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

Optical nonreciprocal response and conversion in a Tavis-Cummings coupling optomechanical system

Yang Jiao Department of Physics, College of Science, Yanbian University, Yanji, Jilin 133002, China    Cheng-Hua Bai School of Physics, Harbin Institute of Technology, Harbin, Heilongjiang 150001, China    Dong-Yang Wang School of Physics, Harbin Institute of Technology, Harbin, Heilongjiang 150001, China    Shou Zhang111E-mail: [email protected] Department of Physics, College of Science, Yanbian University, Yanji, Jilin 133002, China    Hong-Fu Wang222E-mail: [email protected] Department of Physics, College of Science, Yanbian University, Yanji, Jilin 133002, China
Abstract

We propose a scheme to realize optical nonreciprocal response and conversion in a Tavis-Cummings coupling optomechanical system, where a single cavity mode interacts with the vibrational mode of a flexible membrane with an embedded ensemble of two-level quantum emitters. Due to the introduction of the Tavis-Cummings interaction, we find that the phases between the mechanical mode and the optical mode, as well as between the mechanical mode and the dopant mode, are correlated with each other, and further give the analytical relationship between them. By optimizing the system parameters, especially the relative phase between two paths, the optimal nonreciprocal response can be achieved. Under the frequency domain, we derive the transmission matrix of the system analytically based on the input-output relation and study the influence of the system parameters on the nonreciprocal response of the quantum input signal. Moreover, compared with the conventional optomechanical systems, the Tavis-Cummings coupling optomechanical system exhibits richer nonreciprocal conversion phenomena among the optical mode, mechanical mode, and dopant mode, which provide a new applicable way of achieving the phonon-photon transducer and the optomechanical circulator in future practice.

Tavis-Cummings interaction, nonreciprocal response, phonon-photon transducer, optomechanical circulator
pacs:
42.25.Bs, 42.50.Wk, 42.50.Ex

I Introduction

In recent years, cavity optomechanics Aspelmeyer2014 ; Physics.2.40 ; Science.321.1172 (COM) has attracted a lot of attention due to its theoretical and experimental rapid development, which has produced many interesting phenomena, e.g., the detection of gravitational waves 2018AAPPS , mechanical squeezing Bai2019a ; 2008APL92133102 ; Han2019 ; 2008NP4785 ; Bai2019 ; 2016SR638559 , quantum entanglement PRL10 ; PRA16 , mechanical cooling Liu2018 ; PRL21 ; Wang2018a , nonclassical correlations between single photon and phonon N22 , photon blockade Wang2019 , and coherent wavelength conversion S23 ; N24 . The common optomechanical systems study the two-body interaction between the cavity field and mechanical resonator. A novel interaction has been studied lately, which is the three-body interaction among the cavity field, mechanical resonator, and atom. Such hybrid systems are called Tavis-Cummings coupling optomechanical system, which has attracted wide attention. In the hybrid system, the cavity field interacts with mechanical resonator and few-level system simultaneously, which can lead to quantum interference effects and amazing optomechanical phenomena, for instance, EIT and EIA 2019OE277344 ; 2020OE28580 ; 2015OE2311508 , sensing 2019PR71440 ; 2019QE1e10 , quantum repeater 2012PRA85062311 , interaction with an atomic ensemble may be used to produce entanglement, optomechanical cooling, backaction evading measurements of mechanical motion 2007IJTP462550 ; Dantan2014 ; Q73 ; 2007PRA75022312 .

On the other hand, COM-based optical nonreciprocal phenomena have also attracted significant interest in the decades, which means that the transmission of signals in two opposite directions exhibits different characteristics. Optical nonreciprocity has been realized in various optomechanical structures SR43 ; OE44 ; F64 , such as magneto-optical crystals APL45 ; PRL46 ; Li2016 , optical nonlinear systems OL50 ; S51 ; QEe112019 , spatial-symmetry-breaking structures APL52 ; OE53 ; OE54 , parity-time-symmetric structures PRA55 ; Guo2014 ; 2014NP105394 . Nonreciprocal devices based on nonreciprocity, e.g., isolator, directional amplifier, circulator, etc, play significantly important roles in quantum information processing and communication. Specifically, optical isolator is a device that blocks light in one direction but allows light to pass in the opposite direction, which has already been studied in both theory 2017AO562991 and experiment OE32 ; NP27 . The fabrication of low-loss and high Q optical microcavities has achieved great progress using chemo-mechanical polish lithography 2019QE1e9 . The directional amplification has also been studied in a double-cavity optomechanical system with mechanical gain PRA33 , a superconducting microwave circuit with parametric pumps PRX34 ; PRA35 , a triple-cavity optomechanical system with optical gain PRA36 , and a three-mode optomechanical system with an additional mechanical drive OE37 . It has also been realized in an optomechanical circuit via synthetic magnetism and reservoir engineering in experiment NP38 . Circulator has been investigated in three-mode optomechanical systems PRA39 and this technique can be applied to a circuit-QED architecture PRA40 ; Qi2018 and phonon devices NJP41 .

As we all know, the nonreciprocal effect relies strongly on the relative phase between two different paths, which can be usually reduced to the phase at a single path in most theoretical researches. When the phase satisfies the condition of destructive interference, the nonreciprocal phenomenon occurs, which is similar to the optomechanically induced transparency Liu2015 ; AP2018 ; Zheng2019 . The optical nonreciprocal effect can be achieved by the momentum difference between the forward and backward moving light beams in an optomechanical system PRL58 . Li etalet~{}al. proposed a multimode system consisting of two vibrating membranes suspended inside two cavities, which can produce the tunable optical nonreciprocity and be used as a phonon-photon router PRA59 . Hafezi etalet~{}al. pointed out that the nonreciprocal transmission can be achieved via using an unidirectional optical pump in a microring resonator OE60 . We also note that the optical nonreciprocity is studied in coupled spinning optomechanical resonators OE63 . Furthermore, nonreciprocal response has been realized in experiments until now, for instance, silicon chip PRL72 , strong dispersive coupling of a high-finesse cavity to a micromechanical membrane N65 , a dispersive optomechanical system consisting of a membrane inside a cavity NJP66 , and nonlinear coupling in a low-loss system NP67 . Meanwhile, some other studies related to nonreciprocity have been reported such as fast-slow light effects OE61 and nonreciprocity photon blockade PRL71 . On the other hand, the hybrid phonon-photon systems have been studied extensively for the interesting nature due to their potential applications in information processing network 2014PRA90023817 ; Xing2018 ; 2019PRA99013804 . It has been shown that the conversion between phonons and photons can be realized and the optomechanical circulator can be engineered in the optomechanical systems NJP1 .

