This paper was converted on www.awesomepapers.org from LaTeX by an anonymous user.
Want to know more? Visit the Converter page.

Genuine tripartite nonlocality and entanglement in curved spacetime

Shu-Min Wu1, Hao-Sheng Zeng2111Email: [email protected] 1 Department of Physics, Liaoning Normal University, Dalian 116029, China
2 Department of Physics, Hunan Normal University, Changsha 410081, China
Abstract

We study the genuine tripartite nonlocality (GTN) and the genuine tripartite entanglement (GTE) of Dirac fields in the background of a Schwarzschild black hole. We find that the Hawking radiation degrades both the physically accessible GTN and the physically accessible GTE. The former suffers from “sudden death” at some critical Hawking temperature, and the latter approaches to the nonzero asymptotic value in the limit of infinite Hawking temperature. We also find that the Hawking effect cannot generate the physically inaccessible GTN, but can generate the physically inaccessible GTE for fermion fields in curved spacetime. These results show that on the one hand the GTN cannot pass through the event horizon of black hole, but the GTE do can, and on the other hand the surviving physically accessible GTE and the generated physically inaccessible GTE for fermions in curved spacetime are all not nonlocal. Some monogamy relations between the physically accessible GTE and the physically inaccessible GTE are found.

pacs:
04.70.Dy, 03.65.Ud,04.62.+v

I Introduction

The concept of quantum nonlocality was firstly proposed by Einstein, Podolsky and Rosen in 1935 in their famous EPR paradox EPR1935 . Afterwards, Bell established the so-called Bell inequality for conveniently judging the existence of quantum nonlocality L1 . Two quantum systems that admit Bell inequality cannot be regarded as nonlocal, even if they are far apart in space. Contrarily, when the Bell inequality is violated, we say that the two quantum systems are inseparable or have the quantum nonlocality. Quantum nonlocality is a kind of quantum behaviors, which denies the local hidden variable (LHV) model L2 ; L3 ; L4 ; L5 ; L6 ; L7 ; L8 . Quantum nonlocality is a very important quantum resource and has been applied in different fields, such as device-independent quantum computation, communication complexity, quantum cryptography and randomness generation L9 ; L10 ; L11 ; L12 ; L13 ; L14 . In the case of tripartite quantum systems, a more valuable kind of resource is involved which is called GTN. Svetlichny introduced a kind of GTN and found the so-called Svetlichny inequality to detect its existence L15 . Quantum nonlocality and entanglement are inextricably linked. Quantum nonlocality originates from quantum entanglement, but quantum entanglement does not imply quantum nonlocality.

The combination of quantum information science, relativity theory and quantum field theory gives us a deeper understanding of quantum mechanics. It is necessary to understand quantum effects in the relativistic framework, because the world is essentially noninertial or/and curved. Recently, quantum entanglement under relativistic settings received much attention L16 ; L17 ; L18 ; L19 ; L20 ; L21 ; L22 ; L23 ; L24 ; L25 ; L26 ; L27 ; zhx2 ; zhx3 , including the bipartite entanglement for boson fields L16 and fermi fields L17 in the noninertial frames, and the bipartite entanglement in curved spacetime L18 ; L19 ; L20 ; L21 ; L22 ; L23 ; L24 ; L25 ; L26 ; L27 ; zhx2 . Bipartite entanglement under the joint influence of both environmental noise and Unruh effect L28 ; L29 was also studied. Besides bipartite systems, tripartite entanglement in the relativistic framework was also investigated L30 ; L31 ; L32 ; L34 ; L35 ; L36 ; L37 . Note that, in these works, the measure of GTE is based on the concept of logarithmic negativity L38 ; L39 , and the measure of GTE based on concurrence L44 ; L45 was not used in Schwarzschild spacetime.

In this work, we study the properties of GTN and GTE of Dirac fields in the background of a Schwarzschild black hole. Assume that Alice, Bob and Charlie initially share a Greenberger-Horne-Zeilinger-like state. Alice is a Kruskal observer who stays stationarily at an asymptotically flat region, while Bob and Charlie are Schwarzschild observers who hover near the event horizon of the black hole. In addition, there are two imagined observers, anti-Bob and anti-Charlie, in the interior of the event horizon. The Hawking effect would make the information tunnelling from the exterior to the interior of the event horizon, such that a correlated state that involves the above five observers is established. Since the interior of the black hole is causally disconnected from the exterior and observers cannot access to the interior of the event horizon, we thus call the information (including GTN and GTE) that distributes completely in the outside of the event horizon the physically accessible. Otherwise, it is called the physically inaccessible. The main end of this paper is to study the influence of Hawking effect on the physically accessible GTN and GTE, the production of the physically inaccessible GTN and GTE, and the monogamy relationship between the physically accessible and inaccessible information.

The paper is organized as follows. In Sec. II, we briefly recall the measures of GTN and GTE for tripartite quantum systems. In Sec. III, we introduce the quantization of Dirac fields in the background of Schwarzschild black hole. Sec. IV is the main contribution of our work, where the decay of the physically accessible information and the production of the physically inaccessible information, as well as the monogamy relationship between them, under the influence of Hawking effect, are studied. Finally, the last section is devoted to the conclusion.

II Measures of GTN and GTE

II.1 Measure of GTN

Firstly, we briefly review the concept about GTN. Nonlocality in tripartite systems has been considered as the manifestation of genuine tripartite correlations. Generally, local tripartite correlations shared by Alice, Bob and Charlie can be written as

P(a,b,c|x,y,z)=λpλPλ(a|x)Pλ(b|y)Pλ(c|z),\displaystyle P(a,b,c|x,y,z)=\sum_{\lambda}p_{\lambda}P_{\lambda}(a|x)P_{\lambda}(b|y)P_{\lambda}(c|z), (1)

where {λ;pλ}\{\lambda;p_{\lambda}\} is the probability distribution of some hidden variable that controls the outputs a,b,c{0,1}a,b,c\in\{0,1\} of the local measurements performed by Alice, Bob and Charlie on the two-valued variables x,y,z{0,1}x,y,z\in\{0,1\}, and Pλ(a|x)P_{\lambda}(a|x) is the conditional probability for obtaining the output aa when the measurement setting is xx and λ\lambda. If the tripartite correlation cannot be written as the form of Eq.(1), then the tripartite system is said to have GTN.

In 1987, Svetlichny proposed a hybrid local-nonlocal form of correlation to measure GTN L15 . A tripartite correlation under the definition of Svetlichny is called to be local if it admits the following local LHV model

P(a,b,c|x,y,z)\displaystyle P(a,b,c|x,y,z) =\displaystyle= λpλPλ(a|x)Pλ(b,c|y,z)+μpμPμ(b|y)Pu(a,c|x,z)\displaystyle\sum_{\lambda}p_{\lambda}P_{\lambda}(a|x)P_{\lambda}(b,c|y,z)+\sum_{\mu}p_{\mu}P_{\mu}(b|y)P_{u}(a,c|x,z)
+\displaystyle+ νpνPλ(c|z)Pν(a,b|x,y),\displaystyle\sum_{\nu}p_{\nu}P_{\lambda}(c|z)P_{\nu}(a,b|x,y),

where λpλ+μpμ+νpν=1\sum_{\lambda}p_{\lambda}+\sum_{\mu}p_{\mu}+\sum_{\nu}p_{\nu}=1. This kind of correlation is regarded as Svetlichny local, otherwise it is Svetlichny nonlocal L40 ; L41 ; L42 .

Assuming that Alice, Bob and Charlie share a state ρ\rho of some three-qubit system. Alice performs measurements on the observable 𝑨=𝒂𝝈\bm{A}=\bm{a}\cdot\bm{\sigma} and 𝑨=𝒂𝝈\bm{A^{\prime}}=\bm{a^{\prime}}\cdot\bm{\sigma}, where 𝒂=(a1,a2,a3)\bm{a}=(a_{1},a_{2},a_{3}), 𝒂=(a1,a2,a3)3\bm{a^{\prime}}=(a^{\prime}_{1},a^{\prime}_{2},a^{\prime}_{3})\in\mathbb{R}^{3} are any three-dimensional unit vectors, and 𝝈=(σ1,σ2,σ3)\bm{\sigma}=(\sigma_{1},\sigma_{2},\sigma_{3}) is the vector of Pauli matrices. Similar measurements are performed by Bob and Charlie, respectively, on the observable 𝑩\bm{B}, 𝑩\bm{B^{\prime}} and observable 𝑪\bm{C}, 𝑪\bm{C^{\prime}}. Introducing Svetlichny operator

