Corresponding author. E-mail: szhang@ybu.edu.cn
Project supported by the National Natural Science Foundation of China (Grant Nos. 11264042, 61465013, 11465020, and 11165015), the Program for Chun Miao Excellent Talents of Jilin Provincial Department of Education (Grant No. 201316), and the Talent Program of Yanbian University of China (Grant No. 950010001).
We design proposals to generate a remote Greenberger–Horne–Zeilinger (GHZ) state and a W state of nitrogen-vacancy (NV) centers coupled to microtoroidal resonators (MTRs) through noisy channels by utilizing time-bin encoding processes and fast-optical-switch-based polarization rotation operations. The polarization and phase noise induced by noisy channels generally affect the time of state generation but not its success probability and fidelity. Besides, the above proposals can be generalized to n-qubit between two or among n remote nodes with success probability unity under ideal conditions. Furthermore, the proposals are robust for regular noise-changeable channels for the n-node case. This method is also useful in other remote quantum information processing tasks through noisy channels.
Quantum entanglement lies at the heart of quantum information science and technology and is thought to be the key resource to realize quantum computation and quantum information processing (QIP) tasks, [1– 4] such as quantum cryptography, [1] quantum teleportation, [3] quantum secure direct communication, [5] quantum cloning machine, [6] and so on. For the case of a tripartite system, it is well known that there are two kinds of entanglement, one is the Greenberger– Horne– Zeilinger (GHZ) state,
the other is the W state,
The GHZ state is non-equivalent to W state owing to the fact that they cannot be converted to each other by local operations and classical communication. Great efforts have been taken to research both the GHZ and W states since these two states have been shown to have valuable applications in quantum information science.[7– 12] Lots of theoretical and experiment schemes have been proposed and reported to generate these two states.[13– 23]
Generating remote entanglement is critical for constructing a quantum communication network and realizing long-range quantum information processing.[24, 25] Many meaningful schemes have been proposed to generate remote entangled states[26] or implement quantum computation[27, 28] in distant participants. For the theoretical long distance quantum information processing case, the noise is not seriously considered in general. But it is one of the most important problems in experiments and practical applications since the quantum system would inevitably couple with its environment. This interaction will decrease the reliability of the quantum channel or even cause the quantum states to be changed.[29, 30] Some schemes, such as entanglement purification, [31– 34] error correction, [35– 38] and quantum repeaters[39, 40] have been proposed to solve this problem. In contrast to the above schemes, Song et al. proposed a robust scheme[41] to generate a remote atomic W state in a polarization noisy channel through an interesting method, i.e. time-bin encoding, which was first proposed by Brendel et al.[42] and used widely in QIP tasks.[43– 53] The polarization noise in their scheme had no influence on the fidelity of generating the intended state but it would reduce the total success probability to (1/2)n– 1 for n nodes. In addition, Huang et al. proposed a fault tolerant quantum secure direct communication against collective noise.[54] Wu et al. also proposed three quantum dialogue protocols under the condition of collection noise.[55]
Recently, a novel solid-state QIP system, the NV center in diamond, has been examined intensively[56] because it has optical controllability and good long-lived spin triplets, even at room temperature.[57] To generate entanglement, the NV center is always designed to couple to one of the following resonators, i.e. nanomechanical resonators (NAMRs)[58] and MTRs[59, 60] with a quantized whispering-gallery mode (WGM), [19, 61] or embed in cavities.[62, 63] During these apparatus, the NV center coupled to MTRs with a quantized WGM is highly competitive because the required Q factor of the MTR can be surely degraded when it couples to a fiber, which allows the photons to be inputted and outputted on demand through fibers. This feature enables researchers to deterministically generate entanglement of different NV centers in a scalable fashion. Experimentally, by using microspheres, [64, 65] diamond-GaP microdisks, [66] and SiN photonic crystals, [67] a strong coupling has been demonstrated between the NV center and the WGM. Besides, MTRs or MTRs-like equipment have been paid more attention since single-photon input– output process with the MTR coupling to atom has been reported.[59] Moreover, the quantum non-demolition measurement for the electronic spin state of diamond NV centers has been proposed[68] and experimentally demonstrated.[69] These works have shown that an NV center coupled to MTR is a suitable platform for QIP tasks.
Inspired by the above researches, we propose to generate remote an n-qubit GHZ state and a W state of NV centers coupled to MTRs through noisy channels by using time-bin encoding processes and fast optical switches. The polarization and phase noise induced by noisy channels generally affect the time of state generation but not its success probability and fidelity. The structure of this paper is arranged as follows. In Section 2, we briefly introduce the rationale of the basic model. In Section 3, we describe the process to generate the four-qubit GHZ state and W state through a noisy channel in two nodes. In Section 4, the method to generalize the proposals to n-qubit is given through regular noise-changeable channels among n nodes. A discussion and conclusion are given in Section 5.
