Magnetized Kerr–Newman–Taub-NUT spacetimes

We find a new class of exact solutions in the Einstein–Maxwell theory by employing the Ernst magnetization process to the Kerr–Newman–Taub-NUT spacetimes. We study the solutions and find that they are regular everywhere. We also find the quasilocal conserved quantities for the spacetimes, the corresponding Smarr formula and the first law of thermodynamics.


Introduction
Finding the exact solutions to the Einstein-Maxwell theory is always fascinating, as it opens a door to explore the new aspects of the gravitational physics. The exact solutions to the aforementioned theory contain the black hole solutions, such as the Kerr-Newman family, to a more general spacetime solutions of Plebanski-Demianski [1]. Different aspects of those solutions have been studied and reported, in which, some can be related to the real astrophysical phenomena, and others are still in vague. Among the latter, is the spacetime solutions with the NUT parameter, which is considered as the extension of the mass parameter. We note that the conserved quantities, such as the mass and angular momentum in a spacetime with a particular boundary, can be computed, which are related to the symmetry of the spacetime. However, the NUT parameter is not associated to any symmetry of the spacetime, and yet it also leads to some peculiar properties in the spacetime, such as conical singularity and the regular invariants such as squared Riemann tensor, at the origin of the coordinate system. Nevertheless, spacetimes with the NUT parameter has helped to shape our understanding of some gravitational and thermodynamical aspects of gravity theories [2][3][4][5][6][7][8][9][10][11][12][13][14] The Kerr-Newman spacetime is a well known black hole solution in the Einstein-Maxwell theory. The solutions can a e-mail: amg142@campus.usask.ca (corresponding author) b e-mail: haryanto.siahaan@unpar.ac.id be extended to contain the NUT parameter, and usually referred to, as the Kerr-Newman-Taub-NUT spacetimes. Despite the conical singularities in the spacetimes, there are many research works to explore the different aspects of the Kerr-Newman-Taub-NUT black holes. We note the presence of the NUT parameter in the spacetime, leads to the loss of asymptotic flatness, if the corresponding null NUT counterpart has this asymptotic [1]. It can be shown that the Kerr-Newman-Taub-NUT spacetime is a special case of the Plebanski-Demianski spacetime which is considered as one of the most general solution in Einstein-Maxwell theory that can contain black holes [15].
In Einstein-Maxwell theory, it also exists an exact solution describing a universe filled by a homogeneous magnetic field known as the Melvin universe [16]. A black hole solution in this Melvin universe can be obtained by using the Ernst magnetization [17] applied to a known black hole spacetime in Einstein-Maxwell theory as a seed. In fact, performing Ernst magnetization to the Minkowski spacetime can give us the Melvin universe. In general, the magnetization can be done in two ways, namely at the level of perturbation as Wald introduced in [18], and as a strong field as Ernst proposed in [17]. In the Wald prescription, Maxwell field is introduced perturbatively by using the Killing vectors associated to the spacetime, while the presence of homogeneous magnetic field does not change the spacetime solution. We can infer that the magnetization by Wald does not change the asymptotic structure of the magnetized spacetime. The superradiant instability in this weakly magnetized black hole had been investigated in [19], and this type of magnetization for Kerr-NUT-AdS spacetime had been performed in [20]. This Ernst magnetization itself can be viewed as a type of Harrison transformation [21] which maps an old solution to a new one in the theory. A number of aspects of the known magnetized black hole solutions had been reported in literature [22][23][24][25][26][27][28][29][30][31][32][33][34][35][36][37][38], and this shows the importance of such solution in shaping our knowledge on gravity. The most recent ones are the magnetization to Reissner-Nordstrom -Taub-NUT [39] and Kerr-Taub-NUT [40]. The work presented in this paper extends the previous works to magnetizing the Kerr-Newman-Taub-NUT (KNTN) spacetimes, which we refer to as the Melvin-Kerr-Newman-Taub-Nut (MKNTN) spacetimes. Though the idea is straightforward and the mechanism is well understood, but incorporating the functions in the solution are quite challenging.
In this paper, we perform the magnetization procedure to the Kerr-Newman-Taub-NUT spacetimes. We expect to get the magnetized KNTN spacetime solution, whose massless, null NUT, static, and neutral limit, is the Melvin magnetic universe [16]. Some aspects of the spacetime are discussed, such as the deformation of the horizon and the quasilocal conserved quantities associated to the solution.
The organization of this paper is as follows. In Sect. 2, after reviewing the Ernst magnetization process, we construct the MKNTN solutions by employing the Ernst magnetization to the KNTN metric as the seed solution. In Sect. 3, we study some properties of the MKNTN spacetimes. In Sect. 4, we obtain the quasilocal thermodynamical quantities for the MKNTN black holes, as well as the Smarr equation for the MKNTNblack holes and verify the first law of thermodynamics. We consider the natural units c =h = k B = G 4 = 1.

