HTML
--> --> -->In three-dimensional spacetime, there is a typical stationary black hole solution with a negative cosmological constant that was first discovered by Ba?ados-Teitelboim-Zanelli (BTZ) [20]. This black hole solution, because of its similarity and simplicity compared with the (3+1)-dimensional Kerr black hole, has recently received increasing attention. For example, the spinless particles collision around the BTZ black hole has been the subject of study in Ref. [21]. Researchers are interested in the (2+1)-dimensional BTZ black hole, as it can be a good toy model, which helps gain a deeper understanding of the same problem in Kerr spacetime. This is becausethe analytical expression is usually possible in the BTZ background [22–24], while in the Kerr spacetime the same analytical treatment for the same problem is generally very difficult. For example, the collision of fast rotating dust thin shells in the (2+1)-dimension is significantly more simple compared with the (3+1)-dimensional Kerr spacetime [22–25].
In contrast, numerous authors focus on the point particle, whose trajectory is a geodesic. However, a real particle should be an extended body with inclusion of self-interaction. Compared with the spinless particle, the orbit of a spinning test particle is no longer a geodesic, and it has been shown [26–32] that the equations of motion of spinning particles around a given spacetime background are discribed by the Mathisson-Papapetrou-Dixon (MPD) equations [33–35]. By collecting these results, the authors in Ref. [30] show that the collision center-of-mass energy could be divergent for the extremal Kerr black hole. With these motivations, our research in this study is devoted to investigate the collision of spinning particles around the BTZ black hole.
The paper is organized as follows: In Sec. 2, we introduce Mathission-Papapetrou-Dixon (MPD) equations, which describe the spinning particles' motion in curved spacetime, and apply it to the Ba?ados-Teitelboim-Zanelli (BTZ) black hole. In Sec. 3, we obtain the collision center-of-mass energy of spinning particles and find the condition for the divergence of center-of-mass energy with either of the particle possessing critical total angular momentum. Subsequently, in Sec. 4, the motion of spinning particles with critical and subcritical angular momentum near the event horizon is analyzed in detail, and it is shown that a spinning particle with subcritical total angular momentum is allowed to exist on or outside the horizon. In Sec. 5, the collision of two spinning particles with subcritical total angular momenta near the horizon are calculated, and the diverging center-of-mass energy in the critical limit is obtained. Conclusions are provided in Sec. 6. Throughout the paper, we adopt the convention that the speed of light
$ {\rm{d}}s^2 = -g(r){\rm{d}}t^2+{{\rm{d}}r^2\over g(r)}+r^2\left({\rm{d}}\phi-{r_{+}r_{-}\over lr^2}{\rm{d}}t\right)^2, $ ![]() | (1) |
$ g(r) ={(r^2-r_{+}^2)(r^2-r_{-}^2)\over l^2r^2}, $ ![]() | (2) |
$ M = {r_{+}^2+r_{-}^2\over 8Gl^2}, $ ![]() | (3) |