In this paper, we study the nonreciprocal response and conversion effects in a Tavis-Cummings coupling optomechanical system, which is composed of a trapped flexible membrane embedded an ensemble of two-level quantum emitters in an optical cavity. Resorting to the general linearization technique and the Fourier transform, we calculate the transmission matrix analytically and study the nonreciprocal response of the fluctuation signal in the frequency domain. In the present proposal, due to the introduction of the Tavis-Cummings coupling, we find that the phases of the two different paths are correlated each other and further we derive their relation analytically, which is greatly different from the previous studies. By selecting the system parameters appropriately, the prefect nonreciprocal response can be achieved. We also show that the nonreciprocity induces the feasibility to perform signal conversion among the optical mode, mechanical mode, and dopant mode, which implies that the system can be applied as a phonon-photon transducer and an optomechanical circulator. These interesting phenomena indicate that the Tavis-Cummings coupling model has potential applications in experiments, such as integrating the nonreciprocal devices.

The organization of the paper is as follows. In Sec. II, we present the Hamiltonian and dynamical equations of the Tavis-Cummings coupling optomechanical system. In Sec. III, we discuss the effect of the system parameters on nonreciprocal response in detail and show that the system can be applied as a phonon-photon transducer and an optomechanical circulator. Finally, we draw the conclusion in Sec. IV.

II Model and Hamiltonian

Refer to caption
Figure 1: (Color online) (a) Schematic of the Tavis-cummings coupling optomechanical setup, in which a cavity mode interacts with a single vibrational mode of a flexible membrane with an embedded ensemble of two-level quantum emitters that collectively behave as a single bosonic mode. (b) Depiction of the interaction of the Tavis-Cummings coupling optomechanical system given by Hamiltonian in Eq. (6). (c) The effective interactions in the linearized regime of the system Dantan2014 ; Q73 .

As depicted in Fig. 1(a), we consider a Tavis-cummings coupling optomechanical system, where an optical cavity mode interacts with a single vibrational mode of a flexible membrane with an embedded ensemble of two-level quantum emitters. Generally speaking, the operator of two-level quantum emitters is represented by the Pauli operator, here we represent the Pauli operator as a bosonic operator by a bosonic transformation. Under the conditions of sufficiently large two-level quantum emitters number NN and the limit of weak excitation, the lowering operators (raising operators) σ(σ+)\sigma_{-}~{}(\sigma_{+}) of the ensemble can be transformed to a collective bosonic operator a(a)a~{}(a^{\dagger}) in the Holstein-Primakoff representation,

σ\displaystyle\sigma_{-} =\displaystyle= aNaaNa,\displaystyle a\sqrt{N-a^{\dagger}a}\approx\sqrt{N}a, (1)
σ+\displaystyle\sigma_{+} =\displaystyle= aNaaNa,\displaystyle a^{\dagger}\sqrt{N-a^{\dagger}a}\approx\sqrt{N}a^{\dagger}, (3)
σZ\displaystyle\sigma_{Z} =\displaystyle= aaN2,\displaystyle a^{\dagger}a-\frac{N}{2}, (5)

where operators aa and aa^{{\dagger}} obey the bosonic commutation relation [a,aa,a^{{\dagger}}] = 1 Dantan2014 ; Wang2016 ; 2015PRA92033841 . The total Hamiltonian of the system is written as (\hbar = 1)

H\displaystyle H =\displaystyle= H0+Hint+Hdr,\displaystyle H_{0}+H_{\rm{int}}+H_{\rm{dr}}, (6)

with

H0\displaystyle H_{0} =\displaystyle= ωccc+ωaaa+ωmbb,\displaystyle\omega_{c}c^{{\dagger}}c+\omega_{a}a^{{\dagger}}a+\omega_{m}b^{{\dagger}}b, (7)
Hint\displaystyle H_{\rm{int}} =\displaystyle= g0cc(b+b)+[J0+μ0(b+b)](ac+ca),\displaystyle g_{0}c^{{\dagger}}c(b^{{\dagger}}+b)+[J_{0}+\mu_{0}(b^{{\dagger}}+b)](a^{{\dagger}}c+c^{{\dagger}}a), (9)
Hdr\displaystyle H_{\rm{dr}} =\displaystyle= iε(ceiωdtH.c.),\displaystyle i\varepsilon(c^{{\dagger}}e^{-i\omega_{d}t}-\mathrm{H.c.}), (11)

where H0H_{0} describes the free Hamiltonian of the system, HintH_{\rm{int}} represents the interactions of the cavity field with the mechanical membrane via radiation pressure and with the dopant (the embedded ensemble of two-level quantum emitters) via a mechanically modulated Tavis-Cummings coupling, and HdrH_{\rm{dr}} denotes the interaction between the driving field and the cavity field. Here, cc (cc^{{\dagger}}) is the annihilation (creation) operator of the cavity mode with frequency ωc\omega_{c}, aa (aa^{{\dagger}}) is the annihilation (creation) operator of the dopant mode with frequency ωa\omega_{a}, and bb (bb^{{\dagger}}) is the annihilation (creation) operator of the mechanical mode with frequency ωm\omega_{m}. g0g_{0} is the optomechanical coupling strength between the cavity mode and the mechanical mode, J0J_{0} is the coupling strength between the cavity mode and the dopant mode, and μ0\mu_{0} is the Tavis-Cummings coupling strength among the cavity mode, mechanical mode, and dopant mode. ε\varepsilon and ωd\omega_{d} are the amplitude and frequency of the driving field, respectively. In Fig. 1(b), we give the explicit schematic for describing the interaction of the proposed Tavis-Cummings coupling optomechanical system.