Ernst magnetization
Ernst magnetization is a transformation acting on a set of Ernst potentials which can be defined by using some functions appearing in the seed spacetime solution and the accompanying vector field in Einstein-Maxwell theory. The seed solution is typically expressed in the Lewis-Papapetrou-Weyl (LPW) form where f , γ , and ω are function of χ . Here we have used the − + ++ signs convention for the spacetime, and * notation representing the complex conjugation. Using the f function in the LPW line element above, accompanied by the vector A = A μ dx μ , the gravitational Ernst potential, (2.2) and the electromagnetic one can be constructed. The A t component can be obtained after solving Note that the imaginary part of is the vector field which constructs the dual field strength tensor (2.5) whereF μν = ∂ μÃν − ∂ νÃμ .
In equation above, the twist potential is given by the relation Using the Ernst potentials, the following equations can be extracted from the equations of motion in Einstein-Maxwell theory, where = 1 − 2b + b 2 E. (2.13) Here, the constant b is interpreted as the external magnetic field strength in the spacetime. 1 The transformation (2.12) leaves Eqs. (2.7) and (2.8) unchanged for the new potentials E and . In other words, the new metric consisting the functions f and ω , together with the new vector potentials A t and A φ are also solutions to the Einstein-Maxwell field equations.
In particular, the transformed line element (2.1) resulting from the magnetization (2.12) has the components and while the function γ remains unchanged. In Appendix A, we present an example, which shows the differential equations for the function γ , and the invariance of the function γ under the Ernst magnetization process. We note that typical black hole solutions in the Einstein-Maxwell theory, are more compact where they are expressed in the Boyer-Lindquist type coordinates {t, r, x = cos θ, φ}. Consequently, the LPW type metric (2.1) with stationary and axial Killing symmetries will have the metric function that depend on r and x, and the corresponding flat metric line element reads where r = r (r ) and x = x (x). Therefore, the corresponding operator ∇ will read ∇ = √ r ∂ r + i √ x ∂ x . Furthermore we can have ρ 2 = r x , then Eq. (2.4) gives us and The last two equations are useful later in obtaining the A t component associated to the magnetized spacetime according to (2.12). To end some details on magnetization procedure, another equations which will be required to complete the metric are (2.19) and In the following section, we employ this magnetization scheme to the Taub-NUT spacetime.

The Melvin-Kerr-Newman-Taub-NUT spacetimes
To obtain the desired magnetized solution, we use the Ernst potentials that belong to Kerr-Newman-Taub-NUT system, and In the form of LPW line element (2.1), the above spacetime metric, associates to the functions 25) and ρ 2 = x r . From this seed solution, one can construct the corresponding Ernst potentials as follows and (2.28) (2.29) The magnetized Ernst potentials can be obtained from the seed ones above, which yields to the magnetized metric with the new functions f and ω , while γ is unchanged.
In fact, the differential equations for the function γ (r, x), are given by and We explicitly check that Eqs. (2.30) and (2.31) and their counterparts for γ imply the metric function γ (r, x) is the same as γ (r, x), which is given by (2.25).
To summarize the results, the Melvin-Kerr-Newman-NUT black hole is given by together with the Maxwell's field where the metric functions f (r, x) and ω (r, x) are given by (2.34) and Moreover, the components of the Maxwell's field (2.33) are given by and The expressions for a i , b i , . . . , f i , as functions of r , are given in Appendix B.