The NV centers considered here are negatively charged with six electrons from the nitrogen and three carbons surrounding the vacancy. The ground state are electronic spin triplet states | 0⟩ (ms = 0), | + 1⟩ (ms = 1), and | − 1⟩ (ms = − 1). The six excited states are recorded as | A1⟩ , | A2⟩ , | Ex⟩ , | Ey⟩ , | E1⟩ , and | E2⟩ .[61, 70] Ground states | 0⟩ and | ± 1⟩ are split with a gap of 2.87 GHz as a result of spin– spin interactions. The excited states are eigenvectors of the full Hamiltonian, which consists of a spin– spin interaction and a spin– orbit interaction without any perturbation. The pair (A1, A2) is split from the others at about 5.5 GHz owing to the spin– orbit interaction. The gap between states | A⟩ 1 and | A⟩ 2 would be increased to 3.3 GHz through the spin– spin interaction. Consequently, state | A⟩ 2 would decay to sublevels | − 1⟩ and | + 1⟩ with emitting left (L) and right (R) polarization photons because state | A⟩ 2 is robust with the stable symmetric properties in the limit of low strain, which has been demonstrated in Ref. [70]. With the aim of detecting ground state information conveniently and obtaining appropriate state transformation process, an external magnetic field B0 along the NV center symmetry axis should be introduced to the split states | − 1⟩ and | + 1⟩ . The effective energy levels of the NV centers used throughout our proposal are | A⟩ 2, | + 1⟩ , and | – ⟩ , respectively. A sketch drawing is shown in Fig. 1. Considering a process where a single-photon pulse with frequency ω in inputs in an MTR cavity with the cavity mode frequency ω c, and interacts with the NV center through input– output process. Because the qubits are encoded in states | − 1⟩ and | + 1⟩ , the weak excitation condition, i.e. ⟨ σ z⟩ = − 1, should be satisfied throughout this paper. By combining the Heisenberge motion equations of the NV center lowing operators (σ – ) and cavity field operator (a), and the input– output relationship
where ω 0, κ , γ , and g denote the transition frequency between states | A2⟩ and | − 1⟩ , cavity damping rate, NV center dipolar decay rate, and the coupling coefficient between cavity and NV center, respectively. Under the condition ω c = ω 0 = ω in, the reflection coefficients would be simplified to
After choosing coupling strength
after applying a phase shifter with π phase on the output pulse.
We now describe the process of GHZ state generation through a noisy channel via time-bin encoding. The sketch of the whole process is shown in Fig. 2. Assuming that the input photon is prepared in the state
After photon passing through PBS1, NV1, NV2, and PBS2, the state of the whole system is transformed to
where the subscripts L0 and L1 correspond to the possible paths L0 and L1, respectively. A similar representing method will be used in the following discussion. After the optical switches S1 and S2, the whole system state evolves to
It is important that the optical switches S1 and S2, transmitted when photon in path L0 passing through, are adjusted to reflect just before the photon in path L1 passes through. The corresponding sketch of the time control sequences is shown in Fig. 3. Equation (8) implies that the polarization state of the photon is in | R⟩ state before passing through the noisy channel. This merit enables the photon pulse to see the same noise in the noisy channel, which could be expressed as
in which eiφ and eiω are phase noise, μ and ν are polarization noise parameters and satisfy the relation | μ |2+ |ν |2 = 1. Then, the evolution of the whole system after the photon pulse passes through the noisy channel could be written as
After the photon passes through optical switches S3 and S4, the whole system state is changed to
It should be noted that the optical switches S3 and S4, which are always transmitting when the earlier photon pulse (L0, L) passes through, are adjusted to reflect just before the later photon pulse (L1, L) comes. The control sequences of switches S5 and S6, S7 and S8 are the same as S3 and S4. Then, the whole state evolves to
after the photon pulse goes through PBS3, S5– S8. Immediately following, the photon pulse will go through PBS4– PBS8 and the system state is transformed to
From Eq. (13), we can see that photons in path c or d are by far synchronized because the effective path length they passed are both equal to L0+ L1+ L. After the subsequent delay line LD in path d, optical switch S9, PBS9, NV3, and NV4 work, and the state of the system evolves to
Then, the photon pulse in paths e and f will be delivered into path g and h through a 50:50 beam splitter, whose function could be expressed as[74]
And after the photon pulse passes through HWP5 and HWP6, the relevant system state is converted into
in which
and
and subscripts Di (i = 1, 2, 3, 4) of the polarization state mean that this polarization state will be detected by the detector Di. From Eq. (16), we can see the following results. (i) The phase noise produced in channel has no influence on the state generation because it has been converted to a collective phase which has no observable effects. (ii) Polarization noise μ and ν can influence the time of the desired state generation rather than the success probability and fidelity of the state generation because the state would be transformed to the maximum entangled GHZ state after any one of the four detectors clicks at the time (L0+ L1+ L)/c or (L0+ L1+ L+ LD)/c. If detector D1 clicks at time (L0+ L1+ L)/c, then the system state will collapse to state | GHZ⟩ 1, while if detector D2 clicks at time (L0+ L1+ L)/c, then the system state will collapse to state | GHZ⟩ 2. The other six cases and the corresponding GHZ states could be achieved analogously.
