Arxiv:1904.02035V2 [Gr-Qc] 15 Oct 2019 I.Knmtcrsrcin 5 Restrictions Kinematic III

arXiv:1904.02035v2 [gr-qc] 15 Oct 2019 I.Knmtcrsrcin 5 restrictions Kinematic III. V nryetato 7 extraction Energy IV. ∗ ‡ † I oinadcliin fts atce ln the along particles test of collisions and Motion II. .Drvtvsof Derivatives A. .Itouto n ocuin 1 conclusions and Introduction I. [email protected] [email protected]ff.cuni.cz fi[email protected] 4 axis References .Rmrso ls Iciia atce 5 particles critical II class on Remarks D. .Eutoso oinadeetv oeta 3 potential effective and motion of Equations A. .Apiaino osrainlw 7 laws conservation of Application A. .Gnrlfrua o rtclprils5 particles critical for formulae General A. .Mto towards Motion C. .Dsuso fcvas9 caveats of Discussion C. .Ciia atce n olso nry3 energy collision and particles Critical B. .Knmtcrgms8 regimes Kinematic B. .RslsfrteKr-emnslto 6 solution Kerr-Newman the for Results B. nttt fMteaisadMcais aa eea Univ Federal Kazan Mechanics, and Mathematics of Institute particles .Seiccag uo 10 9 9 cutoff charge Specific 3. problem mass Neutral 2. problem feeding Energy 1. xrcino nryfo netea oaigelectrovacu rotating extremal an from energy of Extraction ept h ako nodtoa ieai bounds. t kinematic due can unconditional charge” that of “Wald caveats lack small numerous the find a despite we have Nevertheless, and extr fields. spinning high-energy netic fast of be possibility can the Schnittma which brings the al This including particles test also charge particles. charged hole, the of black for motion electrovacuum the found ing m for were on replicated collision bounds be the can upper in unconditional produced no Bañados-Silkparticle significantly process, the differs BSW of it original analogy However, the an holes. is black This fin Kerr extremal is horizon. particles the to the close of one if Reissner-Nordström spacetime iheeg olsoscnocrfrrdal oigcharg moving radially for occur can collisions High-energy W CONTENTS 1 and etoMliicpia eAto´sc ETA Departa CENTRA, Astrof´ısica – de Multidisciplinar Centro hre nvriy oesvˇkah2 8 0Pau ,C 8, Prague Holeˇsoviˇckách 00 V 180 University, 2, Charles r 2 0 nttt fTertclPyis aut fMteaisa Mathematics of Faculty Physics, Theoretical of Institute V n erycritical nearly and 3 nttt ueirTenc S,Uiesdd eLso – Lisboa de Universidade IST, Técnico – Superior Instituto atcecliin ln h xso symmetry of axis the along collisions Particle ainlUiest,4SooaSur,Kakv602 Ukra 61022, Kharkov Square, Svoboda 4 University, National ± eateto hsc n ehooy hro .N Karazin N. V. Kharkov Technology, and Physics of Department .Hejda, F. vnd oic as1 0901Lso,Portugal Lisboa, 1049-001 1, Pais Rovisco Avenida 1, ∗ .Biˇcák,J. 11 12 2, 3 4 † n .B Zaslavskii B. O. and rcs,wihol ok nieteegshr.How- ergosphere. the Penrose inside the works to only related which directly process, is mo- angular it of moreover, magnitude mentum; and sign specific as with particles seen processes. be other can those The holes for scenario black backreaction.) best-case extremal to the devia- for due survive effects e.g. also simpler geodeticity, can from discussed energies tions was ar- collision it with high 13] processes ex- conditions, [12, bitrarily for weak in relatively addition, under cf. (In that holes; processes 11]. black collision [10, subextremal other ample of to variety even a also applicable is (see there potential Third, effective crit- the (nearly) [9]). by with reflected process particles the Schnittman ical of when variant particular a extracted in introduced on phe- hole, [8] bounds black BSW-type upper a the of from revision of energy the investigation terms to the geometrical led nomena Second, to their related to 7]. due nature [6, own their their and represent on ubiquity processes issue interesting BSW-type theoretical is the intriguing effect” First, an “BSW ways. references). criti- many and the early in summary an this, for for [5] despite in [4] However, 3.1 e.g. Section unfeasible and (see be cism caveats to ener- numerous out to turned to supercollider scale due particle Planck a the as beyond acting gies idea hole original black The a take radius. of horizon we the if to point bound, collision without the centre- grows the energy hole, collision black Kerr of-mass extremal de- an [3] (“critical”) around fine-tuned particles (BSW) involving collisions West for that and scribed Bañados, Silk after attention riy 8KelosaaSre,Kzn400,Russia 420008, Kazan Street, Kremlyovskaya 18 ersity, h rgnlvrino h S ffc eurstest requires effect BSW the of version original The of lot a received 2] [1, process Penrose collisional The .ITOUTO N CONCLUSIONS AND INTRODUCTION I. cincoe oatohsclbakholes, black astrophysical to closer action n h xso eea xrmlrotat- extremal general a of axis the ong ntrso nryetato:ulk for unlike extraction: energy of terms in -ue n h olso on staken is point collision the and e-tuned -yepoeswt eetdfine-tuned reflected with process n-type Ws BW ffc,fis ecie for described first effect, (BSW) -West eso.W hwta hs results these that show We version. d aeteeeg xrcinunfeasible extraction energy the make dts atce nteextremal the in particles test ed s n nryo necpn test escaping an of energy and ass neato ihetra mag- external with interaction o et eF´ısica, de mento ,4, 3, ehRepublic zech dPhysics, nd ‡ UL, ine mbakhole: black um 2 ever, an analogous effect is possible for radially mov- how this approach can be rigorously related to the ex- ing charged particles in the extremal Reissner-Nordström pansion coefficients of the radial equation of motion (in- spacetime [14]. Since this is a non-rotating charged black cluding the relaxation time). In III B we give particular hole (without an ergosphere), the origin of the effect is results for the extremal Kerr-Newman spacetime, which purely electrostatic in this case. Furthermore, for this show that for small values of the black hole charge the electrostatic variant it turned out that, in the test parti- critical particles must be highly relativistic in order to be cle approximation, there is no bound on the mass and the able to approach the horizon radius. energy of an escaping particle produced in the collision [15]. This is in sharp contrast to the original “centrifugal” BSW effect, where unconditional bounds exist [16] In Section IV we deal with the energy extraction. First, (cf. also [17, 18]). we briefly review how to rearrange the conservation laws Astrophysical black holes are expected to be sur- to prove that, for a 2 2 process, a collision of a critical rounded by external magnetic fields, and it has been particle with an incoming→ “usual” (i.e. not fine-tuned) proven in various contexts [19, 20] that a black hole particle necessarily leads to the production of a nearly immersed in an external magnetic field can maintain a critical particle and an incoming usual particle. Then non-zero charge. However, this so-called “Wald charge” we study whether the produced nearly critical particle will be very small, and thus there is a good motivation can escape and extract energy. We find that there are to study generalisations of the radial electrostatic BSW- two threshold values, one for mass and one for energy. If type effect to black holes with a smaller charge than in the nearly critical particle is produced below/above the the extremal Reissner-Nordström case. mass threshold, it is initially outgoing/incoming. Be- There are two natural ways of generalisation. First, low the energy threshold the particle must be produced one can include effects of angular momentum of the par- with such a value of charge that the corresponding criti- ticles and of the dragging from the rotation of the black cal energy will be lower than the actual energy, whereas hole and study overlapping and transition between the above the threshold the critical energy corresponding to electrostatic and the centrifugal BSW-type effect. A de- the charge is above the actual energy. These results qual- tailed analysis of this way of generalisation was given itatively agree with the special case [15]. Here we focus in [21] (concerning only the approach phase of the pro- on comparing the BSW-type process (collision with an in- cess). On the other hand, one can keep the restriction to coming critical particle) and the Schnittman process (col- “purely radial” motion, which is possible in any axially lision with an outgoing, reflected critical particle). For symmetric spacetime for particles moving along the axis instance, a particle that is initially incoming with energy of symmetry. In the present paper, we study this case above the critical energy will fall into a black hole. There- and show that the interesting results for the extremal fore, a particle that is produced with mass above the Reissner-Nordström black hole can be replicated even in threshold must have the energy also above the respective models closer to astrophysical situations. threshold in order to avoid this. In IV C we show that this The paper is organised as follows. In Section II, we may not be generally possible by considering a toy model review the basic features of electrogeodesic motion along of interactions of microscopic particles (cf. the “neutral the axis of symmetry of a general stationary axially sym- mass” problem). However, this problem occurs only for metric black-hole spacetime, including the local defini- the BSW-type kinematics. Thus, the Schnittman vari- tion of the critical particles. We review why they cannot ant again fares better. The problem is actually related to approach the horizon for subextremal black holes and other two caveats for microscopic particles spotted earlier how they cause the divergent behaviour of the centre-of- [15, 22]. As the energy of (nearly) critical particles is pro- mass energy in the limit of the collision point approaching portional to their charge, the (nearly) critical microscopic the horizon radius. In IIC we recall that the trajectory particles need to be highly relativistic (i.e. the “energy of a critical particle is approximated by an exponential feeding” problem), and also the produced particle must relaxation towards the horizon radius. Because of this, have a higher charge than the initial one, which limits any collision event involving a critical particle must al- the efficiency. Finally, we show that the energy feeding ways happen at a radius greater than the horizon radius. problem for microscopic particles may be reduced by six Therefore it makes sense to consider also particles that orders of magnitude if we go from the maximal value of behave approximately as critical at a given collision ra- the black hole charge for the Reissner-Nordström solu- dius (so-called nearly critical particles). In II D we show tion to some minimal value required for the processes to that doubly fine-tuned critical particles with infinite re- be possible. Despite this, the critical microscopic par- laxation time exhibit an inverse power-law behaviour and ticles would still have to be highly relativistic, which is thus approach the horizon radius much more slowly. in sharp contrast to the behaviour for a small black hole In Section III we study restrictions on the values of en- charge (“mega-BSW effect”) seen for the equatorial elec- ergy and charge of critical particles in order for them to trogeodesic case in [21]. Concluding, our analysis of those be able to approach the radius of the degenerate horizon. details motivates further study of energy extraction from Our discussion is based, similarly to [21], on derivatives black holes through the generalised collisional Penrose of a certain effective potential. In the Appendix we show process with charged particles. 3

