–Newman black holes. Gen Relativ Gravit 43:1205–1210. https://doi.org/10.1007/s1071401011368. arXiv:1005.3128 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Arnowitt R, Deser S, Misner CW (1962) The dynamics of general relativity. In: Witten L (ed) Gravitation: an introduction to current research. Wiley, New York, pp 227–265 arXiv:grqc/0405109 Google Scholar
 Ashtekar A, Krishnan B (2002) Dynamical horizons: energy, angular momentum, fluxes and balance laws. Phys Rev Lett 89:261101. https://doi.org/10.1103/PhysRevLett.89.261101. arXiv:grqc/0207080 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Ashtekar A, Krishnan B (2003) Dynamical horizons and their properties. Phys Rev D 68:104030. https://doi.org/10.1103/PhysRevD.68.104030. arXiv:grqc/0308033 ADSMathSciNetCrossRefGoogle Scholar
 Ashtekar A, Krishnan B (2004) Isolated and dynamical horizons and their applications. Living Rev Relativ 7:10. https://doi.org/10.12942/lrr200410. arXiv:grqc/0407042 ADSCrossRefzbMATHGoogle Scholar
 Ashtekar A, Beetle C, Fairhurst S (2000a) Mechanics of isolated horizons. Class Quantum Grav 17:253 arXiv:grqc/9907068 ADSMathSciNetCrossRefGoogle Scholar
 Ashtekar A, Fairhurst S, Krishnan B (2000b) Isolated horizons: Hamiltonian evolution and the first law. Phys Rev D 62:104025 arXiv:grqc/0005083 ADSMathSciNetCrossRefGoogle Scholar
 Bardeen JM, Carter B, Hawking SW (1973) The four laws of black hole mechanics. Commun Math Phys 31:161–170. https://doi.org/10.1007/BF01645742 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Bartnik R (1986) The mass of an asymptotically flat manifold. Commun Pure App Math 39:661–693MathSciNetCrossRefGoogle Scholar
 Bartnik RA, Chruściel PT (2005) Boundary value problems for Diractype equations. J Reine Angew Math 579:13–73. https://doi.org/10.1515/crll.2005.2005.579.13. arXiv:math.DG/0307278 MathSciNetCrossRefzbMATHGoogle Scholar
 Beetle C, Wilder S (2015) A note on axial symmetries. Class Quantum Grav 32:047001. https://doi.org/10.1088/02649381/32/4/047001. arXiv:1401.0075 ADSCrossRefzbMATHGoogle Scholar
 Beig R, Chruściel PT (1996) Killing vectors in asymptotically flat spacetimes: I. Asymptotically translational Killing vectors and the rigid positive energy theorem. J Math Phys 37:1939–1961. https://doi.org/10.1063/1.531497. arXiv:grqc/9510015 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Beig R, O’Murchadha N (1996) Vacuum spacetimes with future trapped surfaces. Class Quantum Grav 13:739–752. https://doi.org/10.1088/02649381/13/4/014. arXiv:grqc/9511070 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Beig R, Schoen RM (2009) On static \(n\)body configurations in relativity. Class Quantum Grav 26:075014. https://doi.org/10.1088/02649381/26/7/075014. arXiv:0811.1727 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Bode T, Laguna P, Matzner R (2011) Superextremal spinning black holes via accretion. Phys Rev D 84:064044. https://doi.org/10.1103/PhysRevD.84.064044. arXiv:1106.1864 ADSCrossRefGoogle Scholar
 Bonnor WB (1980) Equilibrium of charged dust in general relativity. Gen Relativ Gravit 12:453–465. https://doi.org/10.1007/BF00756176 ADSMathSciNetCrossRefGoogle Scholar
 Bonnor WB (1998) A model of a spheroidal body. Class Quantum Grav 15:351. https://doi.org/10.1088/02649381/15/2/009 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Booth I (2005) Black hole boundaries. Can J Phys 83:1073–1099. https://doi.org/10.1139/p05063. arXiv:grqc/0508107 ADSCrossRefGoogle Scholar
 Booth I, Fairhurst S (2008) Extremality conditions for isolated and dynamical horizons. Phys Rev D 77:084005. https://doi.org/10.1103/PhysRevD.77.084005. arXiv:0708.2209 ADSMathSciNetCrossRefGoogle Scholar
 Bray HL, Jauregui JL (2015) Time flat surfaces and the monotonicity of the spacetime Hawking mass. Commun Math Phys 335:285–307. https://doi.org/10.1007/s0022001421622. arXiv:1310.8638 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Bray HL, Jauregui JL, Mars M (2016) Time flat surfaces and the monotonicity of the spacetime Hawking mass II. Ann Henri Poincaré 17:1457–1475. https://doi.org/10.1007/s0002301504202. arXiv:1402.3287 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Bryden ET, Khuri MA (2017) The area–angular momentum–charge inequality for black holes with positive cosmological constant. Class Quantum Grav 34:125017. https://doi.org/10.1088/13616382/aa70fd. arXiv:1611.10287 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 CabreraMunguia I (2015) Binary system of unequal counterrotating Kerr–Newman sources. Phys Rev D 91:044005. https://doi.org/10.1103/PhysRevD.91.044005. arXiv:1505.07080 ADSCrossRefGoogle Scholar
 CabreraMunguia I, Manko V, Ruiz E (2010) Remarks on the mass–angular momentum relations for two extreme Kerr sources in equilibrium. Phys Rev D 82:124042. https://doi.org/10.1103/PhysRevD.82.124042. arXiv:1010.0697 ADSCrossRefGoogle Scholar