S=(𝑨+𝑨)(𝑩𝑪+𝑩𝑪)+(𝑨𝑨)(𝑩𝑪𝑩𝑪),S=(\bm{A}+\bm{A^{\prime}})\otimes(\bm{B}\otimes\bm{C^{\prime}}+\bm{B^{\prime}}\otimes\bm{C})+(\bm{A}-\bm{A^{\prime}})\otimes(\bm{B}\otimes\bm{C}-\bm{B^{\prime}}\otimes\bm{C^{\prime}}),

then the Svetlichny inequality

tr(Sρ)4\displaystyle{\rm tr}(S\rho)\leq 4 (3)

fulfills for any state ρ\rho that admits LHV model, where SS is taken over all possible Svetlichny operators, i.e., the measurements are taken over all the directions of three-dimensional space. Equivalently, if a tripartite state violates the Svetlichny inequality for some Svetlichny operator, then this state is genuinely nonlocal. For convenience, we introduce the Svetlichny value, i.e., the maximum expectation value of Svetlichny operatorL7 ; L8 ; L43

S(ρ)=maxStr(Sρ)\displaystyle S(\rho)=\max_{S}{\rm tr}(S\rho) (4)

to detect the GTN of a given three-qubit state ρ\rho. For the three-qubit XX state with density matrix

ρX=(n1000000c10n20000c2000n300c300000n4c4000000c4m400000c300m3000c20000m20c1000000m1)\displaystyle\rho_{X}=\left(\!\!\begin{array}[]{cccccccc}n_{1}&0&0&0&0&0&0&c_{1}\\ 0&n_{2}&0&0&0&0&c_{2}&0\\ 0&0&n_{3}&0&0&c_{3}&0&0\\ 0&0&0&n_{4}&c_{4}&0&0&0\\ 0&0&0&c^{*}_{4}&m_{4}&0&0&0\\ 0&0&c^{*}_{3}&0&0&m_{3}&0&0\\ 0&c^{*}_{2}&0&0&0&0&m_{2}&0\\ c^{*}_{1}&0&0&0&0&0&0&m_{1}\end{array}\!\!\right) (13)

in the orthogonal basis {|0,0,0,|0,0,1,,|1,1,1}\{|0,0,0\rangle,|0,0,1\rangle,...,|1,1,1\rangle\}, the Svetlichny value can be simply given byL8

S(ρX)=max{82|ci|,4|N|},\displaystyle S(\rho_{X})=\max\{8\sqrt{2}|c_{i}|,4|N|\}, (14)

where N=n1n2n3+n4m4+m3+m2m1N=n_{1}-n_{2}-n_{3}+n_{4}-m_{4}+m_{3}+m_{2}-m_{1}.

II.2 Measure of GTE

GTE can be defined by its opposite of biseparability. We call a tripartite pure state |Ψ|\Psi\rangle is biseparable, if it has a bipartition of the form |Ψ=|ΨA|ΨB|\Psi\rangle=|\Psi_{A}\rangle\otimes|\Psi_{B}\rangle, where |ΨA|\Psi_{A}\rangle and |ΨB|\Psi_{B}\rangle are the monomeric or bipartite pure states. Obviously, a biseparable pure state has at least one pure marginal. If the tripartite state |Ψ|\Psi\rangle is not biseparable with respect to any of its bipartition, then it is called GTE. One can then define the so-called genuine tripartite concurrence C(|Ψ)=minχiχ2[1Tr(ρAχi2)]C(|\Psi\rangle)=\min_{\chi_{i}\in\chi}\sqrt{2[1-\text{Tr}(\rho^{2}_{A_{\chi_{i}}})]} for describing the degrees of the GTE for the pure state |Ψ|\Psi\rangle, where χ={Ai|Bi}\chi=\{A_{i}|B_{i}\} denotes the set of all possible bipartitions of the tripartite system, and ρAχi\rho_{A_{\chi_{i}}} is the reduced density operator of system AA corresponding to the bipartition χi\chi_{i} L44 ; L33 . The GTE for a mixed state ρ\rho can be obtained by a convex roof construction

C(ρ)=inf{pi,|Ψi}ipiC(|Ψi),\displaystyle C(\rho)=\inf_{\{p_{i},|\Psi_{i}\rangle\}}\sum_{i}p_{i}C(|\Psi_{i}\rangle), (15)

where the infimum takes over all possible decompositions ρ=ipi|ΨiΨi|\rho=\sum_{i}p_{i}|\Psi_{i}\rangle\langle\Psi_{i}|. For three-qubit XX states given by Eq.(13), the GTE is given by

C(ρX)=2max{0,|ci|νi},i=1,,4,\displaystyle C(\rho_{X})=2\max\{0,|c_{i}|-\nu_{i}\},i=1,\ldots,4, (16)

where νi=ji4njmj\nu_{i}=\sum_{j\neq i}^{4}\sqrt{n_{j}m_{j}} L45 .

III Quantization of Dirac fields in a Schwarzschild black hole

The Dirac equation under a general background spacetime can be written as L46

[γaea(μ+Γμ)μ]Φ=0,[\gamma^{a}e_{a}{}^{\mu}(\partial_{\mu}+\Gamma_{\mu})]\Phi=0, (17)

where γa\gamma^{a} are the Dirac matrices, the four-vectors eaμe_{a}{}^{\mu} is the inverse of the tetrad eaμe^{a}{}_{\mu}, and Γμ=18[γa,γb]eaebν;μν\Gamma_{\mu}=\frac{1}{8}[\gamma^{a},\gamma^{b}]e_{a}{}^{\nu}e_{b\nu;\mu} are the spin connection coefficients. The metric of the Schwarzschild black hole can be written as

ds2\displaystyle ds^{2} =\displaystyle= (12Mr)dt2+(12Mr)1dr2\displaystyle-(1-\frac{2M}{r})dt^{2}+(1-\frac{2M}{r})^{-1}dr^{2} (18)
+r2(dθ2+sin2θdφ2),\displaystyle+r^{2}(d\theta^{2}+\sin^{2}\theta d\varphi^{2}),

where MM denotes the mass of the black hole. For simplicity, we take ,G,c\hbar,G,c and kk as unity in this paper.

Solving the Dirac equation of Eq.(17) near the event horizon of black hole, a set of positive frequency outgoing solutions for the inside and outside regions of the event horizon can be obtained L19 ; L23 ; L47

Φk,in+ϕ(r)eiωu,\displaystyle\Phi^{+}_{{k},{\rm in}}\sim\phi(r)e^{i\omega u}, (19)
Φk,out+ϕ(r)eiωu,\displaystyle\Phi^{+}_{{k},{\rm out}}\sim\phi(r)e^{-i\omega u}, (20)

where ϕ(r)\phi(r) denotes four-component Dirac spinor, ω\omega is a monochromatic frequency, and u=tru=t-r_{*} with the tortoise coordinate r=r+2Mlnr2M2Mr_{*}=r+2M\ln\frac{r-2M}{2M}. The modes Φk,in+\Phi^{+}_{{k},{\rm in}} and Φk,out+\Phi^{+}_{{k},{\rm out}} are usually called Schwarzschild modes. According to future-directed timelike Killing vector under each region, particles and antiparticles will be classified.

Making an analytic continuation for Eqs.(19) and (20) according to suggestion of Damour and Ruffini, we obtain a complete basis of positive energy modes, i.e., the Kruskal modes L48 . Then, we can use Schwarzschild mode and Kruskal mode to expand the Dirac field, respectively, leading to the Bogoliubov transformations between annihilation operator and creation operator under the Schwarzschild and Kruskal coordinates L49 ; zhx1 . After properly normalizing the state vector, the vacuum state and excited state of the Kruskal particle in the single-mode approximation are given by

|0K\displaystyle|0\rangle_{K} =\displaystyle= (eωT+1)12|0I|0II+(eωT+1)12|1I|1II,\displaystyle(e^{-\frac{\omega}{T}}+1)^{-\frac{1}{2}}|0\rangle_{I}|0\rangle_{II}+(e^{\frac{\omega}{T}}+1)^{-\frac{1}{2}}|1\rangle_{I}|1\rangle_{II},
|1K\displaystyle|1\rangle_{K} =\displaystyle= |1I|0II,\displaystyle|1\rangle_{I}|0\rangle_{II}, (21)

where T=18πMT=\frac{1}{8\pi M} is the Hawking temperature, {|nI}\{|n\rangle_{I}\} and {|nII}\{|n\rangle_{II}\} are the number states for the particle outside the region and the antiparticle inside the region of the event horizon, respectively.