II. MOTION AND COLLISIONS OF TEST (specific energy and specific charge). Denoting PARTICLES ALONG THE AXIS 2 2 W = (ε +˜qAt) N , (4) − We start from a general axially symmetric stationary metric in the form the condition for the motion to be allowed can be stated as W > 0. Furthermore, since N 2 > 0 outside of the 2 2 2 2 2 black hole, we can prescribe the decomposition of W , g = N dt +gϕϕ (dϕ ω dt) +grr dr +gϑϑ dϑ . (1) − − W = (ε V ) (ε V−) , (5) − + − The metric components gϕϕ,grr,gϑϑ and functions N,ω in terms of V± that read are independent of t and ϕ; the metric is suitable to de- scribe an equilibrium state of a black hole. We consider V± = qA˜ t N. (6) also an electromagnetic field with potential in the form − ± In order for W to be non-negative, it must hold either > 6 A = At dt + Aϕ dϕ , (2) ε V+ or ε V−. However, only the first variant is t consistent with p > 0. Thus, we define V V+ and consider only ε > V as the condition for the≡ motion to with At, Aϕ independent of t and ϕ. We assume that the outer black-hole horizon (where N = 0) corresponds to be allowed. ε = V is the condition for a turning point. r = r+. For extremal black holes, we denote the position of their degenerate horizon by r = r . 0 B. Critical particles and collision energy

Conditions (pr)2 > 0 and pt > 0 noted above have A. Equations of motion and effective potential H H further implications. Particles with E + qAt > 0 (At denotes At at r+) can fall into the black hole. For pho- Let us consider the motion of charged test particles tons, this is the sole option as they have E > 0, q = 0. along the axis of symmetry. The (semi)axis forms a two- Thus, unlike in the equatorial case (see e.g. [8, 16]), pho- dimensional submanifold. We can use two integrals of tons along the axis are not so interesting. Turning to motion therein, which are related to the Killing vector massive, charged particles, there is also a possibility for H ∂/∂t and to the normalisation of the momentum. The ax- ε +˜qAt < 0, which corresponds to particles that cannot ial motion is thus fully integrable. The first-order equa- get close to the black hole, so it is also uninteresting for a tions of motion for a particle with rest mass m and charge generalised BSW effect. However, we can consider mas- H q read sive, charged particles with ε +qA ˜ t = 0. These are on the verge between the previous cases, and hence they are usually called critical particles.1 (To complement, parti- t E + qAt r 1 2 2 2 p = , p = σ (E + qAt) m N . cles that are not critical are called usual in the literature.) N 2 N 2g − r rr Critical particles appear to have a turning point at the h (3)i horizon radius, as seen e.g. through the fact that their specific energy, εcr, is equal to the value of the effective Here E has the interpretation of the energy of the parti- potential at the horizon cle and σ = 1 distinguishes the outward/inward radial motion. ± εcr = q˜ At r=r+ = V r=r+ . (7) The motion can be forbidden in some intervals of r due − | | to the presence of the square root in the expression for Nevertheless, their trajectories actually do not reach a pr; we require (pr)2 > 0. Let us assume that the product turning point, which we discuss in the next section(s). 2 N grr (which is equal to the volume element at the axis) Why are the critical particles interesting for collision pro- is finite and non-vanishing at the axis, even for N 0. cesses close to the horizon? → For photons, we put m = q = 0, and their kinematics is The formula for centre-of-mass collision energy reads thus described by only one parameter E. Their motion (see e.g. [10] for more details) is allowed for any E = 0. In order to have pt > 0, we 6 2 2 2 α β restrict to E > 0. E = m + m 2gαβp p . (8) CM 1 2 − (1) (2) On the other hand, the kinematics of massive particles is characterised by two parameters ε E/m andq ˜ q/m Plugging in the equations of axial motion (3), we get ≡ ≡