Some properties of the Melvin-Kerr-Newman-Taub-NUT spacetimes
The largest root of the metric function r describes the outer event horizon of the black hole (2.32), which is given by which implies m 2 + l 2 ≥ a 2 + q 2 , to have a real value for the outer event horizon. The inner event horizon is located at r − = m − m 2 + l 2 − a 2 − q 2 . We note that the inner event horizon exists at a real positive (or zero) value r − ≥ 0, if a 2 + q 2 ≥ l 2 . Combining the former and the latter inequalities, we find the following range for the summation of squares of the rotational parameter and the electric charge of the spacetime (2.32).
The trace of the energy-momentum tensor for the Maxwell's field (2.33) is identically zero. We have verified exactly that the metric (2.32) with the Maxwell's field (2.33) satisfy exactly all the Einstein-Maxwell field equations. The Ricci scalar of the spacetime is identically zero and the Ricci square invariant R = R μν R μν is regular everywhere including r = 0. The expression for the R is very long and so we don't present it here. We also find the Kretschmann invariant where K 1 (r, x) and K 2 (r, x) are two functions with coefficients of the black hole parameters. Though the expression for K 1 (r, x) is very complicated, however 3) The location of curvature/coordinate singularities for the black hole (2.32) can be determined by the equation K 2 (r, x) = 0. In fact, beside ρ(r, x)=0, we find the following equation for the location of singularities, which is expectedly independent of the magnetic field The event horizon and x = ±1 are the roots of ρ(r, x) = 0, where K is regular and finite. Moreover, it seems f (r, x) = 0 or Eq. (3.4) leads to other singular points, however, we verify that at those points, the Kretschmann invariant K remains completely finite. We notice from Eq. (3.4) and Table 1, that the only magnetized spacetimes with the point singularity, at r = 0, are Melvin-Schwarzschild and Melvin-Reissner-Nordstrom spacetimes. All other magnetized spacetimes, i.e. Melvin, Melvin-Kerr, Melvin-NUT, Melvin-Kerr-Newmann, Melvin-Kerr-NUT, Melvin-Reissner-Norstrom-NUT and MKNTN are completely regular at r = x = 0.
In the special case, where all the black hole parameters q, m, a, l approach zero, we find the metric (2.32) reduces to The metric (3.5) describes the spacetime of axisymmetric universe filled by parallel magnetic forcelines known as the Melvin magnetic universe [16]. Interestingly, these forcelines do not contract and collapse under their own gravity, and the corresponding stability against radial perturbations had been investigated in [42]. In fact, the Maxwell's field (2.33) becomes which generates the magnetic field The Ricci scalar of the Melvin space-time (3.5) is identically zero, while the Kretschmann invariant is given by .
Moreover, we find the asymptotic behaviour of the black hole (2.32), where r → ∞, by analyzing all the metric functions at large values of the radial coordinate. We find the asymptotic expressions for the metric functions, as where → 0. Using expressions (3.10)-(3.13), we find the asymptotic metric of the black hole (2.32), as where ds 2 0,r →∞ is the asymptotic of the Melvin universe (3.5). We note that the presence of the NUT charge makes an off-diagonal term to the asymptotic metric of the black hole (2.32).
We consider now the set of 3-dimensional surfaces at a fixed value for the radial coordinate r = . The induced metric on the surface r = , is given by The determinant of the metric (3.15) is .