IV Evolution of the GTN and GTE in Schwarzschild black hole

Consider a Greenberger-Horne-Zeilinger-like (GHZ-like) state of the Dirac fields shared by Alice, Bob and Charlie in the asymptotically flat region

|ΨABC=α|0A0B0C+1α2|1A1B1C,\displaystyle|\Psi\rangle_{ABC}=\alpha|0_{A}0_{B}0_{C}\rangle+\sqrt{1-\alpha^{2}}|1_{A}1_{B}1_{C}\rangle, (22)

where α\alpha is the state parameter that runs from 0 to 11. Now, we assume that Alice still stays at an asymptotically flat region, while Bob and Charlie hover outside the event horizon of the black hole, then we can rewrite Eq.(22) in terms of Kruskal modes for Alice and Schwarzschild modes for Bob and Charlie as

ΨABIBIICICII\displaystyle\Psi_{AB_{I}B_{II}C_{I}C_{II}} =\displaystyle= α(eωT+1)1|0A|1BI|1BII|1CI|1CII\displaystyle\alpha(e^{\frac{\omega}{T}}+1)^{-1}|0\rangle_{A}|1\rangle_{B_{I}}|1\rangle_{B_{II}}|1\rangle_{C_{I}}|1\rangle_{C_{II}} (23)
+\displaystyle+ 1α2|1A|1BI|0BII|1CI|0CII\displaystyle\sqrt{1-\alpha^{2}}|1\rangle_{A}|1\rangle_{B_{I}}|0\rangle_{B_{II}}|1\rangle_{C_{I}}|0\rangle_{C_{II}}
+\displaystyle+ α(eωT+eωT+2)12(|0A|0BI|0BII|1CI|1CII\displaystyle\alpha(e^{\frac{\omega}{T}}+e^{-\frac{\omega}{T}}+2)^{-\frac{1}{2}}(|0\rangle_{A}|0\rangle_{B_{I}}|0\rangle_{B_{II}}|1\rangle_{C_{I}}|1\rangle_{C_{II}}
+\displaystyle+ |0A|1BI|1BII|0CI|0CII)\displaystyle|0\rangle_{A}|1\rangle_{B_{I}}|1\rangle_{B_{II}}|0\rangle_{C_{I}}|0\rangle_{C_{II}})
+\displaystyle+ α(eωT+1)1|0A|0BI|0BII|0CI|0CII.\displaystyle\alpha(e^{-\frac{\omega}{T}}+1)^{-1}|0\rangle_{A}|0\rangle_{B_{I}}|0\rangle_{B_{II}}|0\rangle_{C_{I}}|0\rangle_{C_{II}}.

Generally, this is a 5-partite entangled state consisted by subsystems: subsystem AA observed by Alice, subsystems BIB_{I} and CIC_{I} observed by Bob and Charlie outside the event horizon of black hole, and subsystems BIIB_{II} and CIIC_{II} observed by anti-Bob and anti-Charlie inside the event horizon, respectively.

Since the interior region of black hole is causally disconnected from the exterior region, and Alice, Bob and Charlie cannot access the modes inside the event horizon, we thus call the modes BIB_{I} and CIC_{I} outside the event horizon the accessible modes, and the modes BIIB_{II} and CIIC_{II} inside the event horizon the inaccessible modes. Taking trace over the inaccessible modes on state ΨABIBIICICII\Psi_{AB_{I}B_{II}C_{I}C_{II}}, we obtain the reduced density operator ρABICI\rho_{AB_{I}C_{I}}

ρABICI=(n1000000c10n200000000n300000000n40000000000000000000000000000c1000000m1),\displaystyle\rho_{AB_{I}C_{I}}=\left(\!\!\begin{array}[]{cccccccc}n_{1}&0&0&0&0&0&0&c_{1}\\ 0&n_{2}&0&0&0&0&0&0\\ 0&0&n_{3}&0&0&0&0&0\\ 0&0&0&n_{4}&0&0&0&0\\ 0&0&0&0&0&0&0&0\\ 0&0&0&0&0&0&0&0\\ 0&0&0&0&0&0&0&0\\ c_{1}&0&0&0&0&0&0&m_{1}\end{array}\!\!\right), (32)

where the matrix elements are written by

n1\displaystyle n_{1} =\displaystyle= α2(eωT+1)2,n2=n3=α2(eωT+eωT+2)1\displaystyle\alpha^{2}(e^{-\frac{\omega}{T}}+1)^{-2},n_{2}=n_{3}=\alpha^{2}(e^{\frac{\omega}{T}}+e^{-\frac{\omega}{T}}+2)^{-1}
a4\displaystyle a_{4} =\displaystyle= α2(eωT+1)2,m1=1α2,c1=α1α2(eωT+1)1.\displaystyle\alpha^{2}(e^{\frac{\omega}{T}}+1)^{-2},m_{1}=1-\alpha^{2},c_{1}=\alpha\sqrt{1-\alpha^{2}}(e^{-\frac{\omega}{T}}+1)^{-1}.

According to Eqs.(14) and (16), we obtain the Svetlichny value and GTE for state ρABICI\rho_{AB_{I}C_{I}},

S(ρABICI)=max{8α2(1α2)(eωT+1)1,4|α2(eωT+1)2(eωT1)2+α21|},\displaystyle S(\rho_{AB_{I}C_{I}})=\max\{8\alpha\sqrt{2(1-\alpha^{2})}(e^{-\frac{\omega}{T}}+1)^{-1},4|\alpha^{2}(e^{\frac{\omega}{T}}+1)^{-2}(e^{\frac{\omega}{T}}-1)^{2}+\alpha^{2}-1|\}, (33)

and

C(ρABICI)=2α1α2(eωT+1)1,\displaystyle C(\rho_{AB_{I}C_{I}})=2\alpha\sqrt{1-\alpha^{2}}(e^{-\frac{\omega}{T}}+1)^{-1}, (34)

respectively. Obviously, the Svetlichny value and GTE depend not only on the state parameter α\alpha, but also on the Hawking temperature TT, meaning that the Hawking radiation of black hole will affect the physically accessible GTN and GTE between Alice, Bob and Charlie.

Refer to caption
Refer to caption
Refer to caption
Refer to caption
Figure 1: The Svetlichny value S(ρABICI)S(\rho_{AB_{I}C_{I}}) and the GTE C(ρABICI)C(\rho_{AB_{I}C_{I}}) as functions of the Hawking temperature TT or the initial parameter α2\alpha^{2} for ω=1\omega=1. (a) α=12\alpha=\frac{1}{\sqrt{2}}, and (b) α=16\alpha=\frac{1}{\sqrt{6}}.

In Fig.1, we plot the Svetlichny value S(ρABICI)S(\rho_{AB_{I}C_{I}}) and the GTE C(ρABICI)C(\rho_{AB_{I}C_{I}}) as functions of the Hawking temperature TT for different initial parameter α\alpha. We find that S(ρABICI)S(\rho_{AB_{I}C_{I}}) is larger than 44 at first and then smaller than 44 with the increase of Hawking temperature TT. The critical Hawking temperature for S(ρABICI)=4S(\rho_{AB_{I}C_{I}})=4 is Tc=ωln[2α2(1α2)1]T_{c}=-\frac{\omega}{\ln[2\alpha\sqrt{2(1-\alpha^{2})}-1]}. This implies that the thermal noise introduced by Hawking temperature destroys the physically accessible GTN between Alice, Bob and Charlie, and takes place “sudden death” at the critical temperature TcT_{c}. However, the physically accessible GTE C(ρABICI)C(\rho_{AB_{I}C_{I}}) is a monotonic decreasing function of TT, and has the asymptotic value α1α2\alpha\sqrt{1-\alpha^{2}} in the limit of infinite Hawking temperature. The GTE C(ρABICI)C(\rho_{AB_{I}C_{I}}) never takes place “sudden death” in the finite Hawking temperature. These results suggest that the GTE in the initial state of Eq.(22) can be distinguished into two different parts: nonlocal and local. The nonlocal GTE is destroyed completely by the Hawking effect after the temperature T>TcT>T_{c}, and finally only the local GTE (whole or partial) is preserved. At this stage, no quantum information tasks based on GTN can work, but tasks based on GTE can still work. In other words, GTE is more suitable for relativistic quantum information tasks than GTN.

By comparing Fig.1(a) and (b), we find that both the critical temperature TcT_{c} for the sudden death of GTN and the asymptotic value of C(ρABICI)C(\rho_{AB_{I}C_{I}}) in the infinite temperature depend on the initial GTE in Eq.(22), i.e., parameter α\alpha. The more the initial GTE is, the longer for the death time of GTN is (i.e., the larger the TcT_{c} is ), and the larger for the asymptotic value of C(ρABICI)C(\rho_{AB_{I}C_{I}}) is. In Fig.1(a)(α=12\alpha=\frac{1}{\sqrt{2}}), the critical Hawking temperature (about 1.13) and the asymptotic value (0.5) of C(ρABICI)C(\rho_{AB_{I}C_{I}}) are the maximal. For α=16\alpha=\frac{1}{\sqrt{6}} (Fig.1(b)), they reduce to about 0.34 and 0.37 respectively.