1 Some authors (see e.g. [3]) define the critical particles in a dif- [23]), which is more general. Both notions become compatible ferent way, such that they are on the brink of being able to reach for extremal, asymptotically flat black hole spacetimes. the black hole from infinity. We follow the local definiton (cf. 4

2 2 2 (E1 + q1At) (E2 + q2At) 2 2 2 2 2 2 2 E = m + m +2 σ σ (E + q At) m N (E + q At) m N . (9) CM 1 2 N 2 − 1 2 N 2 1 1 − 1 2 2 − 2 q q In order to consider the N 0 limit (i.e. the collision point arbitrarily close to the horizon radius) in the case of a → H collision involving a critical particle, we examine the expansion of W around r+ with ε +˜qAt = 0:

2 2 2 2 . ∂ N 2 ∂At 1 ∂ N 2 W = (r r+)+ q˜ 2 (r r+) + ... (10) − ∂r − " ∂r − 2 ∂r # − r=r+ r=r+

The first radial derivative of N 2 at the horizon is proportional to surface gravity of the horizon and is non-negative. Let us first consider a generic, subextremal black hole (with non-zero surface gravity). We see from (10) that for some r sufficiently close to r+, expression W will become negative due to the linear term and, therefore, critical particles cannot approach r+ for subextremal black holes (note that W appears under the square root in (3); cf. (4)). In order to consider collisions with (precisely) critical particles arbitrarily close to the horizon radius, we thus have to turn to 2 2 2 2 extremal black holes (see e.g. [21, 23, 24] for more detailed analysis). Then we can use N = (r r0) N˜ , where N˜ can be (at least formally) defined as −

∞ n 2 2 1 ∂ N N˜ (r r )n−2 . (11) ≡ n! ∂rn − 0 n=2 X Evaluating (9) for a collision of a critical particle 1 and a usual particle 2, we ﬁnd that the leading order behavior in the r r limit is → 0 2 2 2 E2 + q2At ∂At 2 ∂At 2 ˜ 2 ECM 2 q1 q1 m1N . (12) ≈ r r0  ˜  ∂r ∓ s ∂r −  − N   r=r0,ϑ=0  

 

The sign corresponds to σ σ = 1. However, for We denoted ∓ 1 2 ± the usual particle one should consider only σ2 = 1 (cf. 2 [25] for detailed reasoning). With this restriction,− the N grr g˜rr 2 . (14) sign means just σ1 = 1. The scenario with incom- ≡ N˜ ing∓ particle 1 (upper sign,∓σ = 1) was ﬁrst described 1 Then the approximate solution valid for late proper times by Ba˜nados, Silk and West for the− extremal Kerr case is an exponential “relaxation” towards r in [3] and was generalised to charged particles in [14]. 0 The collision process with an outgoing critical particle . τ (σ = +1) was introduced by Schnittman [8] in a numer- r = r0 1+exp + ... (15) 1 −τ ical study focused again on uncharged particles in the relax extremal Kerr spacetime. (Analytical treatment of the 1 1 ∂2W = . (16) Schnittman process was considered e.g. in [26, 27].) τ 2 ∂r2 relax s2N˜ g˜rr r=r0

Since no critical particle can ever reach r0, the collision with another particle can only happen at some ra- C. Motion towards r0 and nearly critical particles dius rC > r0. Because of this, the difference between usual and critical particles gets blurred. Indeed, a usual We have seen that for critical particles the centre-of- particle with energy very close to the critical energy will mass collision energy with an usual particle diverges in effectively behave as a critical particle at some radius rC the limit r r0. However, the energy attainable in close to r0 provided that such a thought→ experiment is always finite, although un- bounded, because critical particles are not able to reach ε r 1 C 1 . (17) r0 in a finite proper time. To demonstrate this, let us − ε ∼ r − cr 0 expand the equation of radial motion (3) near the radius of the degenerate horizon, Such particles are called nearly critical. Nearly critical particles with ε<εcr cannot fall into the black hole, and they have a turning point at some pr dr . 1 ∂2W = (r r ) + ... (13) radius smaller than rC. Thus, it makes sense to consider m dτ 0 2 ∂r2 ≡ − − s2N˜ g˜rr also the outgoing nearly critical particles. Furthermore, r=r0

5 if the turning point is much closer to r0 than the desired The other way is to consider the ε > V condition. Let collision point rC or, more precisely, if us recall that the energy εcr of a critical particle is equal to the value of V at r0 (7). Therefore, if the effective ε r 0 < 1 C 1 , (18) potential V grows for r > r0, we will get εcr < V , and − ε ≪ r − cr 0 the motion of the critical particle towards r0 is forbidden. Thus, to see whether a critical particle can approach r , such outgoing nearly critical particles effectively behave 0 we need to check whether the first radial derivative of V as precisely critical at r . This is the motivation behind C at r is negative. Furthermore, we should also look at including outgoing critical particles in the Schnittman 0 the second derivative of V at r , since it will determine process. 0 the trend of V , if the first one is zero. However, both approaches are equivalent. For critical t D. Remarks on class II critical particles particles with p > 0, it can be shown (see Appendix) that One of the least studied aspects of the BSW-like phe- ∂2W ∂V sgn 2 = sgn . (21) nomena is what happens when relaxation time τrelax in ∂r − ∂r r=r0 r=r0 (16) is infinite, i.e when the leading order of the expansion of W in r r0 for a critical particle is the third An analogous statement (cf. (A7)) can be made for class 2 one instead of the− second. The critical particles with II critical particles, and for our present setup, it actually this property are called the “class II” critical particles holds that by Harada and Kimura [28] (“class I” standing for the ∂3W ∂2V generic critical particles with finite τ ). = 6 N˜ . (22) relax ∂r3 − ∂r2 For the class II critical particles the expansion of the r=r0,ϑ=0 r=r0,ϑ=0 radial equation of motion (3) at r turns to 0 Let us proceed with the analysis based on V . For an extremal black hole, it is possible to write down an ar- dr . 3 1 ∂3W = (r r ) 2 + ... (19) bitrary (n-th) order derivative of V with respect to r as dτ − − 0 2 ∂r3 s6N˜ g˜rr follows: r=r0 ∂nV ∂nA ∂n−1N˜ ∂nN˜ This leads to the following approximate solution (which t n = q˜ n + n n−1 + (r r0) n . (23) describes outgoing critical particles for τ and ∂r − ∂r ∂r − ∂r ingoing ones for τ ): → −∞ At r , this simplifies to → ∞ 0 2 n n n−1 ˜ 1 24N˜ g˜ ∂ V ∂ At ∂ N r = r + rr + ... (20) = q˜ + n . (24) 0 2 ∂3W ∂rn − ∂rn ∂rn−1 τ 3 r=r0 ∂r ! ! r=r0,ϑ=0 r=r0