$ J = {r_{+}r_{-}\over 4Gl}, $ ![]() | (4) |
Under the given BTZ spacetime, the spinning particle's motion can be described by MPD equations [28, 30]
$ \frac{D}{D\tau }P^{a} = -\frac{1}{2}R_{bcd}^{a} v ^{b}S^{cd}, $ ![]() | (5) |
$\frac{D}{D\tau }S^{ab} = P^{a} v^{b}-P^{b} v^{a}. $ ![]() | (6) |
To obtain the detailed relation between
$ S^{ab}P_{b} = 0, $ ![]() | (7) |
$ P^a v_a = -m , $ ![]() | (8) |
$ mv^{a}-P^{a} = \frac{S^{ab}R_{bcde}P^{c}S^{de}}{2\left(m^{2}+\dfrac{1}{4}R_{bcde}S^{bc}S^{de}\right)}. $ ![]() | (9) |
There are two Killing vector fields
$ \begin{align} \xi_a& = -\sqrt{g(r)}e_{a}^{(0)}-{r_{+}r_{-}\over lr}e_{a}^{(2)},\\ \phi_a& = r e_{a}^{(2)} , \end{align} $ ![]() | (10) |
$ \begin{align} e_{a}^{(0)}& = \sqrt{g(r)}({\rm{d}}t)_a, \\ e_{a}^{(1)}& ={1\over \sqrt{g(r)}}({\rm{d}}r)_a, \\ e_{a}^{(2)}& =r\left(({\rm{d}}\phi)_a-{r_{+}r_{-}\over lr^2}({\rm{d}}t)_a\right). \end{align} $ ![]() | (11) |
$ Q_{\xi } = P^{a}\xi _{a}-\frac{1}{2}S^{ab}\triangledown_{b}\xi_{a}. $ ![]() | (12) |
$ \begin{align} E_m& =-u^a\xi _a+\frac{1}{2m}S^{ab}\nabla _b\xi _a,\\ J_m& =u^a\phi _a-\frac{1}{2m}S^{ab}\nabla _b\phi _a. \end{align} $ ![]() | (13) |
$ s^2: = \frac{1}{2m^2}S^{ab}S_{ab}, $ ![]() | (14) |
$ S^{(a)(b)} = -m\varepsilon ^{(a)(b)}_{\; \; \; \; \; \; (c)}u^{(c)}s, $ ![]() | (15) |
From Eq. (15), the non-zero components of the spin tensor can be expressed in terms of
$ \begin{align} S^{(0)(1)}& =-msu^{(2)},\\ S^{(0)(2)}& =msu^{(1)},\\ S^{(1)(2)}& =-msu^{(0)}. \end{align} $ ![]() | (16) |
$ E_m = \sqrt{g(r)}u^{(0)}+\left({r_{+}r_{-}\over lr}+{rs\over l^2}\right)u^{(2)}, $ ![]() | (17) |
$ J_m = s\sqrt{g(r)}u^{(0)}+\left({r_{+}r_{-}s\over lr}+r\right)u^{(2)}. $ ![]() | (18) |
$ u^{(0)} = \frac{l \left(l E_m \left(l r^2+r_- r_+ s\right)-J_m \left(l r_- r_++r^2 s\right)\right)}{r \sqrt{\left(r^2-r_-^2\right) \left(r^2-r_+^2\right)} (l^2-s^2)}, $ ![]() | (19) |
$ u^{(2)} = \frac{J_m-E_m s}{r-\dfrac{r s^2}{l^2}}. $ ![]() | (20) |
$ (u^{(1)})^2 = (u^{(0)})^{2}-(u^{(2)})^2-m^2. $ ![]() | (21) |
$ p^t(r) ={{\rm{d}}t\over{\rm{d}}\tau} = {W(r)\over g(r)}, $ ![]() | (22) |
$ p^r(r) = {{\rm{d}}r\over{\rm{d}}\tau} = \rho\sqrt{Y(r)}, $ ![]() | (23) |
$ p^\phi(r) ={{\rm{d}}\phi\over{\rm{d}}\tau} = {r_{+}r_{-}W(r)\over lg(r)r^2}+{l^2(J_m-E_ms)\over r^2(l^2-s^2)}, $ ![]() | (24) |
$ W(r) = \frac{E_{{m}} l \left(l r^2+r_{-} r_{+} s\right)-J_{{m}} \left(l r_{-} r_{+}+r^2 s\right)}{r^2 \left(l^2-s^2\right)}, $ ![]() | (25) |
$ Y(r) = W^2(r)-\left(m^2+\left({J_{{m}}-E_{{m}}s\over r(1-{s^2\over l^2})}\right)^2\right)g(r), $ ![]() | (26) |
We define the critical angular momentum as
$ J_c\equiv{E_ml(lr_{+}+r_{-}s)\over lr_{-}+r_{+}s}, $ ![]() | (27) |
$ W_i(r_{+}) = 0, $ ![]() | (28) |
The timelike constraint of Eq. (22) indicates
$ \begin{split}E^2_{\rm cm}\equiv &-(p^\mu_1(r)+p^\mu_2(r))(p_{1\mu}(r)+p_{2\mu}(r)) \\ =& \;m_1^2+m_2^2 +{W_1(r)W_2(r)-\sqrt{Y_1(r)Y_2(r)}\over g(r)}\\ &-2{l^4(J_{m1}-E_{m1}s_1)(J_{m2}-E_{m1}s_2)\over r^2(l^2-s_1^2)(l^2-s_2^2)}, \end{split} $ ![]() | (29) |
We find that the third term of Eq. (29) is a