To further inspect the behaviors of GTN and GTE of the tripartite subsystem ABICIAB_{I}C_{I} in the regions of the initial parameter α(0,1/2)\alpha\in(0,1/\sqrt{2}) and α(1/2,1)\alpha\in(1/\sqrt{2},1), we plot Fig.1(c) and (d). We can see from Fig.1(c) that the GTE C(ρABICI)C(\rho_{AB_{I}C_{I}}) is symmetrical with respect to α2=1/2\alpha^{2}=1/2 in the parameter region α2[0,1]\alpha^{2}\in[0,1]. In fact, this symmetry can also be verified via the fact that Eq.(34) is covariant under the exchange α21α2\alpha^{2}\longleftrightarrow 1-\alpha^{2}. From Fig.1(d), we see that the Svetlichny value S(ρABICI)S(\rho_{AB_{I}C_{I}}) is obviously asymmetrical with respect to α2=1/2\alpha^{2}=1/2. However, the GTN (i.e., the part of S(ρABICI)4S(\rho_{AB_{I}C_{I}})\geq 4) is symmetrical with respect to α2=1/2\alpha^{2}=1/2. This symmetry also can be verified analytically via Eq.(33). Setting cos2η=(1+eωT)1\cos^{2}\eta=(1+e^{-\frac{\omega}{T}})^{-1}, sin2η=(1+eωT)1\sin^{2}\eta=(1+e^{\frac{\omega}{T}})^{-1} and cos2ζ=α2\cos^{2}\zeta=\alpha^{2}, sin2ζ=1α2\sin^{2}\zeta=1-\alpha^{2}, we have (eωT+1)2(eωT1)2=[(1+eωT)1(1+eωT)1]2=[cos2ηsin2η]2=cos2(2η)(e^{\frac{\omega}{T}}+1)^{-2}(e^{\frac{\omega}{T}}-1)^{2}=[(1+e^{-\frac{\omega}{T}})^{-1}-(1+e^{\frac{\omega}{T}})^{-1}]^{2}=[\cos^{2}\eta-\sin^{2}\eta]^{2}=\cos^{2}(2\eta). Thus the second term in the right hand side of Eq.(33) becomes as 4|cos2ζcos2(2η)+sin2ζ|44|\cos^{2}\zeta\cos^{2}(2\eta)+\sin^{2}\zeta|\leq 4. This means that, in the inspection of the GTN of the tripartite subsystem ABICIAB_{I}C_{I}, the second term in the right hand side of Eq.(33) can be ignored and only the first term need to be considered. Therefore, the GTN is symmetrical with respect to α2=1/2\alpha^{2}=1/2. Note that both the GTE and GTN in Fig.1(c) and (d) decrease when Hawking temperature increases, which are consistent with the results from Fig.1(a) and (b). Naturally, in the region α(0,1/2)\alpha\in(0,1/\sqrt{2}), GTN and GTE change slowly with the increase of the α\alpha; in the region α(1/2,1)\alpha\in(1/\sqrt{2},1), GTN and GTE change steeply with the increase of the α\alpha.

We can also make the similar discussions for other tripartite subsystems. Tracing over the modes BIIB_{II} and CIC_{I} on the state ΨABIBIICICII\Psi_{AB_{I}B_{II}C_{I}C_{II}}, we obtain the reduced density operator ρABICII\rho_{AB_{I}C_{II}} as

ρABICII=(n100000000n20000c2000n300000000n4000000000000000000000c20000m2000000000),\displaystyle\rho_{AB_{I}C_{II}}=\left(\!\!\begin{array}[]{cccccccc}n_{1}&0&0&0&0&0&0&0\\ 0&n_{2}&0&0&0&0&c_{2}&0\\ 0&0&n_{3}&0&0&0&0&0\\ 0&0&0&n_{4}&0&0&0&0\\ 0&0&0&0&0&0&0&0\\ 0&0&0&0&0&0&0&0\\ 0&c_{2}&0&0&0&0&m_{2}&0\\ 0&0&0&0&0&0&0&0\end{array}\!\!\right), (43)

with the matrix elements given by

n1\displaystyle n_{1} =\displaystyle= α2(eωT+1)2,n2=n3=α2(eωT+eωT+2)1\displaystyle\alpha^{2}(e^{-\frac{\omega}{T}}+1)^{-2},n_{2}=n_{3}=\alpha^{2}(e^{\frac{\omega}{T}}+e^{-\frac{\omega}{T}}+2)^{-1}
a4\displaystyle a_{4} =\displaystyle= α2(eωT+1)2,m2=1α2,c2=α1α2(eωT+eωT+2)12.\displaystyle\alpha^{2}(e^{\frac{\omega}{T}}+1)^{-2},m_{2}=1-\alpha^{2},c_{2}=\alpha\sqrt{1-\alpha^{2}}(e^{\frac{\omega}{T}}+e^{-\frac{\omega}{T}}+2)^{-\frac{1}{2}}.

The Svetlichny value and the GTE for this state are given by

S(ρABICII)=max{8α2(1α2)(eωT+eωT+2)12,\displaystyle S(\rho_{AB_{I}C_{II}})=\max\{8\alpha\sqrt{2(1-\alpha^{2})}(e^{\frac{\omega}{T}}+e^{-\frac{\omega}{T}}+2)^{-\frac{1}{2}},
4|α2(eωT+1)2(eωT1)2α2+1|}.\displaystyle 4|\alpha^{2}(e^{\frac{\omega}{T}}+1)^{-2}(e^{\frac{\omega}{T}}-1)^{2}-\alpha^{2}+1|\}. (44)

and

C(ρABICII)=2α1α2(eωT+eωT+2)12,\displaystyle C(\rho_{AB_{I}C_{II}})=2\alpha\sqrt{1-\alpha^{2}}(e^{\frac{\omega}{T}}+e^{-\frac{\omega}{T}}+2)^{-\frac{1}{2}}, (45)

respectively. According to the exchange symmetry for Bob and Charlie, we can get

S(ρABIICI)=S(ρABICII),C(ρABIICI)=C(ρABICII).\displaystyle S(\rho_{AB_{II}C_{I}})=S(\rho_{AB_{I}C_{II}}),C(\rho_{AB_{II}C_{I}})=C(\rho_{AB_{I}C_{II}}). (46)

Thus the analysis for tripartite system ABIICIAB_{II}C_{I} is the same as for tripartite system ABICIIAB_{I}C_{II}.

Refer to caption
Refer to caption
Figure 2: The Svetlichny value S(ρABICII)S(\rho_{AB_{I}C_{II}}) and GTE C(ρABICII)C(\rho_{AB_{I}C_{II}}) as functions of the Hawking temperature TT for ω=1\omega=1. (a) α=12\alpha=\frac{1}{\sqrt{2}} and (b) α=16\alpha=\frac{1}{\sqrt{6}}.

In Fig.2, we plot the Svetlichny value S(ρABICII)S(\rho_{AB_{I}C_{II}}) and the GTE C(ρABICII)C(\rho_{AB_{I}C_{II}}) as functions of the Hawking temperature TT for different initial parameter α\alpha. It is shown that C(ρABICII)C(\rho_{AB_{I}C_{II}}) increases from zero and approaches to the asymptotic value α1α2\alpha\sqrt{1-\alpha^{2}} in the infinite Hawking temperature. This means that the Hawking effect can generate physically inaccessible GTE between modes AA, BIB_{I} and CIIC_{II}, even though they are separated by the event horizon of black hole. Physically, it can be regarded as a kind of entanglement transfer: Initially, there is GTE between modes AA, BIB_{I} and CIC_{I}. Lately, the Hawking effect produce entanglement between the modes CIC_{I} and CIIC_{II}, which is equivalent to an interaction between modes CIC_{I} and CIIC_{II}. This interaction transfers some information from mode CIC_{I} to mode CIIC_{II}. Therefore, the GTE between the modes AA, BIB_{I} and CIIC_{II} is established. Comparing Fig.2 (a) and (b), we find that the produced GTE C(ρABICII)C(\rho_{AB_{I}C_{II}}) depends on the initial accessible GTE in Eq.(22). Under given Hawking temperature, more initially accessible GTE will produce more C(ρABICII)C(\rho_{AB_{I}C_{II}}). In the limit of infinite Hawking temperature, the asymptotic value of C(ρABICII)C(\rho_{AB_{I}C_{II}}) is 0.5 in Fig.2(a) and 0.37 in Fig.2(b) respectively. The figure shows that S(ρABICII)S(\rho_{AB_{I}C_{II}}) is always smaller than 44 for any TT, thus the physically inaccessible GTN between modes AA, BIB_{I} and CIIC_{II} cannot be produced. It also means that the produced inaccessible GTE C(ρABICII)C(\rho_{AB_{I}C_{II}}) is local. Similar analysis is also valid for the tripartite system of modes AA, BIIB_{II} and CIC_{I}. The different behaviors between S(ρABICII)S(\rho_{AB_{I}C_{II}}) and C(ρABICII)C(\rho_{AB_{I}C_{II}}) under the Hawking effect suggest that the information flows of different quantum resources inside and outside the event horizon of a black hole are completely different: The entanglement can pass through the event horizon of black hole, while the nonlocality cannot.