In the rotation frame with Hr=ωd(cc+aa)H_{r}=\omega_{d}(c^{{\dagger}}c+a^{{\dagger}}a), according to the Heisenberg equations of motion, we can get the quantum Langevin equations (QLEs)

c˙\displaystyle\dot{c} =\displaystyle= (iΔc+κ)cig0c(b+b)i[J0+μ0(b+b)]a\displaystyle-(i\Delta_{c}+\kappa)c-ig_{0}c(b^{{\dagger}}+b)-i[J_{0}+\mu_{0}(b^{{\dagger}}+b)]a (14)
+ε+2κcin,\displaystyle+\varepsilon+\sqrt{2\kappa}c_{\rm{in}},
a˙\displaystyle\dot{a} =\displaystyle= (iΔa+γ)ai[J0+μ0(b+b)]c+2γain,\displaystyle-(i\Delta_{a}+\gamma)a-i[J_{0}+\mu_{0}(b^{{\dagger}}+b)]c+\sqrt{2\gamma}a_{\rm{in}}, (16)
b˙\displaystyle\dot{b} =\displaystyle= (iωm+γm)big0cciμ0(ac+ca)\displaystyle-(i\omega_{m}+\gamma_{m})b-ig_{0}c^{{\dagger}}c-i\mu_{0}(a^{{\dagger}}c+c^{{\dagger}}a) (20)
+2γmbin,\displaystyle+\sqrt{2\gamma_{m}}b_{\rm{in}},

where Δi=ωiωd\Delta_{i}=\omega_{i}-\omega_{d} is the detuning of the respective mode (i=c,a)(i=c,a) from the driving frequency. κ\kappa (γ\gamma) is the decay rate of cavity mode cc (dopant mode aa) and γm\gamma_{m} is the mechanical damping rate. cinc_{\rm{in}}, aina_{\rm{in}}, and binb_{\rm{in}} are the input quantum noises with zero mean values. To solve the nonlinear QLEs in Eq. (14), we rewrite the operators as the sum of the mean values and the small quantum fluctuation terms, i.e. c=ξ+δcc=\xi+\delta c, a=α+δaa=\alpha+\delta a, and b=β+δbb=\beta+\delta b. Subsequently, the classical mean values and quantum fluctuations can be treated separately. The equations about steady mean values are given by

(iΔc+κ)ξi[J0+μ0(β+β)]α+ε\displaystyle-(i\Delta_{c}^{\prime}+\kappa)\xi-i[J_{0}+\mu_{0}(\beta^{*}+\beta)]\alpha+\varepsilon =\displaystyle= 0,\displaystyle 0, (21)
(iΔa+γ)αi[J0+μ0(β+β)]ξ\displaystyle-(i\Delta_{a}+\gamma)\alpha-i[J_{0}+\mu_{0}(\beta^{*}+\beta)]\xi =\displaystyle= 0,\displaystyle 0, (23)
(iωm+γm)βig0|ξ|2iμ0(αξ+ξα)\displaystyle-(i\omega_{m}+\gamma_{m})\beta-ig_{0}|{\xi}|^{2}-i\mu_{0}(\alpha^{*}\xi+\xi^{*}\alpha) =\displaystyle= 0,\displaystyle 0, (25)

where Δc=Δc+g0(β+β)\Delta_{c}^{\prime}=\Delta_{c}+g_{0}(\beta^{*}+\beta) is the effective detuning including the frequency shift caused by the mechanical interaction. The solution on α\alpha is

α=iξ[J0+μ0(β+β)](iΔa+γ).\alpha=-\frac{i\xi[J_{0}+\mu_{0}(\beta^{*}+\beta)]}{(i\Delta_{a}+\gamma)}. (26)

While the linearized fluctuations equations around the steady state can be obtained

δc˙\displaystyle\delta\dot{c} =\displaystyle= (iΔc+κ)δciJδaiG(δb+δb)+2κcin,\displaystyle-(i\Delta_{c}^{\prime}+\kappa)\delta c-iJ\delta a-iG(\delta b^{{\dagger}}+\delta b)+\sqrt{2\kappa}c_{\rm{in}}, (27)
δa˙\displaystyle\delta\dot{a} =\displaystyle= (iΔa+γ)δaiJδciμ(δb+δb)+2γain,\displaystyle-(i\Delta_{a}+\gamma)\delta a-iJ\delta c-i\mu(\delta b^{{\dagger}}+\delta b)+\sqrt{2\gamma}a_{\rm{in}}, (29)
δb˙\displaystyle\delta\dot{b} =\displaystyle= (iωm+γm)δbi(Gδc+Gδc)i(μδa+μδa)\displaystyle-(i\omega_{m}+\gamma_{m})\delta b-i(G^{*}\delta c+G\delta c^{{\dagger}})-i(\mu\delta a^{{\dagger}}+\mu^{*}\delta a) (33)
+2γmbin,\displaystyle+\sqrt{2\gamma_{m}}b_{\rm{in}},

and the linearized effective interactions of the system are clearly depicted in Fig. 1(c). Hereafter, for simplicity, we define J=J0+μ0(β+β)J=J_{0}+\mu_{0}(\beta^{*}+\beta), g=g0ξ,μ=μ0ξg=g_{0}\xi,~{}\mu=\mu_{0}\xi, G=giJμ/(iΔa+γ)G=g-iJ\mu/(i\Delta_{a}+\gamma), and specify the operators δcc\delta c\rightarrow c, δaa\delta a\rightarrow a, δbb\delta b\rightarrow b.

For convenience, the linearized QLEs in Eq. (27) can be written in the matrix form

F˙=MF+ΓFin,\dot{F}=MF+\Gamma F_{\rm{in}}, (34)

where the vectors F=[c,a,b,c,a,b]TF=[c,a,b,c^{{\dagger}},a^{{\dagger}},b^{{\dagger}}]^{T} and Fin=[cin,ain,bin,cin,ain,bin]TF_{\rm{in}}=[c_{\rm{in}},a_{\rm{in}},b_{\rm{in}},c_{\rm{in}}^{{\dagger}},a_{\rm{in}}^{{\dagger}},b_{\rm{in}}^{{\dagger}}]^{T} (TT represents the transpose operator), the matrix Γ\Gamma= Diag [2κ,2γ,2γm,2κ,2γ,2γm,][\sqrt{2\kappa},\sqrt{2\gamma},\sqrt{2\gamma_{m}},\sqrt{2\kappa},\sqrt{2\gamma},\sqrt{2\gamma_{m}},], and the coefficient matrix

M=((iΔc+κ)iJiG00iGiJ(iΔa+γ)iμ00iμiGiμ(iωm+γm)iGiμ000iG(κiΔc)iJiG00iμiJ(γiΔa)iμiGiμ0iGiμ(γmiωm)).M=\left(\begin{array}[]{cccccc}-(i\Delta_{c}^{\prime}+\kappa)&-iJ&-iG&0&0&-iG\\ -iJ&-(i\Delta_{a}+\gamma)&-i\mu&0&0&-i\mu\\ -iG^{*}&-i\mu^{*}&-(i\omega_{m}+\gamma_{m})&-iG&-i\mu&0\\ 0&0&iG^{*}&-(\kappa-i\Delta_{c}^{\prime})&iJ&iG^{*}\\ 0&0&i\mu^{*}&iJ&-(\gamma-i\Delta_{a})&i\mu^{*}\\ iG^{*}&i\mu^{*}&0&iG&i\mu&-(\gamma_{m}-i\omega_{m})\\ \end{array}\right). (35)

The system is stable only if the real parts of all the eigenvalues of matrix MM are negative. The stability condition can be derived by applying the Routh-Hurwitz criterion PRA74 , however, whose concrete form is too cumbersome to give here. In the following, all the parameters in the present work have been chosen to make sure the stability of the system.