 CabreraMunguia I, Lämmerzahl C, Macías A (2013) Exact solution for a binary system of unequal counterrotating black holes. Class Quantum Grav 30:175020. https://doi.org/10.1088/02649381/30/17/175020. arXiv:1302.4843 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Callen HB (1985) Thermodynamics and an introduction to thermostatistics, 2nd edn. Wiley, New YorkzbMATHGoogle Scholar
 Carroll SM, Johnson MC, Randall L (2009) Extremal limits and black hole entropy. JHEP 11:109. https://doi.org/10.1088/11266708/2009/11/109. arXiv:0901.0931 ADSMathSciNetCrossRefGoogle Scholar
 Carter B (1973) Black hole equilibrium states. In: Black holes/Les astres occlus (École d’Été Phys. Théor., Les Houches, 1972). Gordon and Breach, New York, pp 57–214Google Scholar
 Cha YS, Khuri MA (2014) Deformations of axially symmetric initial data and the mass–angular momentum inequality. Ann Henri Poincaré 1–56. https://doi.org/10.1007/s0002301403326. arXiv:1401.3384
 Cha YS, Khuri MA (2015) Deformations of charged axially symmetric initial data and the mass–angular momentum–charge inequality. Ann Henri Poincaré 16:2881–2918. https://doi.org/10.1007/s0002301403785. arXiv:1407.3621 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Cha YS, Khuri MA (2017) Transformations of asymptotically AdS hyperbolic initial data and associated geometric inequalities. ArXiv eprints arXiv:1707.09398
 Cha YS, Khuri M, Sakovich A (2016) Reduction arguments for geometric inequalities associated with asymptotically hyperboloidal slices. Class Quantum Grav 33:035009. https://doi.org/10.1088/02649381/33/3/035009. arXiv:1509.06255 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Chen PN, Wang MT, Yau ST (2016) Quasilocal angular momentum and center of mass in general relativity. Adv Theor Math Phys 20:671–682. https://doi.org/10.4310/ATMP.2016.v20.n4.a1. arXiv:1312.0990 MathSciNetCrossRefzbMATHGoogle Scholar
 Christodoulou D (1970) Reversible and irreversible transforations in blackhole physics. Phys Rev Lett 25:1596–1597. https://doi.org/10.1103/PhysRevLett.25.1596 ADSCrossRefGoogle Scholar
 Christodoulou D (1999) The instability of naked singularities in the gravitational collapse of a scalar field. Ann Math 149:183–217MathSciNetCrossRefGoogle Scholar
 Christodoulou D (2008) The formation of black holes in general relativity. In: On recent developments in theoretical and experimental general relativity, astrophysics and relativistic field theories. Proceedings, 12th Marcel Grossmann meeting on general relativity, Paris, France, July 12–18, 2009, vol 1–3, pp 24–34. https://doi.org/10.1142/9789814374552_0002. arXiv:0805.3880
 Chruściel P (1986) Boundary conditions at spatial infinity from a Hamiltonian point of view. In: Topological properties and global structure of spacetime (Erice, 1985). NATO advanced science institutes series B physics, vol 138. Plenum, New York, pp 49–59. http://www.phys.univtours.fr/~piotr/scans
 Chruściel PT (2008) Mass and angular–momentum inequalities for axisymmetric initial data sets I. Positivity of mass. Ann Phys 323:2566–2590. https://doi.org/10.1016/j.aop.2007.12.010. arXiv:0710.3680 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Chruściel PT, Lopes Costa J (2009) Mass, angular–momentum, and charge inequalities for axisymmetric initial data. Class Quantum Grav 26:235013. https://doi.org/10.1088/02649381/26/23/235013. arXiv:0909.5625 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Chruściel PT, Mazzeo R (2015) Initial data sets with ends of cylindrical type: I. The Lichnerowicz equation. Ann Henri Poincaré 16:1231–1266. https://doi.org/10.1007/s000230140339z. arXiv:1201.4937 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Chruściel PT, Nguyen L (2011) A lower bound for the mass of axisymmetric connected black hole data sets. Class Quantum Grav 28:125001. https://doi.org/10.1088/02649381/28/12/125001. arXiv:1102.1175 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Chruściel PT, Delay E, Galloway GJ, Howard R (2001) The area theorem. Ann Henri Poincaré 2:109–178. https://doi.org/10.1007/PL00001029. arXiv:grqc/0001003 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Chruściel PT, Maerten D, Tod P (2006a) Rigid upper bounds for the angular momentum and centre of mass of nonsingular asymptotically antide Sitter spacetimes. JHEP 11:084. https://doi.org/10.1088/11266708/2006/11/084. arXiv:grqc/0606064 ADSMathSciNetCrossRefGoogle Scholar
 Chruściel PT, Reall HS, Tod P (2006b) On Israel–Wilson–Perjes black holes. Class Quantum Grav 23:2519–2540. https://doi.org/10.1088/02649381/23/7/018. arXiv:grqc/0512116 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Chruściel PT, Li Y, Weinstein G (2008) Mass and angularmomentum inequalities for axisymmetric initial data sets. II. Angularmomentum. Ann Phys 323:2591–2613. https://doi.org/10.1016/j.aop.2007.12.011. arXiv:0712.4064 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Chruściel PT, Eckstein M, Nguyen L, Szybka SJ (2011) Existence of singularities in twoKerr black holes. Class Quantum Grav 28:245017. https://doi.org/10.1088/02649381/28/24/245017. arXiv:1111.1448 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Chruściel PT, Costa JL, Heusler M (2012) Stationary black holes: uniqueness and beyond. Living Rev Relativ 15:7. https://doi.org/10.12942/lrr20127. arXiv:1205.6112 ADSCrossRefzbMATHGoogle Scholar