$ \mathop {\lim }\limits_{r \to {r_ + }} 2\frac{{{W_1}(r){W_2}(r) - \sqrt {{Y_1}(r){Y_2}(r)} }}{{g(r)}} = \frac{{{W_2}({r_ + })}}{{{W_1}({r_ + })}}{Z_1} + \frac{{{W_1}({r_ + })}}{{{W_2}({r_ + })}}{Z_2},$ ![]() | (30) |
$ Z_i = \left(m_i^2+\left({J_{mi}-E_{mi}s_i\over r\left(1-\dfrac{s_i^2} {l^2}\right)}\right)^2\right)>0. $ ![]() | (31) |
$ {W_2(r_{+})\over W_1(r_{+})} = {W'_2(r_{+})\over W'_1(r_{+})} = \frac{E_{m1}(l r_-+ r_+ s_2)}{E_{m2} (l r_-+ r_+ s_1)}, $ ![]() | (32) |
$ s_1 = s_c = -{lr_{-}\over r_{+}}. $ ![]() | (33) |
First, we note that for the spinless case [21], a particle with critical total angular momentum
2
4.1.Timelike constraint of $ p^t(r) $![]()
![]()
The first subsection is devoted to the timelike constraint of $ p^t(r) = {W(r)\over g(r)}\geqslant 0, $ ![]() | (34) |
$ W(r) = \frac{l \left(l E_m \left(l r^2+r_- r_+ s\right)-J_m \left(l r_- r_++r^2 s\right)\right)}{r (l^2-s^2)}\geqslant 0. $ ![]() | (35) |
$ J_m\leqslant E_m l = J_c, $ ![]() | (36) |
2
4.2.Radial motion of particle: $ Y(r) $![]()
![]()
Now we come to the radial motion of the particle, starting with the expression of $ {1\over 2}{p^r(r)}^2+V(r) = 0, $ ![]() | (37) |
For a massive particle with
$ \mathop {\lim }\limits_{r \to \infty } Y(r) = - {m^2} \times \infty < 0, $ ![]() | (38) |
2
4.3.Motion of a particle with subcritical total angular momentum
In the last subsection, we already know that particles with critical total angular momentum cannot exist outside the event horizon. Thus, we consider a particle with subcritical total angular momentum $ J_m\equiv J_c-{\tilde{\delta}} = {E_ml(lr_{+}+r_{-}s)\over lr_{-}+r_{+}s}-{\tilde{\delta}}, $ ![]() | (39) |
$ \begin{split}Y(r) =Y_c(r)+\frac{{\tilde{\delta}}(-2E_ml(r^2-r_{+}^2)(l^2-s^2)(lr_{+}+r_{-}s)-(lr_{-}+r_{+}s)(l^2(-r^2+r_{-}^2+r_{+}^2)+2lr_{-}r_{+}s+r^2s^2){\tilde{\delta}})}{r^2(l^2-s^2)^2(lr_{-}+r_{+}s)}, \end{split} $ ![]() | (40) |
$ Y_c(r) = -{(r^2-r_{+}^2)\over r^2}\left({E_m^2l^2(r_{+}^2-r_{-}^2)\over (lr_{-}+r_{+}s)^2}+{m^2(r^2-r_{-})^2\over l^2}\right). $ ![]() | (41) |
$ Y(r_{+}) = \left({(lr_{-}+r_{+}s){\tilde{\delta}}\over r_{+}(l^2-s^2)}\right)^2>0. $ ![]() | (42) |
$ Y_c'(r_+) = \frac{2 \left(r_-^2-r_+^2\right) \left(l^4 E_m^2+m^2 \left(l r_-+r_+ s\right){}^2\right)}{l^2 r_+ \left(l r_-+r_+ s\right){}^2} $ ![]() | (43) |
Then, with subcritical total angular momentum,
$ Y'(r) = Y'_c(r)-{2l{\tilde{\delta}}(-2E_mr_{+}^2(l^2-s^2)(lr_{+}+r_{-}s)+(lr_{-}+r_{+}s)(l(r_{-}^2+r_{+}^2)+2r_{-}r_{+}s){\tilde{\delta}})\over r^3(lr_{-}+r_{+}s)(l^2-s^2)^2}.\\ $ ![]() | (44) |
$ Y'(r_{+}) = D_2{\tilde{\delta}}^2+D_1{\tilde{\delta}}+D_0, $ ![]() | (45) |
$ D_2 = -\frac{2 l \left(l \left(r_-^2+r_+^2\right)+2 r_- r_+ s\right)}{r_+^3 \left(l^2-s^2\right)^2}, $ ![]() | (46) |
$ D_1 = \frac{4 l E_m \left(l r_++r_- s\right)}{r_+ \left(l^2-s^2\right) \left(l r_-+r_+ s\right)}, $ ![]() | (47) |
$ D_0 =\frac{2 \left(r_-^2-r_+^2\right) \left(l^4 E_m^2+m^2 \left(l r_-+r_+ s\right){}^2\right)}{l^2 r_+ \left(l r_-+r_+ s\right){}^2}. $ ![]() | (48) |
$ E_{\max} = \frac{m \sqrt{r_+^2-r_-^2} \sqrt{l r_-^2+l r_+^2+2 r_+ r_- s}}{l^{3/2} r_-}, $ ![]() | (49) |