Now, we discuss the physically inaccessible GTN and GTE between the modes AA, BIIB_{II} and CIIC_{II}. Tracing over the modes BIB_{I} and CIC_{I} on the state ΨABIBIICICII\Psi_{AB_{I}B_{II}C_{I}C_{II}}, we obtain the reduced density operator ρABIICII\rho_{AB_{II}C_{II}} as

ρABIICII=(n100000000n200000000n300000000n4c4000000c4m4000000000000000000000000000),\displaystyle\rho_{AB_{II}C_{II}}=\left(\!\!\begin{array}[]{cccccccc}n_{1}&0&0&0&0&0&0&0\\ 0&n_{2}&0&0&0&0&0&0\\ 0&0&n_{3}&0&0&0&0&0\\ 0&0&0&n_{4}&c_{4}&0&0&0\\ 0&0&0&c_{4}&m_{4}&0&0&0\\ 0&0&0&0&0&0&0&0\\ 0&0&0&0&0&0&0&0\\ 0&0&0&0&0&0&0&0\end{array}\!\!\right), (55)

with the matrix elements given by

n1\displaystyle n_{1} =\displaystyle= α2(eωT+1)2,n2=n3=α2(eωT+eωT+2)1\displaystyle\alpha^{2}(e^{-\frac{\omega}{T}}+1)^{-2},n_{2}=n_{3}=\alpha^{2}(e^{\frac{\omega}{T}}+e^{-\frac{\omega}{T}}+2)^{-1}
a4\displaystyle a_{4} =\displaystyle= α2(eωT+1)2,m4=1α2,c4=α1α2(eωT+1)1.\displaystyle\alpha^{2}(e^{\frac{\omega}{T}}+1)^{-2},m_{4}=1-\alpha^{2},c_{4}=\alpha\sqrt{1-\alpha^{2}}(e^{\frac{\omega}{T}}+1)^{-1}.

The corresponding Svetlichny value and GTE read

S(ρABIICII)=max{8α2(1α2)(eωT+1)1,4|α2(eωT+1)2(eωT1)2+α21|},\displaystyle S(\rho_{AB_{II}C_{II}})=\max\{8\alpha\sqrt{2(1-\alpha^{2})}(e^{\frac{\omega}{T}}+1)^{-1},4|\alpha^{2}(e^{\frac{\omega}{T}}+1)^{-2}(e^{\frac{\omega}{T}}-1)^{2}+\alpha^{2}-1|\}, (56)

and

C(ρABIICII)=2α1α2(eωT+1)1,\displaystyle C(\rho_{AB_{II}C_{II}})=2\alpha\sqrt{1-\alpha^{2}}(e^{\frac{\omega}{T}}+1)^{-1}, (57)

respectively.

Refer to caption
Refer to caption
Figure 3: The Svetlichny value S(ρABIICII)S(\rho_{AB_{II}C_{II}}) and the GTE C(ρABIICII)C(\rho_{AB_{II}C_{II}}) as functions of the Hawking temperature TT for ω=1\omega=1. (a) α=12\alpha=\frac{1}{\sqrt{2}} and (b) α=16\alpha=\frac{1}{\sqrt{6}}.

In Fig.3, we plot the Svetlichny value S(ρABIICII)S(\rho_{AB_{II}C_{II}}) and the GTE C(ρABIICII)C(\rho_{AB_{II}C_{II}}) as functions of the Hawking temperature TT for different initial parameter α\alpha. We find the similar result as observed in the tripartite system of ABICIIAB_{I}C_{II}: The Hawking effect can generate physically inaccessible GTE between the modes AA, BIIB_{II} and CIIC_{II}, but cannot generate the physically inaccessible GTN, i.e., entanglement can pass through the event horizon of black hole, and nonlocality cannot. The more the initial GTE in Eq.(22) is, the more the produced C(ρABIICII)C(\rho_{AB_{II}C_{II}}) is. The produced GTE has the asymptotic value C(ρABIICII)=α1α2C(\rho_{AB_{II}C_{II}})=\alpha\sqrt{1-\alpha^{2}} in the limit of infinite Hawking temperature, which is 0.5 for Fig.3(a) and 0.37 for Fig.3(b) respectively. The mechanism for the production of this GTE is also the result of entanglement transfer.

Finally, we investigate the physically inaccessible GTN and GTE for the system ρABIBII\rho_{AB_{I}B_{II}}. In the bases |000|000\rangle, |100|100\rangle, |010|010\rangle, |001|001\rangle, |101|101\rangle, |111|111\rangle, |110|110\rangle, and |011|011\rangle for AA, BIB_{I} and BIIB_{II}, the density operator ρABIBII\rho_{AB_{I}B_{II}} has its matrix expression

ρABIBII=(n1000000c10000000000000000000000000000000000000000000000m20c1000000m1),\displaystyle\rho_{AB_{I}B_{II}}=\left(\!\!\begin{array}[]{cccccccc}n_{1}&0&0&0&0&0&0&c_{1}\\ 0&0&0&0&0&0&0&0\\ 0&0&0&0&0&0&0&0\\ 0&0&0&0&0&0&0&0\\ 0&0&0&0&0&0&0&0\\ 0&0&0&0&0&0&0&0\\ 0&0&0&0&0&0&m_{2}&0\\ c_{1}&0&0&0&0&0&0&m_{1}\end{array}\!\!\right), (66)

with the matrix elements given by

n1\displaystyle n_{1} =\displaystyle= α2(eωT+1)1,c1=α2(eωT+eωT+2)12\displaystyle\alpha^{2}(e^{-\frac{\omega}{T}}+1)^{-1},c_{1}=\alpha^{2}(e^{\frac{\omega}{T}}+e^{-\frac{\omega}{T}}+2)^{-\frac{1}{2}}
m1\displaystyle m_{1} =\displaystyle= α2(eωT+1)1,m2=1α2.\displaystyle\alpha^{2}(e^{\frac{\omega}{T}}+1)^{-1},m_{2}=1-\alpha^{2}.

The corresponding Svetlichny value and the GTE read

S(ρABIBII)=max{82α2(eωT+eωT+2)12,4|12α2(eωT+1)1|},\displaystyle S(\rho_{AB_{I}B_{II}})=\max\{8\sqrt{2}\alpha^{2}(e^{\frac{\omega}{T}}+e^{-\frac{\omega}{T}}+2)^{-\frac{1}{2}},4|1-2\alpha^{2}(e^{\frac{\omega}{T}}+1)^{-1}|\}, (67)

and

C(ρABIBII)=2α2(eωT+eωT+2)12,\displaystyle C(\rho_{AB_{I}B_{II}})=2\alpha^{2}(e^{\frac{\omega}{T}}+e^{-\frac{\omega}{T}}+2)^{-\frac{1}{2}}, (68)

respectively. According to the exchange symmetry for Bob and Charlie, we can get the Svetlichny value and GTE between modes AA, CIC_{I} and CIIC_{II},

S(ρACICII)=S(ρABIBII),C(ρACICII)=C(ρABIBII).\displaystyle S(\rho_{AC_{I}C_{II}})=S(\rho_{AB_{I}B_{II}}),C(\rho_{AC_{I}C_{II}})=C(\rho_{AB_{I}B_{II}}). (69)

Thus, we just need to analyze the GTN and GTE for the tripartite system ABIBIIAB_{I}B_{II}.

Refer to caption
Refer to caption
Refer to caption
Refer to caption
Figure 4: The Svetlichny value S(ρABIBII)S(\rho_{AB_{I}B_{II}}), the GTE C(ρABIBII)C(\rho_{AB_{I}B_{II}}), the maximal Bell signal (ρBIBII)\mathcal{B}(\rho_{B_{I}B_{II}}), and the bipartite concurrence C(ρBIBII)C(\rho_{B_{I}B_{II}}) as functions of the Hawking temperature TT for ω=1\omega=1. (a) and (c) α=12\alpha=\frac{1}{\sqrt{2}}, and (b) and (d) α=16\alpha=\frac{1}{\sqrt{6}}.

In Fig.4 (a) and (b), we plot the Svetlichny value S(ρABIBII)S(\rho_{AB_{I}B_{II}}) and GTE C(ρABIBII)C(\rho_{AB_{I}B_{II}}) between Alice, Bob and anti-Bob as functions of the Hawking temperature TT for different initial parameter α\alpha. It is shown that Hawking effect can also produce physically inaccessible GTE between modes AA, BIB_{I} and BIIB_{II}, but cannot produce the physically inaccessible GTN between them. The produced C(ρABIBII)C(\rho_{AB_{I}B_{II}}) increases monotonically from zero and reaches the asymptotic value C(ρABIBII)=α2C(\rho_{AB_{I}B_{II}})=\alpha^{2} for TT\rightarrow\infty. Comparing Fig.4 (a) and (b), we find that C(ρABIBII)C(\rho_{AB_{I}B_{II}}) also depends on the initially accessible GTE in Eq.(22). Under given Hawking temperature, more initially accessible GTE can produce more C(ρABIBII)C(\rho_{AB_{I}B_{II}}).