Generally, it is more convenient to reveal the nonreciprocity in the frequency domain. To this end, we perform the Fourier transformations of the operators

o(ω)\displaystyle o(\omega) =\displaystyle= 12π+o(t)eiωt𝑑t,\displaystyle\frac{1}{2\pi}\int_{-\infty}^{+\infty}o(t)e^{i\omega t}\,dt, (36)
o(ω)\displaystyle o^{{\dagger}}(\omega) =\displaystyle= 12π+o(t)eiωt𝑑t,\displaystyle\frac{1}{2\pi}\int_{-\infty}^{+\infty}o^{{\dagger}}(t)e^{i\omega t}\,dt, (38)

and then Eq. (34) in the frequency domain becomes

F(ω)=(M+iωI)1ΓFin(ω),F(\omega)=-(M+i\omega I)^{-1}\Gamma F_{\rm{in}}(\omega), (39)

where II represents the identity matrix. By substituting Eq. (39) into the standard input-output relation oin+oout=κooo_{\rm{in}}+o_{\rm{out}}=\sqrt{\kappa_{o}}o, where oino_{\rm{in}}, oouto_{\rm{out}}, and κo\kappa_{o} denote input operators, output operators, and their corresponding damping rates, respectively. Then the output field vector in the frequency domain is given by

Fout(ω)=U(ω)Fin(ω),F_{\rm{out}}(\omega)=U(\omega)F_{\rm{in}}(\omega), (40)

where the output field vector Fout(ω)=[cout,aout,bout,cout,aout,bout]TF_{\rm{out}}(\omega)=[c_{\rm{out}},a_{\rm{out}},b_{\rm{out}},c_{\rm{out}}^{{\dagger}},a_{\rm{out}}^{{\dagger}},b_{\rm{out}}^{{\dagger}}]^{T} and the coefficient matrix U(ω)U(\omega) is given by

U(ω)=Γ(M+iωI)1ΓI.U(\omega)=-\Gamma(M+i\omega I)^{-1}\Gamma-I. (41)

The spectrum of the output field is defined by

soout(ω)=+𝑑ωoout(ω)oout(ω).s_{o}^{\rm{out}}(\omega)=\int_{-\infty}^{+\infty}d\omega^{\prime}\langle o_{\rm{out}}^{{\dagger}}(\omega^{\prime})o_{\rm{out}}(\omega)\rangle. (42)

Accordingly, the vector of the output field Sout(ω)S_{\rm{out}}(\omega) is written in the form of Sout(ω)=[scout(ω),saout(ω),sbout(ω)]TS_{\rm{out}}(\omega)=[s_{c}^{\rm{out}}(\omega),s_{a}^{\rm{out}}(\omega),s_{b}^{\rm{out}}(\omega)]^{T}. Substituting Eq. (40) into Eq. (42) and making use of correlation functions oin(ω)oin(ω)=soin(ω)δ(ω+ω)\langle o_{\rm{in}}^{{\dagger}}(\omega^{\prime})o_{\rm{in}}(\omega)\rangle=s_{o}^{\rm{in}}(\omega)\delta(\omega+\omega^{\prime}) and oin(ω)oin(ω)=[soin(ω)+1]δ(ω+ω)\langle o_{\rm{in}}(\omega^{\prime})o_{\rm{in}}^{{\dagger}}(\omega)\rangle=[s_{o}^{\rm{in}}(\omega)+1]\delta(\omega+\omega^{\prime}), we can obtain

Sout(ω)=T(ω)Sin(ω)+Sv(ω),S_{\rm{out}}(\omega)=T(\omega)S_{\rm{in}}(\omega)+S_{\rm{v}}(\omega), (43)

where the vector of the input field Sin(ω)=[scin(ω),sain(ω),sbin(ω)]TS_{\rm{in}}(\omega)=[s_{c}^{\rm{in}}(\omega),s_{a}^{\rm{in}}(\omega),s_{b}^{\rm{in}}(\omega)]^{T} and the transmission matrix

T(ω)=(Tcc(ω)Tca(ω)Tcb(ω)Tac(ω)Taa(ω)Tab(ω)Tbc(ω)Tba(ω)Tbb(ω)).T(\omega)=\left(\begin{array}[]{ccc}T_{cc}(\omega)&T_{ca}(\omega)&T_{cb}(\omega)\\ T_{ac}(\omega)&T_{aa}(\omega)&T_{ab}(\omega)\\ T_{bc}(\omega)&T_{ba}(\omega)&T_{bb}(\omega)\\ \end{array}\right). (44)

The element of T(ω)T(\omega) denotes as Tij(ω)T_{ij}(\omega) (i,j=c,a,b)(i,j=c,a,b), which indicates the scattering probability from the input mode jj to the output mode ii and their respective forms are

Tcc(ω)\displaystyle T_{cc}(\omega) =\displaystyle= |U11(ω)|2+|U14(ω)|2,\displaystyle|{U_{11}(\omega)}|^{2}+|{U_{14}(\omega)}|^{2}, (45)
Tca(ω)\displaystyle T_{ca}(\omega) =\displaystyle= |U12(ω)|2+|U15(ω)|2,\displaystyle|{U_{12}(\omega)}|^{2}+|{U_{15}(\omega)}|^{2}, (47)
Tcb(ω)\displaystyle T_{cb}(\omega) =\displaystyle= |U13(ω)|2+|U16(ω)|2,\displaystyle|{U_{13}(\omega)}|^{2}+|{U_{16}(\omega)}|^{2}, (49)
Tac(ω)\displaystyle T_{ac}(\omega) =\displaystyle= |U21(ω)|2+|U24(ω)|2,\displaystyle|{U_{21}(\omega)}|^{2}+|{U_{24}(\omega)}|^{2}, (51)
Taa(ω)\displaystyle T_{aa}(\omega) =\displaystyle= |U22(ω)|2+|U25(ω)|2,\displaystyle|{U_{22}(\omega)}|^{2}+|{U_{25}(\omega)}|^{2}, (53)
Tab(ω)\displaystyle T_{ab}(\omega) =\displaystyle= |U23(ω)|2+|U26(ω)|2,\displaystyle|{U_{23}(\omega)}|^{2}+|{U_{26}(\omega)}|^{2}, (55)
Tbc(ω)\displaystyle T_{bc}(\omega) =\displaystyle= |U31(ω)|2+|U34(ω)|2,\displaystyle|{U_{31}(\omega)}|^{2}+|{U_{34}(\omega)}|^{2}, (57)
Tba(ω)\displaystyle T_{ba}(\omega) =\displaystyle= |U32(ω)|2+|U35(ω)|2,\displaystyle|{U_{32}(\omega)}|^{2}+|{U_{35}(\omega)}|^{2}, (59)
Tbb(ω)\displaystyle T_{bb}(\omega) =\displaystyle= |U33(ω)|2+|U36(ω)|2,\displaystyle|{U_{33}(\omega)}|^{2}+|{U_{36}(\omega)}|^{2}, (61)