(3.16) We note that due to the presence of the NUT charge as well as the magnetic field, the horizon geometry is a distorted sphere. In fact the two-dimensional horizon is given by the line element The two grand circles on the horizon, one at x = 0 and the other passing through x = ±1, have two different circumferences. The former circumference is given by In Fig. 1, we plot the typical behaviour of the equatorial circumference versus the NUT charge l and the magnetic field b, where we set a = 6, m = 7 and q = 1. The latter circumference involves an integral, which we can't find it explicitly as an exact form. In Fig. 2, we plot the result of numerical integration for the circumference, as a function of the magnetic field, where we set the other black hole parameters as a = 6, l = 1, m = 7 and q = 1.
The other interesting surfaces for the black hole (2.32) are the stationary limit surfaces. The stationary limit surfaces r = r s (x) are the roots of equation  which of course, is not solvable by radicals. Hence, we consider the black hole with the same parameters a = 6, b = 2, l = 1, m = 7 and q = 1, which we considered before in this section. The inner and outer event horizons are at r − = 3.4 and r H = 10.6, respectively. In Fig. 3, we plot the function S(r, x) versus r ≥ r H and −1 ≤ x ≤ 1. The outer stationary limit surfaces r = r s (x) are the intersection of the curve with horizontal plane at 0, where S(r s , x) = 0. The outer ergoregion for the black hole (2.32), is the region between r = r H and r = r s (x), where S(r, x) > 0.
In Fig. 4, we plot the function S(r, x) versus r ≤ r − and −1 ≤ x ≤ 1. The inner stationary limit surfaces r s (x) are the intersection of the curve S(r, x) with the horizontal plane at 0, where S(r s , x) = 0. The inner ergoregion for the black hole (2.32), is the region between r = r s (x) and r = r − , where S(r, x) > 0. We also note that for r − < r < r H , the function S(r, x) is positive everywhere, as shown in Fig. 5. Now, we consider the rich structure of the electromagnetic fields, on and outside the event horizon. The electromagnetic field components are given by whereF, the Hodge dual of two-form F, is given in (2.5), and u μ is the 4-velocity of the observer. We find the exact forms for the electromagnetic fields, though their expressions are quite long, and so we do not present them here. In Fig. 6, we plot the typical behaviour of the components of the magnetic field, outside the event horizon, for a black hole with parameters a = 6, q = 1, m = 7, l = 1 and b = 2. Moreover in Fig. 7, we plot the typical behaviour of the components of the electric field, outside the event horizon, for the same black hole parameters. In Figs. 8 and 9, we plot the behaviour of the electromagnetic fields on the event horizon. We notice the minimum and maximum of B r and E x appear quite away from the equatorial plane, however the maximum of B x and the minimum of E r occurs almost on the equatorial plane.
We also plot the polar electromagnetic fields outside the even horizon in Figs. 10 and 11.
We should notice that increasing behaviour in the polar electromagnetic fields doesn't extend to large values of the radial coordinate. In fact, we find the asymptotic behaviour of the electromagnetic fields for r → ∞, as The r (left) and x (right) components of the magnetic field, versus r and x, for a black hole which the event horizon is located at r + = 10.6. We set the black hole parameters as a = 6, q = 1, m = 7, l = 1 and b = 2 Fig. 7 The r (left) and x (right) components of the electric field, versus r and x, for a black hole which the event horizon is located at r + = 10.6. We set the black hole parameters as a = 6, q = 1, m = 7, l = 1 and b = 2