It is possible to solve for the value ofq ˜, for which this ex- This type of trajectory was previously considered in the equatorial geodesic case in [29] (be aware of typographic pression becomes zero, and evaluate also the correspond- errors in equation (91) therein). Because of their inverse ing energy of the critical particle using (7). In particular, power-law behaviour, class II critical particles approach for n = 1, we get r much more slowly than class I critical particles with 0 ˜ their exponential approach. N q˜II = ∂At , (25) ∂r r=r0,ϑ=0

III. KINEMATIC RESTRICTIONS and ˜ A. General formulae for critical particles NAt εII = ∂At . (26) − ∂r r=r0,ϑ=0 Critical particles can, in principle, approach the hori- If we denote zon radius only for extremal black holes. Whether their motion towards r0 is really allowed will depend on their ∂At ∂r values of chargeq ˜ as well as on the properties of a par- α , (27) ≡− NA˜ t ticular extremal black hole spacetime. r=r0,ϑ=0

One way to figure out the conditions for the approach to be allowed is to look at the expansions of the radial equation of motion. For class I critical particles the re- 2 Since equations (19), (20) and (A7) are applicable also to the laxation time τrelax in (16) must be a real number, and equatorial case, they can be used to relate rigorously the results for class II critical particles the square root on the right- about the second derivative of V in [21] (in Sections IV E and V hand side of (19) must also be real. B) to the kinematics of the class II critical particles. 6 and assume α> 0 (this corresponds to a plausible choice Particles moving along the axis of an extremal Kerr- of gauge constant for At), we can state that class I critical Newman black hole are critical if their specific energy particles are allowed to approach r0, whenever αε > 1. and charge are related by For class II critical particles, it holds αε = 1. Plugging (25) into (24) with n = 2, we obtain qQ˜ Q2 + a2 εcr = 2 2 . (34) 2 Q +2a 2 ∂ At p ∂ V 2 ∂N˜ = N˜ ∂r +2 . (28) ∂r2 ∂At ∂r In general, the first radial derivative of V at the de- r=r0 − ∂r ! r=r0,ϑ=0 generate horizon is

Class II critical particles are allowed to approach r0 if ∂V Q3 1 this expression is, for a given spacetime, negative. = q˜ + . (35) ∂r − 2 2 2 2 2 Let us note that for a particle of any kind moving at a r=M (Q +2a ) Q +2a radius rC close to r0, the expansion of V to linear order It becomes zero for particles withp the specific charge can be expressed as given by ˜ V = εcr + NH (1 αεcr) (rC r0)+ ... (29) 3 − − Q2 +2a2 2 q˜II = , (36) Here εcr(˜q) is given by (7); if αεcr(˜q) > 1, the linear Q3 coefficient is negative. However, for particles that behave as nearly critical around rC, their actual energy ε is by and if these particles are critical, their specific energy is definition (17) close to the critical one. Therefore, we can use αε > 1 also as a condition for the existence of escape 1 (Q2 + a2) (Q2 +2a2) ε = . (37) trajectories of nearly critical particles (unless αε is very II ≡ α Q2 close to 1), which is discussed in IV B. p Class I critical particles are allowed to approach r = M, whenever αε > 1. For class II critical particles αε = 1. B. Results for the Kerr-Newman solution Let us note that α 6 1 for any Q and a. Therefore the condition αε > 1 implies ε> 1 (i.e. E>m). No bound For the Kerr-Newman solution with mass M, angular critical particles can approach r = M along the axis of momentum aM (convention a > 0), and charge Q the the extremal Kerr-Newman spacetime.3 Furthermore, we metric (1) reads can see that α Q2. Thus, for Q very small, only highly relativistic critical∼ particles (ε 1) can approach r = M ∆Σ A a 2 along the axis.4 ≫ g = dt2 + sin2 ϑ dϕ 2Mr Q2 dt + − A Σ − A − The second derivative of V at r = M is h i Σ 2 2 + dr + Σ dϑ , ∂2V Q2 2a2 2 Q2 + a2 ∆ 2 2 2 = 2˜qQ Q + a − 3 3 . (30) ∂r 2 2 − 2 2 2 r=M (Q +2a ) (Qp +2a ) p (38) where Inserting (36), or evaluating (28), we get ∆ = r2 2Mr + a2 + Q2 , − ∂2V 4a2 Q2 + a2 2 2 2 Σ = r + a cos ϑ , (31) 2 = 3 . (39) ∂r − 2 2 2 2 r=M Q (Qp +2a ) A 2 2 2 2 2 = r + a ∆a sin ϑ . − This quantity is negative for a = 0, and it blows up for 6 In the extremal case M 2 = Q2 + a2, so ∆ has a double Q 0. Thus, class II critical particles are allowed to → root at r0 M. The electromagnetic potential is approach r = M along the axis, except for the cases of ≡ the extremal Kerr solution (where there are no critical Qr particles moving along the axis whatsoever) and of the A = dt a sin2 ϑ dϕ . (32) − Σ − extremal Reissner-Nordström solution (where V becomes constant for αε = 1). The effective potential for axial electrogeodesic motion (as given in (4.7) in [30]) reads

qQr˜ ∆ V = + . (33) 3 This diﬀers from the equatorial case; see [21]. 2 2 2 2 4 r + a rr + a Similarly, looking at the boundary value for chargeq ˜II (36), we see that only critical particles with |q˜| > 1 can approach r = M. 2 2 3 Let us note that for a = 0, Q = M andq ˜ = sgn Q we And due to Q− dependence in (36), for |Q|≪ M, only critical get V 1. particles with |q˜|≫ ε ≫ 1 can approach r = M. ≡ 7