We now describe the process for four-qubit W state generation between two nodes. The sketch is shown in Fig. 4. The control sequences of optical switches are shown in Fig. 3. Assuming that the initial photon pulse is in state | R⟩ and the four NV centers are all in state | + ⟩ . Thus, the initial system state could be represented as
which would be transformed to
after HWP1, NV1, PBS1, HWP2, and NV2 work. After the photon pulse passes through the time-bin encoding apparatus and optical switches S1 and S2, the system state will evolve to
Suppose the influence on photon pulse induced by noisy channel is same as Eq. (9), and would change system state to
Then from Fig. 4, the system has the following process
The component of the photon pulse in path d would go through an extra delay line LD and the state would be transformed to
The system state will have the following evolving process after the photon pulse interacts with the rest devices
It is easy to see that the polarization noise and phase noise have no influence on the state generation. If detector D2 clicks at time (L1+ L+ L0)/c or (L1+ L+ L0+ LD)/c, then the state of the four NV centers would both collapse to W state determinately.
Next, we shift our attention to illustrate the principle for generalizing the above proposals to n-qubit among two nodes or n (n !⩾ 3) nodes. For the case of n (n ⩾ 3) nodes, the proposals are robust for the regular changeable noise. This merit enables our proposals to have broadened applications in large scale QIP through noisy channels.
In this subsection, we illustrate the process of n-qubit GHZ state generation among n nodes, which are schematically drawn in Fig. 5. The case to generate n-qubit GHZ states between two nodes can be easily derived by following the process in Subsection 3.1 by adding NV centers behind NV 2 or NV 4 in Fig. 2. Thus, we will focus on describing the n nodes case in this subsection.
We first consider the 3-node case, based on which the n-node case can be understood and derived. The initial state is
where
We now discuss the n-node (n > 3) case. For simplicity, we assume that the noises in different channels are the same and can be expressed as | R⟩ → eiφ μ | R⟩ + eiω ν | L⟩ . Initially, suppose that the photon pulse is in state
From the last term in Eq. (25), one can see that if D4 clicks at the time [(n− 1)L0+ (n− 1)L1+ (n− 1)L+ (2n− 1− 1)LD]/c, then the n-qubit GHZ state
The case of the n-qubit W state is schematically drawn in Fig. 8. The case to generate a n-qubit W state between two nodes can be derived similarly according to Subsection 3.2 and Fig. 8(a). Thus, we focus on the case among n nodes. For simplicity, we first consider the three-node case, based on which the n-node case can be deduced. The initial state of the whole system is | ψ ⟩ 1 = | R⟩ | + ⟩ 1| + ⟩ 2| + ⟩ 3. Assuming that the influence of noisy channels is the same as in Subsection 4.1, then we can have the following process
From Eq. (26), we can see that the steady noise caused by channel 1 and the regular changeable noise caused by channel 2 would influence the time of state generation but have no effect on the success probability and fidelity. The n-node case can also be derived in a similar way based on the sketch in Fig. 8(b) and the control sequences of the optical switches in Fig. 7.
We now discuss the n-node (n > 3) case. For simplicity, supposing that the noises in different channels are the same and can be expressed as | R⟩ → eiφ μ | R⟩ + eiω ν | L⟩ . Initially, the photon pulse is in state | R⟩ , and each of the NV centers is in state | + ⟩ . According to Fig. 7 and Fig. 8, the process of state evolution could be written as
From Eq. (27), one can see that the n-qubit W state can be generated if detector D2 clicks at any one of the times [(n– 1)L0+ (n– 1)L1+ (n– 1)L+ ℓ LD]/c (ℓ is an integer which belongs to the interval [0, (2n– 1– 1)]). Moreover, the noises have no influence on state generation since the overall phase could be eliminated.