 Chruściel PT, Mazzeo R, Pocchiola S (2013) Initial data sets with ends of cylindrical type: II. The vector constraint equation. Adv Theor Math Phys 17:829–865. https://doi.org/10.4310/ATMP.2013.v17.n4.a4. arXiv:1203.5138 MathSciNetCrossRefzbMATHGoogle Scholar
 Chruściel PT, Szybka SJ, Tod P (2017) Towards a classification of nearhorizons geometries. ArXiv eprints arXiv:1707.01118
 Compere G (2017) The Kerr/CFT correspondence and its extensions. Living Rev Relativ 20:1. https://doi.org/10.1007/s4111401700032. arXiv:1203.3561 ADSCrossRefzbMATHGoogle Scholar
 Corvino J, Gerek A, Greenberg M, Krummel B (2007) On isoperimetric surfaces in general relativity. Pacific J Math 231:63–84. https://doi.org/10.2140/pjm.2007.231.63 MathSciNetCrossRefzbMATHGoogle Scholar
 Costa JL (2010) Proof of a Dain inequality with charge. J Phys A 43:285202 arXiv:0912.0838 MathSciNetCrossRefGoogle Scholar
 Cvetic M, Gibbons G, Pope C (2011a) Universal area product formulae for rotating and charged black holes in four and higher dimensions. Phys Rev Lett 106:121301. https://doi.org/10.1103/PhysRevLett.106.121301. arXiv:1011.0008 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Cvetic M, Gibbons GW, Pope CN (2011b) More about Birkhoff’s invariant and Thorne’s hoop conjecture for horizons. Class Quantum Grav 28:195001. https://doi.org/10.1088/02649381/28/19/195001. arXiv:1104.4504 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Dafermos M (2005) Spherically symmetric spacetimes with a trapped surface. Class Quantum Grav 22:2221–2232. https://doi.org/10.1088/02649381/22/11/019. arXiv:grqc/0403032 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Dain S (2004) Trapped surfaces as boundaries for the constraint equations. Class Quantum Grav 21:555–573 arXiv:grqc/0308009 ADSMathSciNetCrossRefGoogle Scholar
 Dain S (2006a) Angular momemtum–mass inequality for axisymmetric black holes. Phys Rev Lett 96:101101 arXiv:grqc/0511101 ADSMathSciNetCrossRefGoogle Scholar
 Dain S (2006b) Proof of the (local) angular momemtum–mass inequality for axisymmetric black holes. Class Quantum Grav 23:6845–6855 arXiv:grqc/0511087 ADSMathSciNetCrossRefGoogle Scholar
 Dain S (2006c) A variational principle for stationary, axisymmetric solutions of Einstein’s equations. Class Quantum Grav 23:6857–6871 arXiv:grqc/0508061 ADSMathSciNetCrossRefGoogle Scholar
 Dain S (2008) Proof of the angular momentum–mass inequality for axisymmetric black holes. J Differ Geom 79:33–67 arXiv:grqc/0606105 MathSciNetCrossRefGoogle Scholar
 Dain S (2010) Extreme throat initial data set and horizon area–angular momentum inequality for axisymmetric black holes. Phys Rev D 82:104010. https://doi.org/10.1103/PhysRevD.82.104010. arXiv:1008.0019 ADSCrossRefGoogle Scholar
 Dain S (2011) Geometric inequalities for axially symmetric black holes. ArXiv eprints arXiv:1111.3615
 Dain S (2012) Geometric inequalities for axially symmetric black holes. Class Quantum Grav 29:073001. https://doi.org/10.1088/02649381/29/7/073001. arXiv:1111.3615 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Dain S (2014a) Geometric inequalities for black holes. Gen Relativ Gravit 46:1715. https://doi.org/10.1007/s1071401417151. arXiv:1401.8166 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Dain S (2014b) Inequality between size and angular momentum for bodies. Phys Rev Lett 112:041101. https://doi.org/10.1103/PhysRevLett.112.041101. arXiv:1305.6645 ADSCrossRefGoogle Scholar
 Dain S, Ortiz OE (2009) Numerical evidences for the angular momentum–mass inequality for multiple axially symmetric black holes. Phys Rev D 80:024045. https://doi.org/10.1103/PhysRevD.80.024045. arXiv:0905.0708 ADSCrossRefGoogle Scholar
 Dain S, Reiris M (2011) Area–angularmomentum inequality for axisymmetric black holes. Phys Rev Lett 107:051101. https://doi.org/10.1103/PhysRevLett.107.051101. arXiv:1102.5215 ADSCrossRefGoogle Scholar
 Dain S, Jaramillo JL, Reiris M (2012) Area–charge inequality for black holes. Class Quantum Grav 29:035013 arXiv:1109.5602 ADSMathSciNetCrossRefGoogle Scholar
 Eichmair M (2007) The plateau problem for apparent horizons. J Differ Geom. https://doi.org/10.4310/jdg/1264601035. arXiv:0711.4139
 Eichmair M, Metzger J (2013a) Large isoperimetric surfaces in initial data sets. J Differ Geom 94:159–186. https://doi.org/10.4310/jdg/1361889064. arXiv:1102.2999 MathSciNetCrossRefzbMATHGoogle Scholar
 Eichmair M, Metzger J (2013b) Unique isoperimetric foliations of asymptotically flat manifolds in all dimensions. Invent Math 194:591. https://doi.org/10.1007/s0022201304525. arXiv:1204.6065 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Eichmair M, Galloway GJ, Pollack D (2013) Topological censorship from the initial data point of view. J Differ Geom 95:389–405. https://doi.org/10.4310/jdg/1381931733. arXiv:1204.0278 MathSciNetCrossRefzbMATHGoogle Scholar