IV. ENERGY EXTRACTION the horizon, the following combination of the momentum components cancels up to the first order in r r : − 0 A. Application of conservation laws 2 t r 2 N p σN√grrp (r r ) , (42) − ∼ − 0 Let us now explore, in a simple setup, the possibility whereas with the opposite sign of energy extraction from black holes either by a BSW- type process occurring between particles moving along 2 t r . H N p + σN√grrp =2 E + qAt + ... (43) the axis, or by its Schnittman variant. We shall consider a scenario in which a (nearly) critical particle 1 collides contributes to the zeroth order. In contrast, for the crit- with an incoming usual particle 2 close to the horizon ical particles, or particles that behave as nearly critical radius r0, they interact, and two new particles, 3 and 4, around a desired collision radius rC, both expressions are are produced. We impose the conservation of charge, of the first order in r r , C − 0 2 t r q1 + q2 = q3 + q4 , (40) N p N√grrp (r r ) . (44) ± ∼ C − 0 and the conservation of (both components of) momentum To account consistently (at each order) for the effect of at the point of collision. The time component gives us a particle labeled i, which is not precisely critical, yet the conservation of energy nearly critical, we define a formal expansion:

2 E1 + E2 = E3 + E4 . (41) E Ei = C (r r )+ C (r r ) + ... (45) cr − (i,1) C − 0 (i,2) C − 0 In order to make the best use of the conservation of radial Now, let us sum the conservation laws for the time and momentum, we shall note that for usual particles near radial components of the momenta as follows:

2 t t r r 2 t t r r N p(1) + p(2) + N√grr p(1) + p(2) = N p(3) + p(4) + N√grr p(3) + p(4) . (46) Considering expansion of this formula in r r and using (42),(43) and (44), we reach a conclusion (analogously to C − 0 [15, 16]) that collision between a (nearly) critical particle 1 and an incoming (σ2 = 1) usual particle 2 at a radius 5 − rC close to r0 must necessarily lead to the production of an incoming usual particle, to be denoted 4, and a nearly critical particle, which we will label as 3. Then, the leading (ﬁrst) order of (46), divided by N˜H, implies

2 αE + σ α2E2 m2 = αE C˜ + σ αE C˜ m2 . (47) 1 1 1 − 1 3 − 3 3 3 − 3 − 3 q r

Here again σ = 1 corresponds to the BSW-type pro- as a new parameter6 1 − cess, whereas σ1 = +1 to the Schnittman variant (and A 2 2 2 σ = 1 to outgoing/incoming particle 3). Above we 1 αE1 + σ1 α E m , (49) 3 ≡ 1 − 1 introduced± this parameter will expressq all the dependence on the t ˜ C(3,1) properties of particle 1. Since we assumed p > 0 and C3 , (48) particle 1 cannot be massless, we can make sure that ≡ N˜H A1 > 0. and, for simplification, we chose the particle 1 precisely Because we absorbed the difference between the BSW- critical (E1 = Ecr, and hence C˜1 = 0), which means that type process and the Schnittman variant into the defi- we are using the approximation (18) for the Schnittman nition of the parameter A1, the discussion of kinematic process. regimes in the next section is the same for both. How- All the information about the spacetime coming into ever, if we consider a particular model process, a signifi- (47) is carried by the parameter α (defined in (27)). Fur- cant distinction may appear, as we discuss in IV C. thermore, if we denote the whole left-hand side of (47)

6 In order to keep the same letter for this quantity (introduced in [16] for the vacuum case, and followed e. g. by [15, 18]), we 5 See also [25] for a more detailed discussion on why it is impossible use a diﬀerent font to distinguish it from the components of the to produce outgoing usual particles near the horizon. electromagnetic potential. 8

62.4 3.99

62.2 3.98

ε εcr ε εcr 3.97 62

3.96 61.8

1 1.01 1.02 1.03 1.04 1.05 1 1.01 1.02 1.03 1.04 1.05 r r /r0 /r0

FIG. 1. Effective potential V near the horizon radius r0 = M of the extremal Kerr-Newman black hole with Q/M = 1/√5, a/M = 2/√5. The shading represents nearly critical particles; the varying height of the shaded areas illustrates the varying range of energies of particles that behave as nearly critical at a given r. The red color corresponds to “+” (C˜ > 0), blue to “−” (C˜ < 0). (See (45) and (48) for the definition of C˜.) Left: For particles withq ˜ = 250, V is decreasing, and thus both signs of C˜ are allowed. Right: Forq ˜ = 16, V is increasing, and thus nearly critical particles can exist only with C˜ < 0. It can be seen that particles that behave as nearly critical at lower values of r will get reflected by V at some higher values of r, and thus they cannot escape.

B. Kinematic regimes produced (and also in between these points due to the assumption of V being well approximated by a linear ˜ Equation (47) enables us to determine whether and function). Thus, if a particle 3 with C3 < 0 is not pro- under which circumstances particle 3 can escape and ex- duced as outgoing, it will fall into the black hole. Fur- tract energy from a black hole. Particle 4 necessarily thermore, if the effective potential V is growing at r0, falls into the black hole, which is the essence of a Pen- i.e. αε3 < 1, particle 3 will have a turning point at some rose process. Let us note that particle 3 can actually be radius greater than rC and it will not be able to escape produced in four different kinematic regimes, depending even if it is produced as outgoing. Therefore, in the IN regime the escape is impossible and for OUT it depends− on the combination of the sign of C˜3 and the sign variable on the trend of the effective potential V . (These− findings σ3. Following the classification in [15], we will refer to are summarised in Table I.) the regimes with C3 > 0 as “+”, C3 < 0as“ ”, σ3 = +1 − Second, we should determine to what ranges of pa- as “OUT” and σ3 = 1 as “IN”. − rameters of particle 3 do the different kinematic regimes We analyse the different kinematic regimes from sev- correspond. Then we can infer, whether the impossibility eral points of view. First, we should understand which of escape in the IN regime leads to some bounds on pa- combinations are compatible with particle 3 escaping rameters of the escaping− particles, and more specifically, from the vicinity of the black hole (see Figure 1 for illus- whether it does limit the efficiency of the collisional Pen- tration). For simplicity, let us assume a situation when rose process, which is defined as effective potential V for particle 3 is well approximated by a linear function around r (i.e. that αε is not very E C 3 η = 3 . (50) close to 1). E1 + E2 By definition, C˜3 > 0 implies ε3 < εcr, and hence Solving (47) to express C˜3 and σ3, we get ε3 < V (forbidden motion) at the horizon. Therefore, a particle produced with C˜ > 0 cannot fall into the 1 m2 3 C˜ = αE A + 3 , (51) black hole, and even if it is initially incoming, it must 3 3 − 2 1 A 1 reach a turning point and turn to outgoing. Moreover, 2 m3 since it must hold that ε3 > V at the radius rC where σ = sgn A sgn(A m ) . (52) 3 1 − A ≡ 1 − 3 the particle is produced, for such a particle the effective 1 potential V must be decreasing at r0 (i.e. αε3 > 1). From the second equation we see that the value of pa- Thus, we see that in kinematic regimes OUT+ and IN+ rameter A1 forms a threshold for m3. If the interaction the local escape condition is satisfied automatically. produces particle 3 with a mass above the threshold, the On the other hand, particle 3 with C˜3 < 0 will have particle must be incoming, if its mass is below the thresh- ε3 > V both at the horizon and at the point where it is old, it must be outgoing. 9