The feasibility of our proposals mainly depends on the time-bin encoding and the performance of our NV center, it is thus essential to discuss the experimental feasibility of these two primary components. The accuracy of the time is seriously dependent on the length of the delay lines. Thus, the delay line lengths L0, L1, and LD are critical in our proposals. The time Δ t, Δ t1, and Δ t2 should be big enough to make sure that the status of optical switches has enough time to be changed, which can be achieved by adjusting the lengths of L0, L1, and LD. Meanwhile, the time interval Δ t should be less than L/c to ensure that the noisy channel has a steady influence on the photon pulse with arriving time inside the red-dotted-line-formed rectangle, as shown in Fig. 7. In addition, LD should have a similar value as L1 to ensure L0+ LD> L1, this is the prerequisite condition to unmix the pulses with different noises and arriving times. In addition, The reason why we have not considered the interaction time between photon pulses and NV centers is that it has a smaller level. Sometimes it may even be bigger than (L1– L0)/c. Since the | L⟩ component of photon pulse with different noises and arriving time will inevitably interact with the NV centers, we can add extra delay lines in the paths which | R⟩ component passes through to make the two paths balanced. The only influence of these balanced paths is to shift the control sequences in Fig. 7 to the right for some definite distance, this is the reason why we have not considered the interaction time. Experimentally, time-bin encoding has been widely researched and demonstrated. Brendel et al. have demonstrated the time-bin encoding with time difference (L1– L0)/c in ns orders of magnitude.[42, 43] Recently, Wang et al. have realized the high-speed tomography of time-bin-entangled photons with the time difference 21.3 ns[44] and Takeda et al. have demonstrated it in a time difference 242 ns.[45] Besides, time-bin encoding has been verified as interacting with semiconductor quantum wells, [46] quantum dots, [47– 49] collective atomic excitation in a rare-earth-doped solid, [50] and it has been realized in other QIP tasks.[45, 51] Furthermore, Humphreys et al. have recently demonstrated a robust heralded controlled-phase gate with higher fidelity by using single-spatial time-bin encoding.[52] Donohue et al. have recently reported an effective technique using a nonlinear interaction between chirped entangled time-bin photons and shaped laser pulses to perform projective measurements on arbitrary time-bin states with ps-scale separations to readout the information of time-bin encoding.[53] That is, recent advances in Refs. [52] and Ref. [53] have provided powerful supporting for time-bin-encoding-based QIP tasks.
On the other hand, many efforts have been paid to investigate the performance of the NV centers and the coupled resonators. Although the ideal coupling strength between the NV center and optical resonator is supposed to be on the order of hundreds of megahertz, it is inevitably diminished due to some out-out-control effects.[75] Research on composite microcavity systems where NV centers in bulk diamond couple evanescently to optical resonators[65– 67] makes it possible to retain the excellent properties of NV centers. The experiment in Ref. [66] provided the datum (g, κ , γ )/2π ≈ (0.3, 26, 0.0004) with a relative low-Q value 104. This datum is very close to the requirement of our proposal
The efficiency of our proposals is defined as the probability of the photons to be detected after the generation process. Thus, similar to the scheme proposed in Ref. [76], we use
In conclusion, we have designed theoretical proposals to generate GHZ and W states of NV centers coupled to microresonator through noisy channels with the help of time-bin encoding processes and optical switches. The polarization and phase noise induced by noisy channels generally affect the time of state generation but not its success probability and fidelity, which is a big progress compared with the success probability 1/2 in Ref. [41]. The proposals can also be generalized to n-qubit between two nodes or among n nodes with n – 1 different noisy channels. In addition, this proposal is robust for the regular changeable noise. For the n-qubit state generation, we here only consider the two extreme cases, i.e. two-node case and n-node case. In fact, the n-qubit state can also be generated in more than two and less than n nodes if the operators place more than one NV center in any node. In other words, our proposals provide an alternative avenue for remote large scale QIP through noisy channels via time-bin encoding.
1 |
|
2 |
|
3 |
|
4 |
|
5 |
|
6 |
|
7 |
|
8 |
|
9 |
|
10 |
|
11 |
|
12 |
|
13 |
|
14 |
|
15 |
|
16 |
|
17 |
|
18 |
|
19 |
|
20 |
|
21 |
|
22 |
|
23 |
|
24 |
|
25 |
|
26 |
|
27 |
|
28 |
|
29 |
|
30 |
|
31 |
|
32 |
|
33 |
|
34 |
|
35 |
|
36 |
|
37 |
|
38 |
|
39 |
|
40 |
|
41 |
|
42 |
|
43 |
|
44 |
|
45 |
|
46 |
|
47 |
|
48 |
|
49 |
|
50 |
|
51 |
|
52 |
|
53 |
|
54 |
|
55 |
|
56 |
|
57 |
|
58 |
|
59 |
|
60 |
|
61 |
|
62 |
|
63 |
|
64 |
|
65 |
|
66 |
|
67 |
|
68 |
|
69 |
|
70 |
|
71 |
|
72 |
|
73 |
|
74 |
|
75 |
|
76 |
|