 Epp RJ, McGrath PL, Mann RB (2013) Momentum in general relativity: local versus quasilocal conservation laws. Class Quantum Grav 30:195019. https://doi.org/10.1088/02649381/30/19/195019. arXiv:1306.5500 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Ernst FJ (1968) New formulation of the axially symmetric gravitational field problem. Phys Rev 167:1175–1179. https://doi.org/10.1103/PhysRev.168.1415 ADSCrossRefGoogle Scholar
 Fajman D, Simon W (2014) Area inequalities for stable marginally outer trapped surfaces in Einstein–Maxwelldilaton theory. Adv Theor Math Phys 18:687–707. https://doi.org/10.4310/ATMP.2014.v18.n3.a4. arXiv:1308.3659 MathSciNetCrossRefzbMATHGoogle Scholar
 Gabach Clément ME (2011) Comment on “Horizon area–angular momentum inequality for a class of axially symmetric black holes”. ArXiv eprints arXiv:1102.3834
 Gabach Clément ME (2012) Bounds on the force between black holes. Class Quantum Grav 29:165008. https://doi.org/10.1088/02649381/29/16/165008. arXiv:1201.4099 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Gabach Clément ME, Jaramillo JL (2012) Black hole area–angular momentum–charge inequality in dynamical nonvacuum spacetimes. Phys Rev D 86:064021. https://doi.org/10.1103/PhysRevD.86.064021. arXiv:1111.6248 ADSCrossRefGoogle Scholar
 Gabach Clément ME, Reiris M (2013) On the shape of rotating blackholes. Phys Rev D 88:044031. https://doi.org/10.1103/PhysRevD.88.044031. arXiv:1306.1019 ADSCrossRefGoogle Scholar
 Gabach Clément ME, Jaramillo JL, Reiris M (2013) Proof of the area–angular momentum–charge inequality for axisymmetric black holes. Class Quantum Grav 30:065017. https://doi.org/10.1088/02649381/30/6/065017. arXiv:1207.6761 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Gabach Clément ME, Reiris M, Simon W (2015) The area–angular momentum inequality for black holes in cosmological spacetimes. Class Quantum Grav 32:145006. https://doi.org/10.1088/02649381/32/14/145006. arXiv:1501.07243 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Galloway GJ, O’Murchadha N (2008) Some remarks on the size of bodies and black holes. Class Quantum Grav 25:105009. https://doi.org/10.1088/02649381/25/10/105009. arXiv:0802.3247 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Gannon D (1975) Singularities in nonsimply connected spacetimes. J Math Phys 16:2364–2367. https://doi.org/10.1063/1.522498 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Gannon D (1976) On the topology of spacelike hypersurfaces, singularities, and black holes. Gen Relativ Gravit 7:219–232. https://doi.org/10.1007/BF00763437 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Gibbons GW (1984) The isoperimetric and Bogomolny inequalities for black holes. In: Willmore TJ, Hitchin N (eds) Global Riemannian geometry. Wiley, New York, pp 194–202Google Scholar
 Gibbons GW (1997) Collapsing shells and the isoperimetric inequality for black holes. Class Quantum Grav 14:2905–2915. https://doi.org/10.1088/02649381/14/10/016 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Gibbons GW (1999) Some comments on gravitational entropy and the inverse mean curvature flow. Class Quantum Grav 16:1677–1687. https://doi.org/10.1088/02649381/16/6/302. arXiv:hepth/9809167 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Gibbons GW (2009) Birkhoff’s invariant and Thorne’s hoop conjecture. ArXiv eprints arXiv:0903.1580
 Gibbons GW, Holzegel G (2006) The positive mass and isoperimetric inequalities for axisymmetric black holes in four and five dimensions. Class Quantum Grav 23:6459–6478. https://doi.org/10.1088/02649381/23/22/022. arXiv:grqc/0606116 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Gibbons GW, Hull CM (1982) A Bogomolny bound for general relativity and solitons in \(N=2\) supergravity. Phys Lett B 109:190–194. https://doi.org/10.1016/03702693(82)907511 ADSMathSciNetCrossRefGoogle Scholar
 Gibbons GW, Hawking SW, Horowitz GT, Perry MJ (1983) Positive mass theorems for black holes. Commun Math Phys 88:295–308. https://doi.org/10.1007/BF01213209 ADSMathSciNetCrossRefGoogle Scholar
 Gibbons GW, Kallosh R, Kol B (1996) Moduli, scalar charges, and the first law of black hole thermodynamics. Phys Rev Lett 77:4992–4995. https://doi.org/10.1103/PhysRevLett.77.4992. arXiv:hepth/9607108 ADSCrossRefGoogle Scholar
 Hájiček P (1974) Three remarks on axisymmetric stationary horizons. Commun Math Phys 36:305–320. https://doi.org/10.1007/BF01646202 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Hartle JB, Hawking SW (1972) Solutions of the Einstein–Maxwell equations with many black holes. Commun Math Phys 26:87–101. https://doi.org/10.1007/BF01645696 ADSMathSciNetCrossRefGoogle Scholar
 Hawking SW (1971) Gravitational radiation from collidings black holes. Phys Rev Lett 26:1344–1346. https://doi.org/10.1103/PhysRevLett.26.1344 ADSCrossRefGoogle Scholar