get A1 < αE1, whereas for the Schnittman variant (σ1 = TABLE I. The four kinematic regimes for production of par- 1), it is A < 2αE . Combining with (55), we get µ < E ticle 3. 1 1 1 for the BSW-type eﬀect and µ< 2E for the Schnittman σ3 = +1 σ3 = −1 1 one. Using also the “ ” condition E 6 µ, we get E < OUT+ IN+ − 3 3 ˜ E1 and E3 < 2E1, respectively. Therefore, we see that E3 C3 > 0 m3 < A1, E3 > µ m3 > A1, E3 > µ can never exceed E1 in the OUT regime for the BSW- Guaranteed to escape Guaranteed to escape type process (preventing net energy− extraction), whereas − − OUT IN for the Schnittman variant E3 > E1 is possible in this ˜ C3 6 0 m3 < A1, E3 6 µ m3 > A1, E3 6 µ regime. Escapes if αE3 >m3 Falls inside the black hole

C. Discussion of caveats Turning to parameter C˜3 note that the solution (51) satisﬁes the inequality 1. Energy feeding problem

C˜3 6 αE3 m3 . (53) − High efficiency η of the collisional Penrose process Therefore, if C˜3 > 0, we must have αε3 > 1, as we an- means by definition (50) that we can gain much more ticipated because particles with ε<εcr can be produced energy than we invest. However, despite a high value of only if effective potential V is decreasing at r0. (In gen- η the process may be “inefficient” if the invested energy eral, one can see from (29), (45) and (48) that (53) is itself needs to be high in order for the process to occur. actually the linear order of the expansion in rC r0 of We call this the “energy feeding” problem. There are the condition ε > V .) − two different sources of this problem for particles mov- Let us denote the value of E3 for which C˜3 =0 as µ: ing along the axis. One of them was already mentioned in the discussion below equation (37): for the extremal 2 1 m3 Kerr-Newman spacetime with a small value of charge µ = A1 + . (54) 2α A1 ( Q M), only highly relativistic critical particles can approach| | ≪ r = M along the axis. This does not depend This quantity again represents a threshold. If particle 3 on the nature of the particles. is produced with E3 > µ, it must have such a value of In contrast, the second source of the energy feeding charge that Ecr(q3) > E3; if E3 < µ, it must hold that problem comes into play only if we consider specifi- E (q ) < E . (Here E (q ) m ε (q ); cf. (7).) cr 3 3 cr 3 ≡ 3 cr 3 cally processes involving microscopic particles that ex- A summary of the results about the four kinematic hibit charge quantisation. For all those particles (known regimes is given in Table I. Let us note that these re- in nature) their specific charge q˜ 1. However, the spe- sults resemble those for the special case of the Reissner- cific energy of (nearly) critical| | particles ≫ is proportional Nordström solution studied in [15]. In particular, there to their specific charge (approximately) through relation is still no unconditional upper bound on the energy or (7), or, in particular, by relation (34) for Kerr-Newman mass of particle 3, in contrast with the geodesic (equa- black holes. Therefore, such microscopic particles need torial) case [16, 18]. (Such a possibility is often called to be highly relativistic (ε 1, i.e. E m) in order to the super-Penrose process.) However, the impossibility be (nearly) critical. Since≫ the elementary≫ charge is just of escape in the IN regime means that whenever par- − one order of magnitude short of the Planck mass, they ticle 3 is produced with the mass above the threshold would actually have to be extremely relativistic. This is- A 1, its energy also must be above the threshold µ (which sue was previously noted in [22], and it led the authors therefore acts as a lower bound on E3 in this case). Con- to introduce macroscopic objects acting as critical parti- versely, whenever particle 3 is produced with E3 6 µ, it cles, which would make ε 1 possible (note ε > 1 due A must also have m3 < 1, otherwise it falls into the black to (37)). ∼ hole. These requirements may not be compatible with the properties of a particular type of interaction that is responsible for producing particle 3. This is the third as- 2. Neutral mass problem pect of the kinematic regimes that needs to be examined. In IV C we consider a toy model, where this limitation Although energy extraction by processes involving crit- gets highlighted (the “neutral mass” problem). ical microscopic particles is already unfeasible due to the Before carrying out this discussion, let us further note severe energy feeding problem, there are even further re- one interesting property of the OUT regime. Condition strictions due to particle physics. Since the energy of a m < A (OUT) implies − 3 1 (nearly) critical particle is proportional to its charge, we need q > q in order to have E > E . For micro- αµ < A1 (55) 3 1 3 1 scopic| particles,| | | this means that we need to turn to inter- due to (54). From definition (49), we can derive an upper actions involving atomic nuclei. (Let us note that such bound on A . For the BSW-type process (σ = 1), we processes would actually not benefit from high E due 1 1 − CM 10 to a relatively low binding energy of nuclei, but here we satisfied, e.g. with specific reactions with q3 > q1 > 0, focus on kinematic aspects.) One of the further problems m3 < m1, which are in principle also possible. Never- was noted previously in [15]; stable nuclei have values of theless, we see that there is a strong limitation on the charge in a range that spans just two orders of magnitude. BSW-type processes with microscopic particles. 2 Thus, E3 cannot exceed E1 by more than a factor of 10 . However, if we turn to the Schnittman-type kinemat- However, the problems become much deeper, if we focus ics, the neutral mass problem is circumvented. In partic- specifically on the BSW-type mechanism. The mass of ular, for E1 m1 and σ1 = +1, parameter A1 is large, stable nuclei generally increases faster than their charge namely ≫ due to an increasing share of neutrons (hence “neutral A 2αE + ... (61) mass”). Thus, for our model process with q3 > q1 > 0 1 ≈ 1 and m3 > m1, it will also be more common than the A opposite to have7 Hence we infer m3 < 1 and particle 3 to be produced in the OUT regime. Parameter µ will be large again, but 2 this time dominated by the other term than before, i.e. q3 m3 m3 < < 2 . (56) q1 m1 m1 µ E + ... (62) ≈ 1 Now we should check whether this inequality is consis- Since we assumed q3 > q1 > 0, and hence E3 > E1, tent with particle 3 escaping. The problem again stems particle 3 will be produced in the OUT+ regime and will from the fact that critical microscopic particles are to indeed escape. be immensely relativistic. (At this point we exclude the possibility Q M, i. e. α 1, which is revisited in ≪ ≪ IV C 3.) 3. Specific charge cutoff Namely, for E1 m1 and σ1 = 1 parameter A1 (49) will be very small;≫ it can be approximated− as The problems arising from the fact that critical mi- 2 croscopic particles have to be immensely relativistic can . m1 A1 = + ... (57) be reduced for the extremal Kerr-Newman solution if we 2αE 1 consider Q very small ( Q M). However, we can- | | ≪ Given this, parameter µ (54) gets large, and it is approx- not decrease the required energy arbitrarily, because we imated as run into the other source of the energy feeding problem, which is the proportionality α Q2. Specific charges for 2 ∼ m3 all nuclei are roughly the same (of the same order), say µ E1 2 + ... (58) ≈ m1 qñucl. Because of the critical condition (34), all critical nuclei will also have values of specific energy of the same A Since certainly 1 < m1 and we assumed m1 < m3, it order. Thus, there will be a distinct transition. A will hold that m3 > 1. Thus, our nuclear reaction will Let us first consider a general value ofq ˜. Using (36) occur in the IN regime. Condition E3 > µ, which is we can define a value Q˜min of the specific charge of the Q required for the escape of particle 3 in this regime (cf. black hole Q˜ /M, such that for Q˜ sgnq ˜ < Q˜min all Table I), due to (58) means the critical particles≡ with the given value ofq ˜ would be