Besides GTN and GTE for tripartite systems, we can also study the bipartite nonlocality and entanglement for the considered system under the influence of Hawking effect. The bipartite nonlocality and entanglement may be described by the CHSH inequality and concurrence respectively for qubit systems. We review these concepts in the Appendix A and present all the pairwise bipartite nonlocality and entanglement for our considered system in the appendix B. In Fig.4 (c)-(d), we plot the maximal Bell signal (ρBIBII)\mathcal{B}(\rho_{B_{I}B_{II}}) and bipartite concurrence C(ρBIBII)C(\rho_{B_{I}B_{II}}) between the modes BIB_{I} and BIIB_{II} as functions of the Hawking temperature TT for different initial parameter α\alpha. We find that (ρBIBII)\mathcal{B}(\rho_{B_{I}B_{II}}) is always less than 22, meaning that Hawking effect cannot generate Bell nonlocality between Bob and anti-Bob. However, Hawking effect can generate entanglement between Bob and anti-Bob, i.e., entanglement C(ρBIBII)C(\rho_{B_{I}B_{II}}) can pass through the event horizon of black hole. Actually, this is just the physical nature of the Hawking radiation–produce entangled pairs of particle and antiparticle between the causally disconnected regions. The same analysis is also valid for the reduced state ρCICII\rho_{C_{I}C_{II}}. In fact, from the calculation in Appendix B, we can find that all the reduced bipartite subsystems in state ΨABIBIICICII\Psi_{AB_{I}B_{II}C_{I}C_{II}} have no nonlocality, meaning that nonlocality cannot pass through the event horizon of black hole. Except for C(ρBIBII)C(\rho_{B_{I}B_{II}}) and C(ρCICII)C(\rho_{C_{I}C_{II}}), there is no pairs of entanglement in any other reduced bipartite subsystems. This result is easy to understand: Initially, there is no entanglement between modes AA and BB, modes AA and CC, modes BB and CC. The Hawking radiation is essentially a local operation on BIBIIB_{I}B_{II} or CICIIC_{I}C_{II}, it of course can only produce quantum entanglement between modes BIB_{I} and BIIB_{II} or CIC_{I} and CIIC_{II}, and cannot produce any other kinds of bipartite entanglement.

From the discussions of Figs.1-4, we find that the physically accessible GTN, i.e., S(ρABICI)4S(\rho_{AB_{I}C_{I}})-4 reduces with Hawking temperature and suffers from a “sudden death” at some critical Hawking temperature, but the physically inaccessible GTN is never generated by the Hawking effect, i.e., S(ρABICII)S(\rho_{AB_{I}C_{II}}), S(ρABIICI)S(\rho_{AB_{II}C_{I}}), S(ρABIICII)S(\rho_{AB_{II}C_{II}}), S(ρABIBII)S(\rho_{AB_{I}B_{II}}), and S(ρACICII)S(\rho_{AC_{I}C_{II}}) are all less than 4. This result suggests that the GTN can not be redistributed. However, the GTE behaves differently. When the physically accessible GTE reduces with Hawking temperature, at the same time, the physically inaccessible GTE is generated. This result implies that the GTE may be redistributable through the Hawking effect. To manifest this inference, we try to find some monogamy relation for the GTE. Through careful inspection, we find three monogamy relations between the physically accessible GTE and the physically inaccessible GTE,

C(ρABICI)+C(ρABIICII)=2α1α2,\displaystyle C(\rho_{AB_{I}C_{I}})+C(\rho_{AB_{II}C_{II}})=2\alpha\sqrt{1-\alpha^{2}}, (70)
C(ρABICI)2+C(ρABIICI)2+C(ρABICII)2+C(ρABIICII)2=4α2(1α2),\displaystyle C(\rho_{AB_{I}C_{I}})^{2}+C(\rho_{AB_{II}C_{I}})^{2}+C(\rho_{AB_{I}C_{II}})^{2}+C(\rho_{AB_{II}C_{II}})^{2}=4\alpha^{2}(1-\alpha^{2}), (71)
α2[C(ρABICI)2+C(ρABIICII)2]+(1α2)[C(ρABIBII)2+C(ρACICII)2]=4α2(1α2),\displaystyle\alpha^{2}[C(\rho_{AB_{I}C_{I}})^{2}+C(\rho_{AB_{II}C_{II}})^{2}]+(1-\alpha^{2})[C(\rho_{AB_{I}B_{II}})^{2}+C(\rho_{AC_{I}C_{II}})^{2}]=4\alpha^{2}(1-\alpha^{2}), (72)

where 2α1α22\alpha\sqrt{1-\alpha^{2}} is the initial GTE in state of Eq.(22). These monogamy relations reflect the restrictions in the redistribution of entanglement from physically accessible to physically inaccessible patterns. Especially the Eq.(70) shows that the total sum of the physically accessible GTE C(ρABICI)C(\rho_{AB_{I}C_{I}}) and the physically inaccessible GTE C(ρABIICII)C(\rho_{AB_{II}C_{II}}) is equal to the initial GTE. The monogamy relations are important for understanding the transfer of quantum information in relativistic spacetime.

Besides the above monogamy relations, we also find that the physically accessible entanglement fulfills the following Coffman-Kundu-Wootters monogamy inequality

C(ρijk)2C(ρij)2+C(ρik)2,\displaystyle C(\rho_{ijk})^{2}\geq C(\rho_{ij})^{2}+C(\rho_{ik})^{2}, (73)

where (i,j,ki,j,k) denote all the permutations of the three modes A,BI,CIA,B_{I},C_{I}. These Coffman-Kundu-Wootters monogamy inequalities reflect the distribution of the physically accessible entanglement in the environment of Schwarzschild black hole.

V Conclusions

The effect of Hawking radiation on the GTN and GTE for Dirac fields in Schwarzschild spacetime has been investigated. It has been shown that Hawking effect degrades both the physically accessible GTN and the physically accessible GTE, where the former takes place “sudden death” at some critical Hawking temperature, and the latter approaches to the nonzero asymptotic value in the infinite Hawking temperature. This means that on the one hand the surviving physically accessible GTE is not nonlocal, and on the other hand the GTE has more resistance to the Hawking noise than GTN.

Further investigation has demonstrated that the GTN is not redistributable, but GTE can be redistributed through Hawking effect. With the growth of the Hawking temperature, the physically accessible GTN decreases, but no physically inaccessible GTN is generated. The GTE however behaves differently: With the loss of the physically accessible GTE, the physically inaccessible GTE is generated continually through Hawking effect. Further, the physically accessible GTE and the physically inaccessible GTE fulfil some monogamy relations. All these phenomena suggest that the GTE is redistributable. We can regard the redistribution of entanglement as a kind of phenomenon of information tunnelling, i.e. the flow of quantum entanglement can pass through the event horizon of black hole, but the flow of quantum nonlocality can not. This result has been demonstrated also by the pairwise bipartite nonlocality and entanglement in the underlying system considered in this paper.

Note that the authors in reference L31 studied the tripartite entanglement in environment of Schwarzschild black hole used the measure of π\pi-tangle. They found that the physically accessible π\pi-tangle decreases with Hawking temperature and approaches to a nonzero asymptotic value for the infinite Hawking temperature. Similar Coffman-Kundu-Wootters monogamy inequalities to Eq.(73) for π\pi-tangle was found. These results agree with ours.

Quantum entanglement and nonlocality are the important manifestations of quantum correlation. They have potential applications in various fields of science. We expect our research can present helps for these applications and enriches the theory of the relativistic quantum information science.

Acknowledgements.
This work is supported by the National Natural Science Foundation of China (Grant Nos. 1217050862, 11275064), and the Construct Program of the National Key Discipline.

Appendix A CHSH inequality and concurrence for bipartite systems

In this appendix, we review the concepts of CHSH inequality and concurrence for bipartite systems. The CHSH inequality is an important tool for judging quantum nonlocality. It is considered to be both necessary and sufficient conditions of the separability for bipartite pure states. The key element for the CHSH inequality is the Bell operator defined by L50

CHSH=𝒂𝝈(𝒃+𝒃)𝝈+𝒂𝝈(𝒃𝒃)𝝈,\displaystyle\mathcal{B}_{CHSH}=\bm{a}\cdot\bm{\sigma}\otimes(\bm{b}+\bm{b^{\prime}})\cdot\bm{\sigma}+\bm{a^{\prime}}\cdot\bm{\sigma}\otimes(\bm{b}-\bm{b^{\prime}})\cdot\bm{\sigma}, (74)

where 𝒂\bm{a}, 𝒂\bm{a^{\prime}}, 𝒃\bm{b} and 𝒃\bm{b^{\prime}} are unit vectors in 3\mathbb{R}^{3}, and 𝝈=(σ1,σ2,σ3)\bm{\sigma}=(\sigma_{1},\sigma_{2},\sigma_{3}) is the vector of Pauli matrices. For any bipartite mixed state ρ\rho, the well-known CHSH inequality can be expressed as

(ρ)=|Tr(ρCHSH)|2.\displaystyle\mathcal{B}(\rho)=|{\rm Tr}(\rho\mathcal{B}_{CHSH})|\leq 2. (75)

CHSH inequality holds for any state ρ\rho that admits local hidden variable model, and the violation of it implies the Bell nonlocality of the underlying state. In practical applications, we need to find the maximal Bell signal (ρ)\mathcal{B}(\rho), which can be equivalently expressed for two-qubit systems as

(ρ)=2maxi<j(𝒵i+𝒵j),\displaystyle\mathcal{B}(\rho)=2\sqrt{\max_{i<j}(\mathcal{Z}_{i}+\mathcal{Z}_{j})}, (76)

where 𝒵i\mathcal{Z}_{i} and 𝒵j\mathcal{Z}_{j} are the two largest eigenvalues of U(ρ)=TρTTρU(\rho)=T^{\rm T}_{\rho}T_{\rho}, and the correlation matrix is defined by T=(tij)T=(t_{ij}) with tij=Tr[ρσiσj]t_{ij}={\rm Tr}[\rho\sigma_{i}\otimes\sigma_{j}]. Bell nonlocality can be witnessed by the maximum violation of CHSH inequality.