 Hayward SA (2011) Involute, minimal, outer and increasingly trapped surfaces. Int J Mod Phys D 20:401–411. https://doi.org/10.1142/S0218271811018718. arXiv:0906.2528 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Hayward SA, Shiromizu T, Nakao K (1994) A cosmological constant limits the size of black holes. Phys Rev D 49:5080–5085. https://doi.org/10.1103/PhysRevD.49.5080. arXiv:grqc/9309004 ADSMathSciNetCrossRefGoogle Scholar
 Hennig J (2014) Geometric relations for rotating and charged AdS black holes. Class Quantum Grav 31:135005. https://doi.org/10.1088/02649381/31/13/135005. arXiv:1402.5198 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Hennig J, Ansorg M (2009) The inner Cauchy horizon of axisymmetric and stationary black holes with surrounding matter in Einstein–Maxwell theory: study in terms of soliton methods. Ann Henri Poincaré 10:1075–1095. https://doi.org/10.1007/s0002300900120. arXiv:0904.2071 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Hennig J, Neugebauer G (2011) Nonexistence of stationary twoblackhole configurations: the degenerate case. Gen Relativ Gravit 43:3139–3162. https://doi.org/10.1007/s1071401112280. arXiv:1103.5248 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Hennig J, Ansorg M, Cederbaum C (2008) A universal inequality between the angular momentum and horizon area for axisymmetric and stationary black holes with surrounding matter. Class Quantum Grav 25:162002. https://doi.org/10.1088/02649381/25/16/162002 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Hennig J, Cederbaum C, Ansorg M (2010) A universal inequality for axisymmetric and stationary black holes with surrounding matter in the Einstein–Maxwell theory. Commun Math Phys 293:449–467. https://doi.org/10.1007/s002200090889y. arXiv:0812.2811 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Herzlich M (1998) The positive mass theorem for black holes revisited. J Geom Phys 26:97–111. https://doi.org/10.1016/S03930440(97)000405 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Hildebrandt S, Kaul H, Widman KO (1977) An existence theorem for harmonic mappings of Riemannian manifolds. Acta Math 138:1–16. https://doi.org/10.1007/BF02392311 MathSciNetCrossRefzbMATHGoogle Scholar
 Hod S (2015) Bekenstein’s generalized second law of thermodynamics: the role of the hoop conjecture. Phys Lett B 751:241–245. https://doi.org/10.1016/j.physletb.2015.10.052. arXiv:1511.03665 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Hollands S (2012) Horizon area–angular momentum inequality in higher dimensional spacetimes. Class Quantum Grav 29:065006. https://doi.org/10.1088/02649381/29/6/065006. arXiv:1110.5814 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Horowitz GT (1984) The positive energy theorem and its extensions. In: Flaherty FJ (ed) Asymptotic behavior of mass and spacetime geometry. Lecture notes in physics, vol 202. Springer, Berlin, pp 1–21. https://doi.org/10.1007/BFb0048063
 Huang LH, Schoen R, Wang MT (2011) Specifying angular momentum and center of mass for vacuum initial data sets. Commun Math Phys 306:785–803. https://doi.org/10.1007/s0022001112959. arXiv:1008.4996 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Huisken G (1998) Evolution of hypersurfaces by their curvature in Riemannian manifolds. In: Rehmann U (ed) Proceedings of the international congress of mathematicians (ICM), vol II, Bielefeld. Documenta mathematica, extra volume, pp 349–360. http://eudml.org/doc/227789
 Huisken G, Ilmanen T (2001) The inverse mean curvature flow and the Riemannian Penrose inequality. J Differ Geom 59:352–437. https://doi.org/10.4310/jdg/1090349447 MathSciNetCrossRefzbMATHGoogle Scholar
 Immerman JD, Baumgarte TW (2009) Trumpetpuncture initial data for black holes. Phys Rev D 80:061501. https://doi.org/10.1103/PhysRevD.80.061501. arXiv:0908.0337 ADSMathSciNetCrossRefGoogle Scholar
 Jackson JD (1999) Classical electrodynamics, 3rd edn. Wiley, HobokenzbMATHGoogle Scholar
 Jaramillo JL (2013) Area inequalities for stable marginally trapped surfaces. In: Sánchez M, Ortega M, Romero A (eds) Recent trends in Lorentzian geometry. Springer proceedings in mathematics and statistics, vol 26. Springer, New York, pp 139–161. https://doi.org/10.1007/9781461448976_5. arXiv:1201.2054
 Jaramillo JL, Gourgoulhon E, CorderoCarrion I, Ibanez JM (2008) Trapping horizons as inner boundary conditions for black hole spacetimes. Phys Rev D 77:047501. https://doi.org/10.1103/PhysRevD.77.047501. arXiv:0709.2842 ADSMathSciNetCrossRefGoogle Scholar
 Jaramillo JL, Reiris M, Dain S (2011) Black hole area–angular momentum inequality in nonvacuum spacetimes. Phys Rev D 84:121503. https://doi.org/10.1103/PhysRevD.84.121503. arXiv:1106.3743 ADSCrossRefGoogle Scholar
 Khuri M, Weinstein G (2016) The positive mass theorem for multiple rotating charged black holes. Calc Var 55:33. https://doi.org/10.1007/s0052601609698. arXiv:1502.06290