Fig. 8
The r (left) and x (right) components of the magnetic field, on the event horizon, versus x, for a black hole which the event horizon is located at r + = 10.6. We set the black hole parameters as a = 6, q = 1, m = 7, l = 1 and b = 2 for an observer with the 4-velocity u μ = (1, 0, 0, 0).

Thermodynamics of the Melvin-Kerr-Newman-Taub-NUT spacetimes
In this section, we discuss the thermodynamical quantities for the black hole (2.32) and then construct the mass of the black hole according to the Smarr relation. We should emphasis that using the term "black hole" for the metric (2.32), with the NUT charge parameter l must Fig. 9 The r (left) and x (right) components of the electric field, on the event horizon, versus x, for a black hole which the event horizon is located at r + = 10.6. We set the black hole parameters as a = 6, q = 1, m = 7, l = 1 and b = 2

Fig. 10
The r (left) and x (right) components of the polar magnetic field, outside the event horizon, versus r , for a black hole which the event horizon is located at r + = 10.6. We set the black hole parameters as a = 6, q = 1, m = 7, l = 1 and b = 2. In the left figure, the down curve is for x = 1 and the up curve is for x = −1. In the right figure, the down curve is for x = −1 and the up curve is for x = 1

Fig. 11
The r (left) and x (right) components of he polar electric field, outside the event horizon, versus r , for a black hole which the event horizon is located at r + = 10.6. We set the black hole parameters as a = 6, q = 1, m = 7, l = 1 and b = 2. In both figures, the down curve is for x = 1 and the up curve is for x = −1 be taken with some cautions, to avoid contradictions with the black hole uniqueness theorems. The event horizon is a global concept and, therefore, requires asymptotic flatness to be well defined. On the other hand, the NUT charge parameter l makes the spacetime (2.32) asymptotically locally flat, and so violates the global flatness condition. Hence, in general, a global event horizon does not exist. However, since the spacetime (2.32) possesses many physical quantities, similar to a black hole (such as event horizons,. . .), we often refer to the spacetime (2.32), as a black hole. 2 The surface gravity for the black hole (2.32) is given by where ξ μ is the Killing vector ξ μ = ∂ ∂t + H ∂ ∂φ and H is the angular velocity of the horizon, which is given by and We note that for b = 0, we recover the angular velocity for the Kerr-Newman-NUT black hole [31]. In Figs. 12 and 13, we plot the angular velocity of the horizon versus different black hole parameters.
We find the surface gravity of the black hole is the same as the Kerr-Newman-NUT black hole, and so the Hawking temperature is The Coulomb potential on the horizon is given by where we add a constant term H to the potential to make it regular at r → ∞. The expression for the H is long and so we do not present it explicitly here. The area of the horizon is given by where is the two-dimensional hypersurface, parameterized with (θ, φ) in (2.32) and E ⊥ is the normal component of the electric field on . We find where +36 l 2 m 2 q 3 r 4 − 36 l 2 mq 5 r 3 − 48 l 2 mq 3 r 5 + 12 l 2 q 7 r 2 +28 l 2 q 5 r 4 + 16 l 2 q 3 r 6 + q 7 r 4 . (4.12) Moreover, the functions s i , i = 0, . . . , 4 are given by (4.13) In the limit of l → 0, we recover exactly the results of [31] for the electric charge. In Figs. 14 and 15 we plot the electric charge (4.10) of the black hole, versus different black hole parameters. We then find the angular momentum J of the black hole (2.32), according to [31] * dP, (4.14) where P μ = ∂ ∂φ is the Killing vector in φ-direction. We find the integrand in (4.14) is given by (4.15) A straightforward and lengthy calculation shows that we get  (4.17) where  .17) is where J is given by (4.16), and J N and J S are contributions to the angular momentum from NUT charges on the north and south poles. The total mass for the black hole (2.32) is given by where T μ = ∂ ∂t is the time-like Killing vector. Evaluating the integral in Eq. (4.19), where b = 0, we find (4.20) Moreover, the thermodynamic potentials ψ N and ψ S are given by [43] ψ N = 1 4π where ξ μ N is the Killing vector which generates the south pole Killing horizon, due to the presence of the NUT charge. The functions N and S are given by (4.25) Using the form of the Killing vectors on the north and south pole horizons, we find where (r, x, ) is given by As expected, we notice that neither the magnetic field nor the other parameters of the black hole (except the NUT charge) contribute to the thermodynamic potentials ψ N and ψ S , where r → ∞. The independence of ψ N and ψ S from other parameters of the black hole and the magnetic field, is consistent with the notion of considering them as the "surface gravity" over the NUT tubes [43], similar to κ as the surface gravity (4.1) over the horizon r H . The thermodynamic charges N N and N S are given by where the integrals are over the very narrow NUT tubes T N and T S , along the positive and negative z-axis, where r H ≤ z < ∞, x 1 and −∞ < z ≤ −r H , x −1, respectively. We find the integrand in Eqs. (4.31) and (4.32), is given by Evaluating the integrals in (4.31) and (4.32), we find very long expressions for the N N and N S in terms of the black hole parameters a, l, m and q and the magnetic field parameter b which are unfeasible to present here. We verify that in the special cases, the thermodynamic NUT potentials reduce exactly to the well-known results in [43].
As we notice from Eq. (3.14), the NUT charge contributes to an off-diagonal term in the asymptotic of MKNTN black holes (2.32). As a result, we find two contributions to the total angular momentum (4.18) from the north and south NUT tubes, which are given by respectively, where P μ = ∂ ∂φ is the space-like Killing vector. We find the integrand in Eqs. (4.34) and (4.35), is given by Evaluating the integrals in (4.34) and (4.35), we find very long expressions for the J N and J S in terms of the black hole parameters a, l, m and q and the magnetic field parameter b which are unfeasible to present here. We verify that in the special cases, the thermodynamic NUT potentials reduce exactly to the well-known results in [43].
To complete the calculation, we also mention that the area S N and S S of the north and south NUT tubes are given by where Z stands for N and S, respectively. After calculating the integrals in (4.37), (4.38), we find that S N and S S , are independent of the magnetic field parameter b, and are given by where r ∞ − r H is the length of the north (south) NUT tube along the positive (negative) z-axis.  .17) is indeed the Smarr relation for the MKNTN black hole (2.32). Moreover, by construction of different thermodynamical quantities for the black hole (2.32), and the Smarr equation (4.17), we find the first law of thermodynamics, as given by We also verify that Eq. (4.41) reduces exactly to the wellknow first law of thermodynamics for the Kerr-NUT and Melvin-Kerr-Newman black holes in [43] and [31].