where Uvw(ω)U_{vw}(\omega) (v,w=1,,6v,w=1,...,6) is the element in the vv-th row and the ww-th column of the matrix U(ω)U(\omega) in Eq. (41). And Sv(ω)=[scv(ω),sav(ω),sbv(ω)]TS_{\rm{v}}(\omega)=[s_{c}^{\rm{v}}(\omega),s_{a}^{\rm{v}}(\omega),s_{b}^{\rm{v}}(\omega)]^{T} in Eq. (43) is the output spectrum caused by the input vacuum field and its elements are

scv(ω)\displaystyle s_{c}^{\rm{v}}(\omega) =\displaystyle= |U14(ω)|2+|U15(ω)|2+|U16(ω)|2,\displaystyle|{U_{14}(\omega)}|^{2}+|{U_{15}(\omega)}|^{2}+|{U_{16}(\omega)}|^{2}, (62)
sav(ω)\displaystyle s_{a}^{\rm{v}}(\omega) =\displaystyle= |U24(ω)|2+|U25(ω)|2+|U26(ω)|2,\displaystyle|{U_{24}(\omega)}|^{2}+|{U_{25}(\omega)}|^{2}+|{U_{26}(\omega)}|^{2}, (64)
sbv(ω)\displaystyle s_{b}^{\rm{v}}(\omega) =\displaystyle= |U34(ω)|2+|U35(ω)|2+|U36(ω)|2.\displaystyle|{U_{34}(\omega)}|^{2}+|{U_{35}(\omega)}|^{2}+|{U_{36}(\omega)}|^{2}. (66)

III RESULTS AND DISCUSSION

III.1 Nonreciprocal response

Refer to caption
Figure 2: (Color online) The relation between the relative phase θ\theta and the phase ϕ\phi. The blue solid line represents |θπ/2|/π|\theta-\pi/2|/\pi while the orange dashed line represents |θ3π/2|/π|\theta-3\pi/2|/\pi. The other parameters are Δc=Δa=ωm=10κ\Delta_{c}^{\prime}=\Delta_{a}=\omega_{m}=10\kappa and g=J=|μ|=γ=γm=κg=J=|\mu|=\gamma=\gamma_{m}=\kappa.

In this section, we turn to discuss the nonreciprocal transmission between the optical mode and the dopant mode. Based on above definition, it is easy to find that the coupling strength JJ between the cavity mode and the dopant mode is real while both the effective optomechanical coupling GG and interaction strength μ\mu are complex. Due to the existence of the Tavis-Cummings coupling (μ00\mu_{0}\neq 0), as shown in Fig. 1(c), we cannot simply determine the respective phase at the paths of BB and CC. In the following, the relation of the two phases is derived analytically. If we rewrite μ=μ0ξ=|μ|eiϕ\mu=\mu_{0}\xi=|\mu|e^{i\phi},

G=giJ|μ|(cosϕ+isinϕ)iΔa+γ=|G|eiϕ,G=g-\frac{iJ|\mu|(\cos\phi+i\sin\phi)}{i\Delta_{a}+\gamma}=|G|e^{i\phi^{\prime}}, (67)

where ϕ\phi and ϕ\phi^{\prime} are the arguments of μ\mu and GG, respectively. We specify the relative phase between the paths of BB and CC is θ=ϕϕ\theta=\phi-\phi^{\prime}. Generally, when the relative phase between the two paths is π/2\pi/2 or 3π/23\pi/2, the optimal nonreciprocal response occurs PRA39 ; PRA36 ; PRA59 . In Fig. 2, we plot the relative phase θ\theta with respect to the phase ϕ\phi, where the parameters are given as g=J=|μ|=γ=γm=κg=J=|\mu|=\gamma=\gamma_{m}=\kappa and Δa=Δc=ωm=10κ\Delta_{a}=\Delta_{c}^{\prime}=\omega_{m}=10\kappa. One can note that the optimal value of the relative phase is θπ/2=0\theta-\pi/2=0 or θ3π/2=0\theta-3\pi/2=0. Obviously, it does not appear at ϕ=π/2\phi=\pi/2 and ϕ=3π/2\phi=3\pi/2, but at ϕ=0.47π\phi=0.47\pi and ϕ=1.53π\phi=1.53\pi. Meanwhile, according to the previous calculation and analysis of the optomechanical coupling strength GG, we can find that the relative phase θ=ϕϕ\theta=\phi-\phi^{\prime} of the two different paths is related to the detuning of the dopant mode Δa\Delta_{a}, the interaction strength between the optical mode and the dopant mode JJ, and the effective optomechanical coupling gg. Therefore, we can modulate the relative phase by adjusting the corresponding system parameters Δa\Delta_{a}, JJ or gg for practical system NatureC ; PRL114093602 ; PRL105220501 . Next, we discuss the nonreciprocal response when ϕ=0.47π\phi=0.47\pi and 1.53π1.53\pi.

Refer to caption
Figure 3: (Color online) Scattering probabilities Tca(ω)T_{ca}(\omega) (blue dashed line) and Tac(ω)T_{ac}(\omega) (orange solid line) as functions of the frequency of the incoming signal ω\omega for different effective optomechanical coupling rates gg. These panels reveal the situations of two different phases: (a), (c), (e), and (g) for ϕ=0.47π\phi=0.47\pi while (b), (d), (f), and (h) for ϕ=1.53π\phi=1.53\pi. From the top panels to bottom panels, the parameter gg is chosen as 0.01κ,0.5κ,1.0κ0.01\kappa,0.5\kappa,1.0\kappa, and 1.5κ1.5\kappa, respectively. The other parameters are the same as in Fig. 2.