2 forbidden to approach r = M. Using (34) or (37) we can E3 m 3 also evaluate a corresponding speciﬁc energy εmin. We > 2 . (59) E1 m1 obtain

1 As both energies are (approximately) proportional to the 2 q˜ 3 2 ˜ 3 respective charges by the same factor, this translates to Qmin = 2 , εmin = | | 1+ q˜ . (63) s1+ q˜ 3 √2 | | the relation | | q 2 However, for critical nuclei withq ˜ 1, we can use q3 m3 nucl ≫ > 2 . (60) approximate expressions q1 m1 2 √ 3 However, this is the inequality opposite to (56). There- ˜ . 2 (˜qnucl) Qmin = 3 , εmin . (64) fore, we conclude that in our “common nuclear process”, √qñucl ≈ √2 particle 3 will be produced in the IN regime (E3 < µ) − 17 ˜ and it will fall into the black hole. Condition (60) can be Sinceq ñucl is around 5 10 , we get Qmin of order 10−6 and ε around 5 ·1011. Therefore, for extremal min · Kerr-Newman black holes with Q˜ = Q˜min, the energy feeding problem for microscopic particles is reduced by 7 Inequality (56) could be the “rule of thumb” even for macro- six orders of magnitude as compared with the extremal scopic particles, as it is harder to hold together larger amounts Reissner-Nordström case (where εcr =q ñucl). Neverthe- of charge. less, ε 1 in any case. Thus, we can never have min ≫ 11 non-relativistic critical microscopic particles approaching CA16104 GWverse. J. B. acknowledges support from the r = M along the axis of an extremal Kerr-Newman black Grant Agency of the Czech Republic, Grant No. GACRˇ hole. This is very different from the “mega-BSW” effect 19-01850S. The work of O. Z. was performed accord- described in Section V E of [21] for equatorial charged ing to the Russian Government Program of Competitive critical particles. Growth of Kazan Federal University.

Appendix A: Derivatives of W and V ± ACKNOWLEDGEMENTS Since the relation (5) among W and V± is the same for both equatorial and axial motion, we can build on F. H. is grateful for the continued support of his super- what was derived in [21] (in particular in the Appendix visor at CENTRA, Professor J. P. S. Lemos. F. H. thanks therein). Let us start with equation (34) of [21], which Funda¸c˜ao para a Ciˆencia e a Tecnologia (Portugal) for states funding his Ph.D. study at CENTRA through Grant No. ∂2W ∂V ∂V PD/BD/113477/2015 awarded in the framework of the =2 + − . (A1) Doctoral Programme IDPASC-Portugal, and for travel ∂r2 ∂r ∂r r=r0,ε=εcr r=r0 support provided through UID/FIS/00099/2013 – CEN-

TRA. F. H. is also a proud member of COST Action Taking the third radial derivative of (5) above, we get

3 3 2 2 3 ∂ W ∂ V+ ∂ V+ ∂V− ∂V+ ∂ V− ∂ V− = (ε V−)+3 +3 (ε V ) . (A2) ∂r3 − ∂r3 − ∂r2 ∂r ∂r ∂r2 − − + ∂r3

If we evaluate this relation for critical particles (ε = εcr) at the radius of the degenerate horizon (where V+ = V− = εcr), it simpliﬁes to ∂3W ∂2V ∂V ∂V ∂2V =3 + − + + − . (A3) ∂r3 ∂r2 ∂r ∂r ∂r2 r=r0,ε=εcr r=r0

Relations (A1) and (A3) have implications valid for both the equatorial and the axial motion, some of which can be further simpliﬁed in the axial case.

General case

Because V+ > V− outside the horizon, though V+ = V− on the horizon, it must hold that ∂V ∂V + > − . (A4) ∂r ∂r r=r0 r=r0

Using this with (A1), we arrive at the following two logical statements : ∂2W ∂V ∂V < 0 + > 0 & − < 0 , (A5) ∂r2 ⇐⇒ ∂r ∂r r=r0,ε=εcr r=r0 ! r=r0 !

∂2W ∂V ∂V > 0 + < 0 or − > 0 . (A6) ∂r2 ⇐⇒ ∂r ∂r r=r0,ε=εcr r=r0 ! r=r0 !

t It is easy to check that the two variants in the second statement correspond to the critical particle having p > 0 or pt < 0, respectively.8 Thus, with restriction to pt > 0, equation (21) follows. Using (A4) also with (A3), we get a statement analogous to (21) for class II critical particles:

∂V ∂2W ∂3W ∂2V ε = ε & + =0= =0 & sgn = sgn + . (A7) cr ∂r ⇒ ∂r2 ∂r3 − ∂r2 r=r0 r=r0 ! r=r0 r=r0 !