 Khuri M, Weinstein G, Yamada S (2017) Proof of the Riemannian Penrose inequality with charge for multiple black holes. J Differ Geom 106:451–498. https://doi.org/10.4310/jdg/1500084023. arXiv:1409.3271 MathSciNetCrossRefzbMATHGoogle Scholar
 Khuri MA (2009) The hoop conjecture in spherically symmetric spacetimes. Phys Rev D 80:124025. https://doi.org/10.1103/PhysRevD.80.124025. arXiv:0912.3533 ADSMathSciNetCrossRefGoogle Scholar
 Khuri MA (2015a) Existence of black holes due to concentration of angular momentum. JHEP 2015(06):188. https://doi.org/10.1007/JHEP06(2015)188. arXiv:1503.06166
 Khuri MA (2015b) Inequalities between size and charge for bodies and the existence of black holes due to concentration of charge. J Math Phys 56:112503. https://doi.org/10.1063/1.4936149. arXiv:1505.04516 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Khuri MA, Weinstein G (2013) Rigidity in the positive mass theorem with charge. J Math Phys 54:092501. https://doi.org/10.1063/1.4820469. arXiv:1307.5499 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Komar A (1959) Covariant conservation laws in general relativity. Phys Rev 113:934–936. https://doi.org/10.1103/PhysRev.113.934 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Kunduri HK, Lucietti J (2009) A classification of nearhorizon geometries of extremal vacuum black holes. J Math Phys 50:082502. https://doi.org/10.1063/1.3190480. arXiv:0806.2051 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Kunduri HK, Lucietti J (2013) Classification of nearhorizon geometries of extremal black holes. Living Rev Relativ 16:8. https://doi.org/10.12942/lrr20138. arXiv:1306.2517 ADSCrossRefzbMATHGoogle Scholar
 Lee C (1976) A restriction on the topology of Cauchy surfaces in general relativity. Commun Math Phys 51:157. https://doi.org/10.1007/BF01609346 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Lewandowski J, Pawlowski T (2003) Extremal isolated horizons: a local uniqueness theorem. Class Quantum Grav 20:587–606. https://doi.org/10.1088/02649381/20/4/303. arXiv:grqc/0208032 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Maeda K, Koike T, Narita M, Ishibashi A (1998) Upper bound for entropy in asymptotically de Sitter spacetime. Phys Rev D 57:3503–3508. https://doi.org/10.1103/PhysRevD.57.3503. arXiv:grqc/9712029 ADSMathSciNetCrossRefGoogle Scholar
 Maerten D (2006) Positive energymomentum theorem for AdSasymptotically hyperbolic manifolds. Ann Henri Poincaré 7:975–1011. https://doi.org/10.1007/s0002300602739 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Majumdar SD (1947) A class of exact solutions of Einstein’s field equations. Phys Rev 72:390. https://doi.org/10.1103/PhysRev.72.390 ADSMathSciNetCrossRefGoogle Scholar
 Malec E (1992) Isoperimetric inequalities in the physics of black holes. Acta Phys Pol B 22:829ADSMathSciNetGoogle Scholar
 Malec E, Xie N (2015) BrownYork mass and the hoop conjecture in nonspherical massive systems. Phys Rev D 91:081501. https://doi.org/10.1103/PhysRevD.91.081501. arXiv:1503.01354 ADSCrossRefGoogle Scholar
 Malec E, Mars M, Simon W (2002) On the Penrose inequality for general horizons. Phys Rev Lett 88:121102. https://doi.org/10.1103/PhysRevLett.88.121102. arXiv:grqc/0201024 ADSMathSciNetCrossRefGoogle Scholar
 Manko VS, Ruiz E (2001) Exact solution of the doubleKerr equilibrium problem. Class Quantum Grav 18:L11–L15. https://doi.org/10.1088/02649381/18/2/102 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Manko VS, Ruiz E, Sadovnikova MB (2011) Stationary configurations of two extreme black holes obtainable from the Kinnersley–Chitre solution. Phys Rev D 84:064005. https://doi.org/10.1103/PhysRevD.84.064005. arXiv:1105.2646 ADSCrossRefGoogle Scholar
 Manko VS, Rabadan RI, Ruiz E (2013) The Breton–Manko equatorially antisymmetric binary configuration revisited. Class Quantum Grav 30:145005. https://doi.org/10.1088/02649381/30/14/145005. arXiv:1302.7110 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Manko VS, Rabadán RI, SanabriaGómez JD (2014) Stationary black diholes. Phys Rev D 89(6):064049. https://doi.org/10.1103/PhysRevD.89.064049. arXiv:1311.2326 ADSCrossRefGoogle Scholar
 Mars M (2009) Present status of the Penrose inequality. Class Quantum Grav 26:193001. https://doi.org/10.1088/02649381/26/19/193001. arXiv:0906.5566 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Mars M (2014) Stability of marginally outer trapped surfaces and geometric inequalities. In: Bičák J, Ledvinka T (eds) General relativity, cosmology and astrophysics: perspectives 100 years after Einstein’s stay in Prague. Fundamental theories of physics, vol 177. Springer, Cham, pp 191–208. https://doi.org/10.1007/9783319063492_8
 Mars M, Senovilla JMM (1993) Axial symmetry and conformal Killing vectors. Class Quantum Grav 10:1633–1647. https://doi.org/10.1088/02649381/10/8/020. arXiv:grqc/0201045 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Meeks W III, Simon L, Yau ST (1982) Embedded minimal surfaces, exotic spheres, and manifolds with positive Ricci curvature. Ann Math 2(116):621–659. https://doi.org/10.2307/2007026 MathSciNetCrossRefzbMATHGoogle Scholar