We now focus on how the nonreciprocal response is tuned by the optomechanical coupling strength gg. Figure 3 shows the scattering probabilities Tca(ω)T_{ca}(\omega) and Tac(ω)T_{ac}(\omega) as functions of the frequency of the incoming signal ω\omega for different coupling rates gg. It is obvious to find that the scattering probabilities of the mutually inverse processes TcaT_{ca} and TacT_{ac} reveal opposite nonreciprocal effects at two different phases. When ϕ=0.47π\phi=0.47\pi, the scattering probability from mode aa to mode cc is almost the same as the one from mode cc to mode aa [i.e., Tca(ω)Tac(ω)T_{ca}(\omega)\approx T_{ac}(\omega)] when the optomechanical coupling strength gg is weak (i.e., g=0.01κg=0.01\kappa), as shown in Fig. 3(a). The scattering probability TacT_{ac} gradually increases with the enhancement of the optomechanical coupling strength gg while the opposite direction TcaT_{ca} decreases [i.e., Tac(ω)>Tca(ω)T_{ac}(\omega)>T_{\rm{ca}}(\omega)]. Finally, the nonreciprocal response becomes more obvious and the system reaches the optimal nonreciprocal response when g=κg=\kappa at ω=ωm\omega=\omega_{m} [i.e., Tac(ω)1,Tca(ω)0T_{ac}(\omega)\approx 1,T_{\rm{ca}}(\omega)\approx 0], as shown in Fig. 3(e). However, if we keep increasing the optomechanical coupling strength gg, the nonreciprocal response gradually becomes weak. While for ϕ=1.53π\phi=1.53\pi, the system exhibits the opposite phenomena.

Refer to caption
Figure 4: (Color online) Scattering probabilities Tca(ω)T_{ca}(\omega) (blue dashed line) and Tac(ω)T_{ac}(\omega) (orange solid line) as functions of the frequency of the incoming signal ω\omega for different coupling coefficient JJ. These panels reveal the situations of two different phases: (a), (c), (e), and (g) for ϕ=0.47π\phi=0.47\pi while (b), (d), (f), and (h) for ϕ=1.53π\phi=1.53\pi. From the top panels to bottom panels, the parameter JJ is chosen as 0.01κ,0.5κ,1.0κ0.01\kappa,0.5\kappa,1.0\kappa, and 1.5κ1.5\kappa, respectively. In all subfigures, g=κg=\kappa and the other parameters are the same as in Fig. 2.

In Fig. 4, the scattering probabilities Tca(ω)T_{ca}(\omega) and Tac(ω)T_{ac}(\omega) are shown as functions of the frequency of the incoming signal ω\omega for different interaction strength JJ. It is shown that the scattering probabilities of the two mutually inverse processes TcaT_{ca} and TacT_{ac} still reveal opposite nonreciprocal effects at ϕ=0.47π\phi=0.47\pi and ϕ=1.53π\phi=1.53\pi. When JκJ\ll\kappa (e.g., J=0.01κJ=0.01\kappa), as depicted in Figs. 4(a) and 4(b), Tca(ω)Tac(ω)T_{ca}(\omega)\approx T_{\rm{ac}}(\omega). In this parameter regime, the system almost does not exhibit the nonreciprocal response. With the increase of the coupling strength JJ, the nonreciprocal response becomes distinct and the system reaches the optimal nonreciprocal effect when J=κJ=\kappa at ω=ωm\omega=\omega_{m}. As shown in Figs. 4(g) and 4(h), if further increasing coupling strength JJ, the nonreciprocal phenomenon becomes weak, which is similar with the case in Fig. 3.

Refer to caption
Figure 5: (Color online) Scattering probabilities Tca(ω)T_{ca}(\omega) (blue dashed line) and Tac(ω)T_{ac}(\omega) (orange solid line) as functions of the frequency of the incoming signal ω\omega for different mechanical damping rates γm\gamma_{m}. These panels reveal the situations of two different phases: (a), (c), (e), and (g) for ϕ=0.47π\phi=0.47\pi while (b), (d), (f), and (h) for ϕ=1.53π\phi=1.53\pi. From the top panels to bottom panels, the parameter γm\gamma_{m} is chosen as 0.01κ,0.5κ,1.0κ0.01\kappa,0.5\kappa,1.0\kappa, and 1.5κ1.5\kappa, respectively. In all subfigures, g=κg=\kappa and the other parameters are the same as in Fig. 2.

In Fig. 5, we plot the scattering probabilities Tca(ω)T_{ca}(\omega) and Tac(ω)T_{ac}(\omega) for different mechanical damping rates γm\gamma_{m} as functions of the frequency of the incoming signal ω\omega. In the case of ϕ=0.47π\phi=0.47\pi, when the mechanical damping rate γm\gamma_{m} is smaller than the decay rate of the cavity κ\kappa, as shown in Fig. 5(a), TcaTacT_{ca}\approx T_{\rm{ac}}, which indicates that there is no obvious nonreciprocity. With the increase of the mechanical damping rate γm\gamma_{m}, the nonreciprocal response becomes prominent and the system shows the optimal nonreciprocal effect at γm=κ\gamma_{m}=\kappa, as shown in Fig. 5(e). However, when ϕ=1.53π\phi=1.53\pi, contrary to the case of ϕ=0.47π\phi=0.47\pi, Tca>TacT_{ca}>T_{\rm{ac}} occurs once increasing the mechanical damping rate γm\gamma_{m} and the optimal effect corresponds to γm=κ\gamma_{m}=\kappa at ω=ωm\omega=\omega_{m}. The nonreciprocal response also becomes weak with increasing the mechanical damping rate γm\gamma_{m} continuously at two different phases.

III.2 Phonon-Photon transducer and optomechanical circulator

Refer to caption
Figure 6: (Color online) Scattering probabilities Tcb(ω)T_{cb}(\omega) (blue dashed line) and Tbc(ω)T_{bc}(\omega) (orange solid line) as functions of the frequency of the incoming signal ω\omega for different phases ϕ\phi. The two panels reveal the situations of two different phases: (a) for ϕ=0.47π\phi=0.47\pi while (b) for ϕ=1.53π\phi=1.53\pi. The other parameters are the same as in Fig. 2.
Refer to caption
Figure 7: (Color online) Scattering probabilities Ttc(ω)T_{tc}(\omega), Tta(ω)T_{ta}(\omega), and Ttb(ω)T_{tb}(\omega) (t=c,at=c,~{}a, and bb) as functions of the frequency of the incoming signal ω\omega for different phases ϕ\phi. These panels reveal the situations of two different phases: (a), (c), and (e) for ϕ=0.47π\phi=0.47\pi while (b), (d), and (f) for ϕ=1.53π\phi=1.53\pi. The other parameters are the same as in Fig. 2.