8 Note that p > 0 corresponds to ε > V+ and p < 0 to ε 6 V , t t ∂ V+/∂r| < 0 corresponds to critical particles with p > 0 and − r0 t and for critical particles their energy εcr = V+| = V | . Thus r0 r0 ∂ V− ∂r − / |r0 > 0 to those with pt < 0. 12

Axial case Using the value ofq ˜ for class II critical particles (25), we get For motion along the axis (A7) can be further reﬁned. ∂V From deﬁnition (6), we can calculate − = 2 N˜ , (A9) ∂r − r=r0,ϑ=0 r=r0,ϑ=0

∂V− ∂At = q˜ + N˜ . (A8) and if we plug the result into (A3), we arrive at (22). ∂r − ∂r r=r0,ϑ=0 r=r0,ϑ=0

[1] T. Piran, J. Shaham, J. Katz, High efficiency of the Pen- [16] T. Harada, H. Nemoto, U. Miyamoto, Upper limits of rose mechanism for particle collisions, Astrophys. J. Lett. particle emission from high-energy collision and reaction 196, L107-L108 (1975). near a maximally rotating Kerr black hole, Phys. Rev. D [2] T. Piran, J. Shaham, Upper bounds on collisional Pen- 86, 024027 (2012), 10 pages, arXiv:1205.7088 [gr-qc]. rose processes near rotating black-hole horizons, Phys. [17] M. Bejger, T. Piran, M. Abramowicz, F. H˚akanson, Col- Rev. D 16, 1615–1635 (1977). lisional Penrose Process near the Horizon of Extreme [3] M. Bañados, J. Silk, S. M. West, Kerr Black Holes as Kerr Black Holes, Phys. Rev. Lett. 109, 121101 (2012), Particle Accelerators to Arbitrarily High Energy, Phys. 5 pages, arXiv:1205.4350 [astro-ph.HE]. Rev. Lett. 103, 111102 (2009), 4 pages, arXiv:0909.0169 [18] O. B. Zaslavskii, Energetics of particle collisions near [hep-ph]. dirty rotating extremal black holes: Banados-Silk-West [4] E. Berti, V. Cardoso, L. Gualtieri, F. Pretorius, U. Sper- effect versus Penrose process, Phys. Rev. D 86, 084030 hake, Comment on “Kerr Black Holes as Particle Acceler- (2012), 14 pages, arXiv:1205.4410 [gr-qc]. ators to Arbitrarily High Energy”, Phys. Rev. Lett. 103, [19] R. M. Wald, Black hole in a uniform magnetic field, Phys. 239001 (2009), 1 page, arXiv:0911.2243 [gr-qc]. Rev. D 10, 1680-1685 (1974). [5] T. Harada, M. Kimura, Black holes as particle accelera- [20] F. J. Ernst, W. J. Wild, Kerr black holes in a magnetic tors: a brief review, Classical and Quantum Gravity 31, universe, J. Math. Phys. 17, 182-184 (1976). 243001 (2014), 17 pages, arXiv:1409.7502 [gr-qc]. [21] F. Hejda, J. Biˇcák, Kinematic restrictions on particle col- [6] O. B. Zaslavskii, Acceleration of particles by black lisions near extremal black holes: A unified picture, Phys. holes—a general explanation, Classical Quant. Grav. 28, Rev. D 95, 084055 (2017), 22 pages, arXiv:1612.04959 105010 (2011), 7 pages, arXiv:1011.0167 [gr-qc]. [gr-qc]. [7] O. B. Zaslavskii, Acceleration of particles by black holes: [22] H. Nemoto, U. Miyamoto, T. Harada, T. Kokubu, Es- Kinematic explanation, Phys. Rev. D 84, 024007 (2011), cape of superheavy and highly energetic particles pro- 6 pages, arXiv:1104.4802 [gr-qc]. duced by particle collisions near maximally charged [8] J. D. Schnittman, Revised Upper Limit to Energy Ex- black holes, Phys. Rev. D 87, 127502 (2013), 4 pages, traction from a Kerr Black Hole, Phys. Rev. Lett. 113, arXiv:1212.6701 [gr-qc]. 261102 (2014), 5 pages, arXiv:1410.6446 [astro-ph.HE]. [23] O. B. Zaslavskii, Acceleration of particles as a universal [9] E. Berti, R. Brito, V. Cardoso, Ultrahigh-Energy Debris property of rotating black holes, Phys. Rev. D 82, 083004 from the Collisional Penrose Process, Phys. Rev. Lett. (2010), 5 pages, arXiv:1007.3678 [gr-qc]. 114, 251103 (2015), 5 pages, arXiv:1410.8534 [gr-qc]. [24] A. A. Grib, Yu. V. Pavlov, On the collisions between [10] A. A. Grib, Yu. V. Pavlov, On particle collisions near particles in the vicinity of rotating black holes, JETP rotating black holes, Gravitation and Cosmology 17, 42– Letters 92, 125-129 (2010), arXiv:1004.0913 [gr-qc]. 46 (2011), arXiv:1010.2052 [gr-qc]. [25] O. B. Zaslavskii, General limitations on trajectories suit- [11] T. Harada, M. Kimura, Collision of an innermost stable able for super-Penrose process, Europhys. Lett. 111, circular orbit particle around a Kerr black hole, Phys. 50004 (2015), 4 pages, arXiv:1506.06527 [gr-qc]. Rev. D 83, 024002 (2011), 11 pages, arXiv:1010.0962 [gr- [26] K. Ogasawara, T. Harada, U. Miyamoto, High efficiency qc]. of collisional Penrose process requires heavy particle [12] I. V. Tanatarov, O. B. Zaslavskii, Bañados-Silk-West ef- production, Phys. Rev. D 93, 044054 (2016), 9 pages, fect with nongeodesic particles: Extremal horizons, Phys. arXiv:1511.00110 [gr-qc]. Rev. D 88, 064036 (2013), 14 pages, arXiv:1307.0034 [gr- [27] O. B. Zaslavskii, Maximum efficiency of the collisional qc]. Penrose process, Phys. Rev. D 94, 064048 (2016), 9 [13] I. V. Tanatarov, O. B. Zaslavskii, Bañados-Silk-West pages, arXiv:1607.00651 [gr-qc]. effect with nongeodesic particles: Nonextremal hori- [28] T. Harada, M. Kimura, Collision of two general geodesic zons, Phys. Rev. D 90, 067502 (2014), 5 pages, particles around a Kerr black hole, Phys. Rev. D 83 arXiv:1407.7463 [gr-qc]. 084041 (2011), 9 pages, arXiv:1102.3316 [gr-qc]. [14] O. B. Zaslavskii, Acceleration of particles by nonrotating [29] O. B. Zaslavskii, Near-horizon circular orbits and ex- charged black holes?, JETP Letters 92, 571-574 (2010), tremal limit for dirty rotating black holes, Phys. Rev. arXiv:1007.4598 [gr-qc]. D 92, 044017 (2015), 12 pages, arXiv:1506.00148 [gr-qc]. [15] O. B. Zaslavskii, Energy extraction from extremal [30] J. Biˇcák, Z. Stuchl´ık, V. Balek, The motion of charged charged black holes due to the Banados-Silk-West ef- particles in the field of rotating charged black holes and fect, Phys. Rev. D 86, 124039 (2012), 4 pages, naked singularities, Bulletin of the Astronomical Insti- arXiv:1207.5209 [gr-qc]. tutes of Czechoslovakia 40, 65-92 (1989).