 Moreschi OM, Sparling GAJ (1984) On the positive energy theorem involving mass and electromagnetic charges. Commun Math Phys 95:113–120. https://doi.org/10.1007/BF01215757 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Murchadha NO (1986) How large can a star be? Phys Rev Lett 57:2466–2469. https://doi.org/10.1103/PhysRevLett.57.2466 ADSCrossRefGoogle Scholar
 Murchadha NO, Tung RS, Xie N, Malec E (2010) The BrownYork mass and the Thorne hoop conjecture. Phys Rev Lett 104:041101. https://doi.org/10.1103/PhysRevLett.104.041101. arXiv:0912.4001 ADSMathSciNetCrossRefGoogle Scholar
 Nester JM, Meng FF, Chen CM (2004) Quasilocal centerofmass. J Korean Phys Soc 45:S22–S25. https://doi.org/10.3938/jkps.45.22. arXiv:grqc/0403103 CrossRefGoogle Scholar
 Neugebauer G, Hennig J (2009) Nonexistence of stationary twoblackhole configurations. Gen Relativ Gravit 41:2113–2130. https://doi.org/10.1007/s1071400908408. arXiv:0905.4179 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Neugebauer G, Hennig J (2012) Stationary twoblackhole configurations: a nonexistence proof. J Geom Phys 62:613–630. https://doi.org/10.1016/j.geomphys.2011.05.008. arXiv:1105.5830 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Neugebauer G, Hennig J (2014) Stationary blackhole binaries: a nonexistence proof. In: Bičák J, Ledvinka T (eds) General relativity, cosmology and astrophysics: perspectives 100 years after Einstein’s stay in Prague. Fundamental theories of physics, vol 177. Springer, Cham, pp 209–228. https://doi.org/10.1007/9783319063492_9. arXiv:1302.0573
 Nordström G (1918) On the energy of the gravitational field in Einstein’s theory. Verhandl Koninkl Ned Akad Wetenschap, Afdel Natuurk, Amsterdam 26:1201–1208zbMATHGoogle Scholar
 Papapetrou A (1945) A static solution of the equations of the gravitational field for an arbitrary chargedistribution. Proc R Irish Acad A 51:191–204MathSciNetGoogle Scholar
 Penrose R (1965) Gravitational collapse and spacetime singularities. Phys Rev Lett 14:57–59. https://doi.org/10.1103/PhysRevLett.14.57 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Penrose R (1973) Naked singularities. Ann NY Acad Sci 224:125–134. https://doi.org/10.1111/j.17496632.1973.tb41447.x ADSCrossRefzbMATHGoogle Scholar
 Penrose R, Floyd RM (1971) Extraction of rotational energy from a black hole. Nature 229:177–179. https://doi.org/10.1038/physci229177a0 ADSCrossRefGoogle Scholar
 Reiris M (2014a) On extreme Kerrthroats and zero temperature black holes. Class Quantum Grav 31:025001. https://doi.org/10.1088/02649381/31/2/025001. arXiv:1209.4530 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Reiris M (2014b) On the shape of bodies in general relativistic regimes. Gen Relativ Gravit 46:1777. https://doi.org/10.1007/s1071401417770. arXiv:1406.6938 ADSCrossRefzbMATHGoogle Scholar
 Reissner H (1916) Über die Eigengravitation des elektrischen Feldes nach der Einsteinschen Theorie. Ann Phys 50:106–120. https://doi.org/10.1002/andp.19163550905 CrossRefGoogle Scholar
 Rogatko M (2014) Mass angular momentum and charge inequalities for black holes in Einstein–Maxwellaxion–dilaton gravity. Phys Rev D 89:044020. https://doi.org/10.1103/PhysRevD.89.044020. arXiv:1402.3376 ADSCrossRefGoogle Scholar
 Rogatko M (2017) Angular momentum, mass and charge inequalities for black holes in Einstein–Maxwell gravity with dark matter sector. ArXiv eprints arXiv:1701.07643
 Schoen R, Yau ST (1979) On the proof of the positive mass conjecture in general relativity. Commun Math Phys 65:45–76. https://doi.org/10.1007/BF01940959 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Schoen R, Yau ST (1981) Proof of the positive mass theorem. II. Commun Math Phys 79:231–260. https://doi.org/10.1007/BF01942062 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Schoen R, Yau ST (1983) The existence of a black hole due to condensation of matter. Commun Math Phys 90:575–579. https://doi.org/10.1007/BF01216187. https://projecteuclid.org/euclid.cmp/1103940419
 Schoen R, Yau S (2017) Positive scalar curvature and minimal hypersurface singularities. ArXiv eprints arXiv:1704.05490
 Schoen R, Zhou X (2013) Convexity of reduced energy and mass angular momentum inequalities. Ann Henri Poincaré 14:1747–1773. https://doi.org/10.1007/s0002301302401 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Senovilla JMM (2008) A reformulation of the hoop conjecture. Europhys Lett 81:20004. https://doi.org/10.1209/02955075/81/20004. arXiv:0709.0695 ADSMathSciNetCrossRefGoogle Scholar
 Senovilla JMM (2011) Trapped surfaces. Int J Mod Phys D 20:2139. https://doi.org/10.1142/S0218271811020354. arXiv:1107.1344 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Senovilla JMM, Garfinkle D (2015) The 1965 Penrose singularity theorem. Class Quantum Grav 32:124008. https://doi.org/10.1088/02649381/32/12/124008. arXiv:1410.5226 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Shiromizu T, Nakao K, Kodama H, Maeda KI (1993) Can large black holes collide in de Sitter spacetime? An inflationary scenario of an inhomogeneous universe. Phys Rev D 47:R3099–R3102. https://doi.org/10.1103/PhysRevD.47.R3099 ADSCrossRefGoogle Scholar