In this section, we show how the proposed Tavis-Cummings coupling optomechanical system can be applied as a photon-phonon transducer and optomechanical circulator. Based on the scattering probabilities Tcb(ω)T_{cb}(\omega) and Tbc(ω)T_{bc}(\omega) in Eq. (45) and the above parameters for producing the optimal nonreciprocal response, as shown in Fig. 6, we obtain the phonon-photon conversion rates Tcb(ω)T_{cb}(\omega) and Tbc(ω)T_{bc}(\omega). It is obvious that the proposed system can realize the nonreciprocal conversion between the optical and mechanical modes, which indicates that the present Tavis-Cummings coupling optomechanical system can be viewed as a phonon-photon transducer and its converted processes can be reversed by changing the phase of the interaction strength.

Figure 7 depicts the scattering probabilities Ttc(ω)T_{tc}(\omega), Tta(ω)T_{ta}(\omega), and Ttb(ω)T_{tb}(\omega) (t=c,a,b)(t=c,a,b) as functions of the frequency of the incoming signal ω\omega for different phases based on the optimal nonreciprocal conditions. When ϕ=0.47π\phi=0.47\pi, one can notice that Tac(ω)Tba(ω)Tcb(ω)1T_{ac}(\omega)\approx T_{\rm{ba}}(\omega)\approx T_{\rm{cb}}(\omega)\approx 1 but the other scattering probabilities are approximately equal to zero at ω=ωm\omega=\omega_{m}, as shown in Figs. 7(a), 7(c), and 7(e). However, as shown in Figs. 7(b), 7(d), and 7(f), once ϕ=1.53π\phi=1.53\pi, Tbc(ω)Tca(ω)Tab(ω)1T_{bc}(\omega)\approx T_{\rm{ca}}(\omega)\approx T_{\rm{ab}}(\omega)\approx 1 while the others are approximately equal to zero at ω=ωm\omega=\omega_{m}. This indicates that the signal is transferred from one mode to another along a clockwise direction cabcc\rightarrow a\rightarrow b\rightarrow c or counterclockwise direction cbacc\rightarrow b\rightarrow a\rightarrow c, which depends on the phase ϕ=0.47π\phi=0.47\pi or ϕ=1.53π\phi=1.53\pi. Therefore, the proposed Tavis-Cummings coupling optomechanical system also exhibits the optomechanical circulator behavior and the circulator direction is determined by the phase ϕ\phi.

IV Conclusions

In conclusion, we have studied the nonreciprocal response effect and signal conversion in a Tavis-Cummings coupling optomechanical system. Using the general linearization technique and the Fourier transform, we analytically calculate the transmission matrix of the quantum input signal in the frequency domain. Different from the previous schemes, we find that, due to the introduction of the Tavis-Cummings interaction, both the effective optomechanical coupling strength and the interaction strength between the dopant mode and the mechanical mode are complex and their phases are correlated each other. To obtain the optimal nonreciprocal response, we analytically get the relation of the two phases and derive the relative phase between the two different paths. By selecting the suitable system parameters, especially the relative phase, the optimal nonreciprocal response can be effectively achieved. Furthermore, based on this interesting property, we further discuss the nonreciprocal conversion phenomena among the optical mode, mechanical mode, and dopant mode, which can be used as a phonon-photon transducer and an optomechanical circulator. We hope that the proposed proposal would have potential applications in quantum information processing network.

ACKNOWLEDGMENTS

This work was supported by the National Natural Science Foundation of China under Grant Nos. 61822114, 61465013, and 11465020.