Concluding remarks
In this work, we have constructed a new class of the exact solutions to the Einstein-Maxwell theory in four dimensions which describes the immersion of the Kerr-Newman-Taub-NUT spacetimes in an external magnetic field. The solutions are obtained by applying the Ernst magnetization procedure to the four dimensional Kerr-Newman-Taub-NUT spacetime as the seed. We discuss the properties of the MKNTN spacetimes and show that they are completely regular at r = 0, x = 0. Particularly, in addition to the extensive investigation on the space-time structure and the ergoregions for the MKNTN black hole, we also study the behaviour of the electromagnetic fields in the magnetized spacetime solutions. We find that the horizon has a non-trivial topology which leads to the eccentric horizon. The thermodynamical quasi-local conserved quantities of the spacetime are obtained, though they are generally quite complicated functions of the five independent parameters of MKNTN black holes, namely m, a, l, q and b. In this paper, we also establish the Smarr formula for the MKNTN black holes. Finally, we study the thermodynamics of the MKNTN spacetimes, and show the corresponding first law of thermodynamics.
Acknowledgements This work was supported by the Natural Sciences and Engineering Research Council of Canada.

Data Availability Statement
This manuscript has associated data in a data repository. [Authors' comment: The paper is posted on arXiv:2103.04865].
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecomm ons.org/licenses/by/4.0/. Funded by SCOAP 3 .

A The differential equations for γ (ρ, z) in the LPW metric
Let us first consider the LPW metric (2.1) with dχ dχ * as given in (2.9), namely (A.1) If the system is electrovacuum, the accompanying vector field A μ obeys Eq. (2.4), which then constructs the Ernst electromagnetic potential , as defined in (2.3). Now let us write the Einstein equations for the electrovacuum system as In fact, the Ernst equations (2.7) and (2.8) are obtained from the last equations. Particularly, from the E ρz component of (A.2), we have while from E ρρ and E zz , we find Obviously, the LPW metric functions for the Minkowski spacetime are f = ρ 2 , γ = ln ρ, and ω = 0. Accordingly, the Ernst potentials for this seed solution are E = ρ 2 and = 0, which give us = 1 + b 2 ρ 2 . The magnetized Ernst potentials then read On the other hand, the associated vector solution obtained from has the components A φ = − −1 bρ 2 andÃ φ = 0, which lead to A t = 0, according to (2.4). Suppose we consider the function γ in the Melvin spacetime, changes under the Ernst magnetization. To find γ , we need to solve the remaining non-zero equations (A.2) for the Melvin spacetime with (A.6) and (A.7). They are given by and We notice that the absence of external magnetic parameter b in the last three equations tells us that γ obeys the same differential equations as γ . Hence, we can conclude γ for the Melvin spacetime is the same as γ for the Minkowski spacetime. We also note that we get the same result γ = γ , by plugging Eqs. (A.6) and (A.7) in (A.3) and (A.4). In regard to the Melvin-Kerr-Newman-Taub-NUT spacetime (2.32) with the Boyer-Lindquist type coordinate (t, r, x = cos θ, φ), we find the following differential equations for γ (r, x) from the Einstein equations, We have explicitly checked that the seed solution { f, ω, } that belongs to Kerr-Newman-Taub-NUT system, and the magnetized version f , ω , (as presented in Sect. 2.2), obey (A.11) and (A.12), with the same γ function, as given by (2.25).