For two-qubit X-state,

ρX=(ρ1100ρ140ρ22ρ2300ρ23ρ330ρ1400ρ44),\displaystyle\rho^{X}=\left(\!\!\begin{array}[]{cccccccc}\rho_{11}&0&0&\rho_{14}\\ 0&\rho_{22}&\rho_{23}&0\\ 0&\rho_{23}&\rho_{33}&0\\ \rho_{14}&0&0&\rho_{44}\end{array}\!\!\right), (81)

with all elements ρij\rho_{ij} being real, the three eigenvalues corresponding to the matrix U(ρ)=TρTTρU(\rho)=T^{\rm T}_{\rho}T_{\rho} are,

𝒵1=4(|ρ14|+|ρ23|)2,𝒵2=4(|ρ14||ρ23|)2,𝒵3=(|ρ11||ρ22||ρ33|+|ρ44|)2.\displaystyle\mathcal{Z}_{1}=4(|\rho_{14}|+|\rho_{23}|)^{2},\mathcal{Z}_{2}=4(|\rho_{14}|-|\rho_{23}|)^{2},\mathcal{Z}_{3}=(|\rho_{11}|-|\rho_{22}|-|\rho_{33}|+|\rho_{44}|)^{2}. (82)

As 𝒵1\mathcal{Z}_{1} is greater than 𝒵2\mathcal{Z}_{2}, so the maximal Bell signal reads

(ρX)=max{1,2},\displaystyle\mathcal{B}(\rho^{X})=\max\{\mathcal{B}_{1},\mathcal{B}_{2}\}, (83)

with 1=2𝒵1+𝒵2\mathcal{B}_{1}=2\sqrt{\mathcal{Z}_{1}+\mathcal{Z}_{2}} and 2=2𝒵1+𝒵3\mathcal{B}_{2}=2\sqrt{\mathcal{Z}_{1}+\mathcal{Z}_{3}} L51 ; L52 .

In addition, the concurrence for quantifying the entanglement of the two-qubit X-state of Eq.(81) can be calculated through the expression,

C(ρX)=2max{|ρ14|ρ22ρ33,|ρ23|ρ11ρ44},\displaystyle C(\rho^{X})=2\max\{|\rho_{14}|-\sqrt{\rho_{22}\rho_{33}},|\rho_{23}|-\sqrt{\rho_{11}\rho_{44}}\}, (84)

where ρij\rho_{ij} is the element of density matrix ρX\rho^{X} L45 .

Appendix B Pairwise quantum nonlocality and entanglement in state ΨABIBIICICII\Psi_{AB_{I}B_{II}C_{I}C_{II}}

We now study the pairwise bipartite quantum nonlocality and entanglement in the state ΨABIBIICICII\Psi_{AB_{I}B_{II}C_{I}C_{II}}. By tracing over the irrelevant modes on state ΨABIBIICICII\Psi_{AB_{I}B_{II}C_{I}C_{II}}, we obtain all the pairwise density operators as

ρBIBII=(α2(eωT+1)00α2eωT+eωT+20000001α20α2eωT+eωT+200α2(eωT+1)),\displaystyle\rho_{B_{I}B_{II}}=\left(\!\!\begin{array}[]{cccccccc}\frac{\alpha^{2}}{(e^{-\frac{\omega}{T}}+1)}&0&0&\frac{\alpha^{2}}{\sqrt{e^{\frac{\omega}{T}}+e^{-\frac{\omega}{T}}+2}}\\ 0&0&0&0\\ 0&0&1-\alpha^{2}&0\\ \frac{\alpha^{2}}{\sqrt{e^{\frac{\omega}{T}}+e^{-\frac{\omega}{T}}+2}}&0&0&\frac{\alpha^{2}}{(e^{\frac{\omega}{T}}+1)}\end{array}\!\!\right), (89)
ρABI=ρACI=(α2(eωT+1)0000α2(eωT+1)0000000001α2),\displaystyle\rho_{AB_{I}}=\rho_{AC_{I}}=\left(\!\!\begin{array}[]{cccccccc}\frac{\alpha^{2}}{(e^{-\frac{\omega}{T}}+1)}&0&0&0\\ 0&\frac{\alpha^{2}}{(e^{\frac{\omega}{T}}+1)}&0&0\\ 0&0&0&0\\ 0&0&0&1-\alpha^{2}\end{array}\!\!\right), (94)
ρBICI=(α2(eωT+1)20000α2(eωT+eωT+2)0000α2(eωT+eωT+2)0000α2(eωT+1)2+1α2,),\displaystyle\rho_{B_{I}C_{I}}=\left(\!\!\begin{array}[]{cccccccc}\frac{\alpha^{2}}{(e^{-\frac{\omega}{T}}+1)^{2}}&0&0&0\\ 0&\frac{\alpha^{2}}{(e^{\frac{\omega}{T}}+e^{-\frac{\omega}{T}}+2)}&0&0\\ 0&0&\frac{\alpha^{2}}{(e^{\frac{\omega}{T}}+e^{-\frac{\omega}{T}}+2)}&0\\ 0&0&0&\frac{\alpha^{2}}{(e^{\frac{\omega}{T}}+1)^{2}}+1-\alpha^{2},\end{array}\!\!\right), (99)
ρBIICII=(α2(eωT+1)2+1α20000α2(eωT+eωT+2)0000α2(eωT+eωT+2)0000α2(eωT+1)2,),\displaystyle\rho_{B_{II}C_{II}}=\left(\!\!\begin{array}[]{cccccccc}\frac{\alpha^{2}}{(e^{-\frac{\omega}{T}}+1)^{2}}+1-\alpha^{2}&0&0&0\\ 0&\frac{\alpha^{2}}{(e^{\frac{\omega}{T}}+e^{-\frac{\omega}{T}}+2)}&0&0\\ 0&0&\frac{\alpha^{2}}{(e^{\frac{\omega}{T}}+e^{-\frac{\omega}{T}}+2)}&0\\ 0&0&0&\frac{\alpha^{2}}{(e^{\frac{\omega}{T}}+1)^{2}},\end{array}\!\!\right), (104)
ρABII=ρACII=(α2(eωT+1)0000α2(eωT+1)00001α200000),\displaystyle\rho_{AB_{II}}=\rho_{AC_{II}}=\left(\!\!\begin{array}[]{cccccccc}\frac{\alpha^{2}}{(e^{-\frac{\omega}{T}}+1)}&0&0&0\\ 0&\frac{\alpha^{2}}{(e^{\frac{\omega}{T}}+1)}&0&0\\ 0&0&1-\alpha^{2}&0\\ 0&0&0&0\end{array}\!\!\right), (109)
ρBICII=ρBIICI=(α2(eωT+1)20000α2(eωT+eωT+2)0000α2(eωT+eωT+2)+1α20000α2(eωT+1)2).\displaystyle\rho_{B_{I}C_{II}}=\rho_{B_{II}C_{I}}=\left(\!\!\begin{array}[]{cccccccc}\frac{\alpha^{2}}{(e^{-\frac{\omega}{T}}+1)^{2}}&0&0&0\\ 0&\frac{\alpha^{2}}{(e^{\frac{\omega}{T}}+e^{-\frac{\omega}{T}}+2)}&0&0\\ 0&0&\frac{\alpha^{2}}{(e^{\frac{\omega}{T}}+e^{-\frac{\omega}{T}}+2)}+1-\alpha^{2}&0\\ 0&0&0&\frac{\alpha^{2}}{(e^{\frac{\omega}{T}}+1)^{2}}\end{array}\!\!\right). (114)

Using Eqs.(83) and (84), it is easy to find the maximal Bell signal and concurrence for the corresponding pairwise qubit systems,