$ \begin{split} {\tilde{\delta}}_{\rm L}& =\frac{l^3 E_m r_+^2 (l^2-s^2) \left(l r_++r_- s\right)-\sqrt{\Delta}}{l^3 \left(l r_-+r_+ s\right) \left(l \left(r_-^2+r_+^2\right)+2 r_- r_+ s\right)},\\ {\tilde{\delta}}_{\rm R}& =\frac{l^3 E_m r_+^2 (l^2-s^2) \left(l r_++r_- s\right)+\sqrt{\Delta}}{l^3 \left(l r_-+r_+ s\right) \left(l \left(r_-^2+r_+^2\right)+2 r_- r_+ s\right)}, \\ \Delta =& l^3 r_+^2 (l^2-s^2)^2 \left(l r_-+r_+ s\right){}^2 \\ &\times\left(l^3 E_m^2 r_-^2+m^2 \left(r_-^2-r_+^2\right) \left(l \left(r_-^2+r_+^2\right)+2 r_- r_+ s\right)\right). \end{split} $ ![]() | (50) |
For the spinless particle, the infinite center-of-mass energy collision occurs at the extreme point of
$ r_m = r_{+}\left(1+{Y'(r_{+})l^2\over 2r_{+}m^2}\right)^{1\over 4}. $ ![]() | (51) |
$ r_m\geqslant r_{+}, $ ![]() | (52) |
After applying the extremal condition
$ \begin{split} {\tilde{\delta}}_{\rm L}& =\frac{1}{2} E_m \left(l-s-\sqrt{(l-s)^2}\right),\\ {\tilde{\delta}}_{\rm R}& =\frac{1}{2} E_m \left(l-s+\sqrt{(l-s)^2}\right). \end{split}$ ![]() | (53) |
Since
In Fig. 1,

$ \lim_{{\tilde{\delta}}_1\rightarrow 0}E_{\rm cm}^2(r_{m}) = m_1^2+m_2^2+Q-{2l^4(E_{m1}(l-s_1))J_{m2}\over r_+^2(l^2-s_1^2)(l^2-s_2^2)}, $ ![]() | (54) |
$ Q \equiv \mathop {\lim }\limits_{r \to {r_m}} 2\frac{{{W_1}({r_m}){W_2}({r_m}) - \sqrt {{Y_1}({r_m}){Y_2}({r_m})} }}{{g({r_m})}}.$ ![]() | (55) |
$ \begin{split} Q& = \lim_{{\tilde{\delta}}_1\rightarrow 0}{W_2(r_m)\over \dot{g}(r_m)}\left(2\dot{W}_1(r_m)-{\dot{Y}_1(r_m)\over \sqrt{Y_1(r_m)}}\right)\\ & = \lim_{{\tilde{\delta}}_1\rightarrow 0}{W_2(r_m)\over \dot{g}(r_m)}\left(2\dot{W}_1(r_m)-2\sqrt{\dot{W}_1(r_m)\over(l-s_1)}\right), \end{split}$ ![]() | (56) |
$ \lim_{{\tilde{\delta}}_1\rightarrow 0}\dot{W}_1(r_m) = \frac{E_{m1}^2 l^4+m_1^2 r_{+}^2 (l+s_1)^2}{m_1^2 r_{+}^2 (l-s_1) (l+s_1)^2}, $ ![]() | (57) |
$ \lim_{{\tilde{\delta}}_1\rightarrow 0}\dot{g}(r_m) = \frac{2 {\tilde{\delta}}_1 E_{m1}^2 l^4}{m_1^4 r_{+}^2 \left(l^2-s_1^2\right)^2}. $ ![]() | (58) |
$ \begin{split} Q& = \lim_{{\tilde{\delta}}_1\rightarrow 0}{W_2(r_m)\over \dot{g}(r_m)}\left(2\dot{W}_1(r_m)-2\sqrt{\dot{W}_1(r_m)\over(l-s_1)}\right) \\&= 2k\lim_{{\tilde{\delta}}_1\rightarrow 0}{W_2(r_m)\over \dot{g}(r_m)}\dot{W}_1(r_m)\\ & = k\lim_{{\tilde{\delta}}_1\rightarrow 0}W_2(r_m)\frac{m_1^2 (l-s_1) \left(E_{m1}^2 l^4+m_1^2 r_{+}^2 (l+s_1)^2\right)}{ {\tilde{\delta}}_1 E_{m1}^2 l^4}. \end{split} $ ![]() | (59) |
$\begin{split} k = 1-\sqrt{\dot{W}_1(r_m)\over(l-s_1)}/\dot{W}_1(r_m) = 1-\sqrt{\frac{m_1^2 r_{+}^2 (l+s_1)^2}{E_{m1}^2 l^4+m_1^2 r_{+}^2 (l+s_1)^2}}>0. \end{split}$ ![]() | (60) |
Notably, there are still many important issues that need to be investigated in the future. For example, inspired by the BSW mechanism, people found that the efficiency of extracting energy from a rotating black hole, which is usually called the Penrose process, can be significantly improved, especially for spinning particles [12, 13, 19, 36]. Therefore, with the BSW mechanism for spinning particles, studying the corresponding Penrose process becomes possible. We hope to address this issue in the near future.