 Simon W (2012) Bounds on area and charge for marginally trapped surfaces with a cosmological constant. Class Quantum Grav 29:062001. https://doi.org/10.1088/02649381/29/6/062001. arXiv:1109.6140 ADSCrossRefzbMATHGoogle Scholar
 Sorce J, Wald RM (2017) Gedanken experiments to destroy a black hole. II. Kerr–Newman black holes cannot be overcharged or overspun. Phys Rev D 96(10):104014. https://doi.org/10.1103/PhysRevD.96.104014. arXiv:1707.05862 ADSCrossRefGoogle Scholar
 Szabados LB (2004) Quasilocal energymomentum and angular momentum in GR: a review article. Living Rev Relativ 7:4. https://doi.org/10.12942/lrr20044 ADSCrossRefzbMATHGoogle Scholar
 Thorne KS (1972) Nonspherical gravitational collapse: a short review. In: Klauder J (ed) Magic without magic: John Archibald Wheeler. A collection of essays in honor of his sixtieth birthday. W.H. Freeman, San Francisco, pp 231–258Google Scholar
 Tod KP (1983) All metrics admitting supercovariantly constant spinors. Phys Lett B 121:241–244. https://doi.org/10.1016/03702693(83)907979 ADSMathSciNetCrossRefGoogle Scholar
 Tung RS (2009) Energy and angular momentum in strong gravitating systems. Int J Mod Phys A A24:3538–3544. https://doi.org/10.1142/S0217751X09047168. arXiv:0903.1036 ADSCrossRefzbMATHGoogle Scholar
 Visser M (2013) Area products for black hole horizons. Phys Rev D 88:044014. https://doi.org/10.1103/PhysRevD.88.044014. arXiv:1205.6814 ADSCrossRefGoogle Scholar
 Wald R (1999) Gravitational collapse and cosmic censorship. In: Iyer BR, Bhawal B (eds) Black holes, gravitational radiation and the universe. Fundamental theories of physics, vol 100. Kluwer Academic, Dordrecht, pp 69–85. https://doi.org/10.1007/9789401709347_5. arXiv:grqc/9710068
 Wald RM (1984) General relativity. The University of Chicago Press, ChicagoCrossRefGoogle Scholar
 Wald RM (2001) The thermodynamics of black holes. Living Rev Relativ 4:6. https://doi.org/10.12942/lrr20016. arXiv:grqc/9912119 ADSCrossRefzbMATHGoogle Scholar
 Wali KC (1982) Chandrasekhar vs. Eddington—an unanticipated confrontation. Phys Today 35:33. https://doi.org/10.1063/1.2914790 CrossRefGoogle Scholar
 Weinstein G (1996) \(N\) black hole stationary and axially symmetric solutions of the Einstein–Maxwell equations. Commun Part Diff Eq 21:1389–1430. https://doi.org/10.1080/03605309608821232. arXiv:grqc/9412036 MathSciNetCrossRefzbMATHGoogle Scholar
 Wieland W (2017) Quasilocal gravitational angular momentum and centre of mass from generalised Witten equations. Gen Relativ Gravit 49:38. https://doi.org/10.1007/s1071401722004. arXiv:1604.07428 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Witten E (1981) A new proof of the positive energy theorem. Commun Math Phys 80:381–402. https://doi.org/10.1007/BF01208277 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Woolgar E (1999) Bounded area theorems for higher genus black holes. Class Quantum Grav 16:3005–3012. https://doi.org/10.1088/02649381/16/9/316. arXiv:grqc/9906096 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Yazadjiev S (2013a) Horizon area–angular momentum–charge–magnetic fluxes inequalities in 5D Einstein–Maxwelldilaton gravity. Class Quantum Grav 30:115010. https://doi.org/10.1088/02649381/30/11/115010. arXiv:1301.1548 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Yazadjiev SS (2013b) Area–angular momentum–charge inequality for stable marginally outer trapped surfaces in 4D Einstein–Maxwelldilaton theory. Phys Rev D 87:024016. https://doi.org/10.1103/PhysRevD.87.024016. arXiv:1210.4684 ADSCrossRefGoogle Scholar
 Yodzis P, Seifert HJ, Müller zum Hagen H (1973) On the occurrence of naked singularities in general relativity. Commun Math Phys 34:135–148. https://doi.org/10.1007/BF01646443 ADSMathSciNetCrossRefGoogle Scholar
 Yoon JH (2004) New Hamiltonian formalism and quasilocal conservation equations of general relativity. Phys Rev D 70:084037. https://doi.org/10.1103/PhysRevD.70.084037. arXiv:grqc/0406047 ADSMathSciNetCrossRefGoogle Scholar
 Yoshino H (2008) Highly distorted apparent horizons and the hoop conjecture. Phys Rev D 77:041501. https://doi.org/10.1103/PhysRevD.77.041501. arXiv:0712.3907 ADSMathSciNetCrossRefGoogle Scholar
 Zangwill A (2013) Modern electrodynamics. Cambridge University Press, CambridgezbMATHGoogle Scholar
 Zhang X (1999) Angular momentum and positive mass theorem. Commun Math Phys 206:137–155. https://doi.org/10.1007/s002200050700 ADSMathSciNetCrossRefzbMATHGoogle Scholar
 Zhou X (2012) Mass angular momentum inequality for axisymmetric vacuum data with small trace. ArXiv eprints arXiv:1209.1605
Copyright information
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.