(ρBIBII)=max{42α2(eωT+eωT+2)12,2(2α21)2+4α2(eωT+eωT+2)12},\displaystyle\mathcal{B}(\rho_{B_{I}B_{II}})=\max\{4\sqrt{2}\alpha^{2}(e^{\frac{\omega}{T}}+e^{-\frac{\omega}{T}}+2)^{-\frac{1}{2}},2\sqrt{(2\alpha^{2}-1)^{2}+4\alpha^{2}(e^{\frac{\omega}{T}}+e^{-\frac{\omega}{T}}+2)^{-\frac{1}{2}}}\}, (115)
(ρABI)=(ρACI)=2[α2(eωT+1)1(eωT1)+1α2]<2,\displaystyle\mathcal{B}(\rho_{AB_{I}})=\mathcal{B}(\rho_{AC_{I}})=2[\alpha^{2}(e^{\frac{\omega}{T}}+1)^{-1}(e^{\frac{\omega}{T}}-1)+1-\alpha^{2}]<2, (116)
(ρBICI)=(ρBIICII)=2[α2(eωT+1)2(eωT1)2+1α2]<2,\displaystyle\mathcal{B}(\rho_{B_{I}C_{I}})=\mathcal{B}(\rho_{B_{II}C_{II}})=2[\alpha^{2}(e^{\frac{\omega}{T}}+1)^{-2}(e^{\frac{\omega}{T}}-1)^{2}+1-\alpha^{2}]<2, (117)
(ρABII)=(ρACII)=2[α2(eωT+1)1(eωT1)1+α2]<2,\displaystyle\mathcal{B}(\rho_{AB_{II}})=\mathcal{B}(\rho_{AC_{II}})=2[\alpha^{2}(e^{\frac{\omega}{T}}+1)^{-1}(e^{\frac{\omega}{T}}-1)-1+\alpha^{2}]<2, (118)
(ρBICII)=(ρBIICI)=2[α2(eωT+1)2(eωT1)21+α2]<2,\displaystyle\mathcal{B}(\rho_{B_{I}C_{II}})=\mathcal{B}(\rho_{B_{II}C_{I}})=2[\alpha^{2}(e^{\frac{\omega}{T}}+1)^{-2}(e^{\frac{\omega}{T}}-1)^{2}-1+\alpha^{2}]<2, (119)

and

C(ρBIBII)=max{0,2α2(eωT+eωT+2)12},\displaystyle C(\rho_{B_{I}B_{II}})=\max\{0,2\alpha^{2}(e^{\frac{\omega}{T}}+e^{-\frac{\omega}{T}}+2)^{-\frac{1}{2}}\}, (120)
C(ρABI)=C(ρACI)=0,\displaystyle C(\rho_{AB_{I}})=C(\rho_{AC_{I}})=0, (121)
C(ρBICI)=C(ρBIICII)=0,\displaystyle C(\rho_{B_{I}C_{I}})=C(\rho_{B_{II}C_{II}})=0, (122)
C(ρABII)=C(ρACII)=0,\displaystyle C(\rho_{AB_{II}})=C(\rho_{AC_{II}})=0, (123)
C(ρBICII)=C(ρBIICI)=0.\displaystyle C(\rho_{B_{I}C_{II}})=C(\rho_{B_{II}C_{I}})=0. (124)

References

  • (1) A. Einstein, B. Podolsky, and N. Rosen, Phys. Rev. 47, 777 (1935).
  • (2) J. S. Bell, Physics 1, 195 (1964).
  • (3) N. Brunner, D. Cavalcanti, S. Pironio, V. Scarani, S. Wehner, Rev.Mod. Phys. 86, 419 (2014).
  • (4) S. J. Freedman, J. F. Clauser, Phys. Rev. Lett. 28, 938 (1972).
  • (5) A. Aspect, P. Grangier, G. Roger, Phys. Rev. Lett. 47, 460 (1981).
  • (6) B. Hensen, et al, Nature 526, 682 (2015).
  • (7) J. Handsteiner, et al, Phys. Rev. Lett. 118, 060401 (2017).
  • (8) Z. Su, L. Li, J. Ling, Quantum Inf Process 17, 85 (2018).
  • (9) K. Wang, Y, Liang, Z. J. Zheng, Quantum Inf. Process 19, 140 (2020).
  • (10) H. Buhrman, R. Cleve, S. Massar, R. D. Wolf, Rev. Mod. Phys. 82, 665 (2010).
  • (11) C. H. Bennett, G. Brassard, N. D. Mermin, Phys. Rev. Lett. 68, 557 (1992).
  • (12) C. Dhara, G. Prettico, A. Antonio, Phys. Rev. A 88, 052116 (2013).
  • (13) J. Barrett, L. Hardy, A. Kent, Phys. Rev. Lett. 95, 010503 (2005).
  • (14) C. H. Bennett, et al, Phys. Rev. A 59, 1070 (1999).
  • (15) S. Popescu, D. Rohrlich, Found. Phys. 24, 379 (1994).
  • (16) G. Svetlichny, Phys. Rev. D 35, 3066 (1987).
  • (17) I. Fuentes-Schuller and R. B. Mann, Phys. Rev. Lett. 95, 120404 (2005).
  • (18) P. M. Alsing, I. Fuentes-Schuller, R. B. Mann, and T. E. Tessier, Phys. Rev. A 74, 032326 (2006).
  • (19) J. Wang, Q. Pan, S. Chen, J. Jing, Phys. Lett. B 677, 186 (2009).
  • (20) E. Martín-Martínez, L. J. Garay, and J. León, Phys. Rev. D 82, 064006 (2010).
  • (21) J. Wang, Q. Pan, J. Jing, Annals Phys. 325, 1190 (2010).
  • (22) E. Martín-Martínez, L. J. Garay, and J. León, Phys. Rev. D 82, 064028 (2010).
  • (23) S. Kanno, J. P. Shock, and J. Soda, Phys. Rev. D 94, 125014 (2016).
  • (24) J. Wang, Q. Pan, J. Jing, Phys. Lett. B 692, 202 (2010).
  • (25) Z. Tian, J. Jing, J. High. Energy. Phys. 04, 109 (2013).
  • (26) B. Nasr Esfahani, M. Shamirzaie, M. Soltani, Phys. Rev. D 84, 025024 (2011).
  • (27) Z. Tian, J. Jing, J. High. Energy. Phys. 07, 089 (2014).
  • (28) X. Liu, Z. Tian, J. Wang, J. Jing, Phys. Rev. D 97, 105030 (2018).
  • (29) S. M. Wu, H. S. Zeng, Int. J. Theor. Phys. 59, 861 (2020).
  • (30) S. M. Wu, H. S. Zeng, Class. Quantum Grav. 37, 115003 (2020).
  • (31) J. Wang, J. Jing, Phys. Rev. A 82, 032324 (2010).
  • (32) S. Khan and M. K. Khan, J. Phys. A: Math. Theor. 44, 045305 (2011).
  • (33) J. Wang, and J. Jing, Phys. Rev. A 83, 022314 (2011).
  • (34) S. Xu, X. K. Song, J. D. Shi, and L. Ye, Phys. Rev. D 89, 065022 (2014).
  • (35) Y. Dai, Z. Shen, and Y. Shi, Phys. Rev. D 94, 025012 (2016).
  • (36) A. J. Torres-Arenas, Q. Dong, G. H. Sun, W. C. Qiang, S. H. Dong, Phys. Lett. B 789, 93 (2019).
  • (37) W. Zhang, J. Jing, arXiv:1103.4903.
  • (38) S. Khan, Journal of Modern Optics 59, 250 (2012).
  • (39) M. R. Hwang, D. Park, and E. Jung, Phys. Rev. A 83, 012111 (2011).
  • (40) M. B. Plenio, Phys. Rev. Lett. 95, 090503 (2005).
  • (41) Y. C. Ou, H. Fan, Phys. Rev. A 75, 062308 (2007).
  • (42) Z. H. Ma, Z. H. Chen, J. L. Chen, C. Spengler, A. Gabriel, and M. Huber, Phys. Rev. A 83, 062325 (2011).
  • (43) S. M. Hashemi Rafsanjani, M. Huber, C. J. Broadbent, and J. H. Eberly, Phys. Rev. A 86, 062303 (2012).
  • (44) R. Gallego, L. E. Würflinger, A. Acín, M. Navascués, Phys. Rev. Lett. 109, 070401 (2012).
  • (45) J. D. Bancal, J. Barrett, N. Gisin, S. Pironio, Phys. Rev. A 88, 0440102 (2013).
  • (46) K. Mukherjee, B. Paul, D. Sarkar, J. Phys. A: Math. Theor. 48, 465302 (2015).
  • (47) K. Wang, Z. J. Zheng, Sci. Rep. 10, 6621 (2020).
  • (48) D. R. Brill and J. A. Wheeler, Rev. Mod. Phys. 29, 465 (1957).
  • (49) W. C. Qiang, G. H. Sun, Q. Dong, and S. H. Dong, Phys. Rev. A 98, 022320 (2018).
  • (50) J. Jing, Phys. Rev. D 70, 065004 (2004).
  • (51) T. Damoar and R. Ruffini, Phys. Rev. D 14, 332 (1976).
  • (52) S. M. Barnett and P. M. Radmore, Methods in Theoretical Quantum Optics, 67-80 (Oxford University Press, New York), 1997.
  • (53) S. M. Wu, Z. C. Li, H. S. Zeng, EPL-Europhys. Lett. 129, 40002 (2020).
  • (54) J. F. Clauser, M. A. Horne, A. Shimony, and R. A. Holt, Phys. Rev. Lett. 23, 880 (1969).
  • (55) P. Shadbolt, et al, Sci. Rep. 2, 470 (2012).
  • (56) W. Y. Sun, D. Wang, J. D. Shi, L. Ye, Sci. Rep. 7, 39651 (2017).