References

  • (1) M. Aspelmeyer, T. J. Kippenberg, and F. Marquardt, Rev. Mod. Phys. 86, 1391–1452 (2014).
  • (2) F. Marquardt and S. M. Girvin, Physics 2, 40 (2009).
  • (3) T. J. Kippenberg and K. J. Vahala, Science 321, 1172–1176 (2008).
  • (4) G. Kang, AAPPS Bulletin 28, 25–29 (2018).
  • (5) C. H. Bai, D. Y. Wang, S. Zhang, and H. F. Wang, Sci. China-Phys. Mech. Astron. 62, 970311 (2019).
  • (6) W. Y. Huo and G. L. Long, Appl. Phys. Lett. 92, 133102 (2008).
  • (7) X. Han, D. Y. Wang, C. H. Bai, W. X. Cui, S. Zhang, and H. F. Wang, Phys. Rev. A 100, 033812 (2019).
  • (8) S. Etaki, M. Poot, I. Mahboob, K. Onomitsu, H. Yamaguchi and H. S. J. van der Zant, Nat. Phys. 4, 785–788 (2008).
  • (9) C. H. Bai, D. Y. Wang, S. Zhang, S. T. Liu, and H. F. Wang, Photon. Res. 7, 1229–1239 (2019).
  • (10) D. Y. Wang, C. H. Bai, H. F. Wang, A. D. Zhu, and S. Zhang, Sci. Rep. 6, 38559 (2016).
  • (11) G. L. Wang, L. Huang, Y. C. Lai, and C. Grebogi, Phys. Rev. Lett. 112, 110406 (2014).
  • (12) R. X. Chen, L. T. Shen, Z. B. Yang, H. Z. Wu, and S. B. Zheng, Phys. Rev. A 89, 023843 (2014).
  • (13) Y. M. Liu, C. H. Bai, D. Y. Wang, T. Wang, M. H. Zheng, H. F. Wang, A. D. Zhu, and S. Zhang, Opt. Express 26, 6143–6157 (2018).
  • (14) R. W. Peterson, T. P. Purdy, N. S. Kampel, R. W. Andrews, P. Yu, K. W. Lehnert, and C. A. Regal, Phys. Rev. Lett. 116, 063601 (2016).
  • (15) D. Y. Wang, C. H. Bai, S. T. Liu, S. Zhang, and H. F. Wang, Phys. Rev. A 98, 023816 (2018).
  • (16) R. Riedinger, S. Hong, R. A. Norte, J. A. Slater, J. Shang, A. G. Krause, V. Anant, M. Aspelmeyer, and S. Gröblacher, Nature (London) 530, 313 (2016).
  • (17) D. Y. Wang, C. H. Bai, S. T. Liu, S. Zhang, and H. F. Wang, Phys. Rev. A 99, 043818 (2019).
  • (18) S. Hong, R. Riedinger, I. Marinković, A. Wallucks, S. G. Hofer, R. A. Norte, M. Aspelmeyer, and S. Gröblacher, Science 358, 203–206 (2017).
  • (19) J. T. Hill, A. H. Safavi Naeini, J. Chan, and O. Painter, Nat. Commun. 3, 1196 (2012).
  • (20) T. Wang, Y. Hu, C. Du, and G. Long, Opt. Express 27, 7344–7353 (2019).
  • (21) G. Qin, H. Yang, X. Mao, J. Wen, M. Wang, D. Ruan, and G. Long, Opt. Express 28, 580–592 (2020).
  • (22) F. Lei, M. Gao, C. Du, Q. Jing, and G. Long, Opt. Express 23, 11508–11517 (2015).
  • (23) G. Qin, M. Wang, J. Wen, D. Ruan, and G. Long, Photon. Res. 7, 1440–1446 (2019).
  • (24) D. E. Liu, Quantum Eng. 1, e10 (2019).
  • (25) T. J. Wang, S. Y. Song, and G. L. Long, Phys. Rev. A 85, 062311 (2012).
  • (26) J. M. Ma, Z. Y. Jiao, and N. Li, Int. J. Theor. Phys. 46, 2550–2559 (2007).
  • (27) A. Dantan, B. Nair, G. Pupillo, and C. Genes, Phys. Rev. A 90, 033820 (2014).
  • (28) O. Černotík, C. Genes, and A. Dantan, Quantum Sci. Technol. 4, 024002 (2019).
  • (29) A. Retzker, E. Solano, and B. Reznik, Phys. Rev. A 75, 022312 (2007).
  • (30) Z. Q. Wang, L. Shi, Y. Liu, X. B. Xu, and X. L. Zhang, Sci. Rep. 5, 8657 (2015).
  • (31) H. Q. Qiu, J. J. Dong, L. Liu, and X. L. Zhang, Opt. Express 25, 8975–8985 (2017).
  • (32) X. B. Yan, H. L. Lu, F. Gao, and L. Yang, Front. Phys. 14, 52601 (2019).
  • (33) J. Fujita, M. Levy, R. M. Osgood, L. Wilkens, and H. Dötsch, Appl. Phys. Lett. 76, 2158–2160 (2000).
  • (34) F. D. M. Haldane and S. Raghu, Phys. Rev. Lett. 100, 013904 (2008).
  • (35) Y. Li, J. Liu, L. Cao, and Q. Liu, Sci. China-Technol. Sci. 59, 1899–1910 (2016).
  • (36) M. Soljačić, C. Luo, J. D. Joannopoulos, and S. Fan, Opt. Lett. 28, 637–639 (2003).
  • (37) L. Fan, J. Wang, L. T. Varghese, H. Shen, B. Niu, Y. Xuan, A. M. Weiner, and M. Qi, Science 335, 447–450 (2012).
  • (38) R. Osellame, Quantum Eng. 1, e11 (2019).
  • (39) A. E. Miroshnichenko, E. Brasselet, and Y. S. Kivshar, Appl. Phys. Lett. 96, 063302 (2010).
  • (40) C. Wang, C. Z. Zhou, and Z. Y. Li, Opt. Express 19, 26948–26955 (2011).
  • (41) K. Xia, M. Alamri, and M. S. Zubairy, Opt. Express 21, 25619–25631 (2013).
  • (42) H. Ramezani, T. Kottos, R. ElGanainy, and D. N. Christodoulides, Phys. Rev. A 82, 043803 (2010).
  • (43) Z. Guo, H. Cao, and L. Lu, Sci. China-Phys. Mech. Astron. 57, 1835–1839 (2014).
  • (44) B. Peng, Ş. Özdemir, F. Lei, F. Monifi, M. Gianfreda, G. Long, S. Fan, F. Nori, C M. Bender and L. Yang, Nat. Phys. 10, 394–398 (2014).
  • (45) A. Rangelov and S. Longhi, Appl. Opt. 56, 2991–2994 (2017).
  • (46) Y. Shoji, M. Ito, Y. Shirato, and T. Mizumoto, Opt. Express 20, 18440–18448 (2012).
  • (47) L. Bi, J. J. Hu, P. Jiang, D. H. Kim, G. F. Dionne, L. C. Kimerling, and C. A. Ross, Nat. Photonics 5, 758 (2011).
  • (48) M. Wang, R. Wu, J. Lin, J. Zhang, Z. Fang, Z. Chai and Y. Cheng, Quantum Eng. 1, e9 (2019).
  • (49) X. Z. Zhang, L. Tian, and Y. Li, Phys. Rev. A 97, 043818 (2018).
  • (50) K. M. Sliwa, M. Hatridge, A. Narla, S. Shankar, L. Frunzio, R. J. Schoelkopf, and M. H. Devoret, Phys. Rev. X 5, 041020 (2015).
  • (51) F. Lecocq, L. Ranzani, G. A. Peterson, K. Cicak, R. W. Simmonds, J. D. Teufel, and J. Aumentado, Phys. Rev. Applied 7, 024028 (2017).
  • (52) C. Jiang, L. N. Song, and Y. Li, Phys. Rev. A 97, 053812 (2018).
  • (53) Y. Li, Y. Y. Huang, X. Z. Zhang, and L. Tian, Opt. Express 25, 18907–18916 (2017).
  • (54) K. J. Fang, J. Luo, A. Metelmann, M. H. Matheny, F. Marquardt, A. A. Clerk, and O. Painter, Nat. Phys. 13, 465 (2017).
  • (55) X. W. Xu and Y. Li, Phys. Rev. A 91, 053854 (2015).
  • (56) J. Koch, A. A. Houck, K. L. Hur, and S. M. Girvin, Phys. Rev. A 82, 043811 (2010).
  • (57) L. Qi, Y. Xing, J. Cao, X. X. Jiang, C. S. An, A. D. Zhu, S. Zhang, and H. F. Wang, Sci. China-Phys. Mech. Astron. 61, 080313 (2018).
  • (58) S. J. M. Habraken, K. Stannigel, M. D. Lukin, P. Zoller, and P. Rabl, New J. Phys. 14, 115004 (2012).
  • (59) Y. C. Liu, Y. F. Xiao, X. Luan, and C. W. Wong, Sci. China-Phys. Mech. Astron. 58, 1–6 (2015).
  • (60) T. Wang, M. H. Zheng, C. H. Bai, D. Y. Wang, A. D. Zhu, H. F. Wang, and S. Zhang, Ann. Phys. (Berlin) 530, 1800228 (2018).
  • (61) M. H. Zheng, T. Wang, D. Y. Wang, C. H. Bai, S. Zhang, C. S. An, and H. F. Wang, Sci. China-Phys. Mech. Astron. 62, 950311 (2019).
  • (62) S. Manipatruni, J. T. Robinson, and M. Lipson, Phys. Rev. Lett. 102, 213903 (2009).
  • (63) G. L. Li, X. Xiao, Y. Li, and X. G. Wang, Phys. Rev. A 97, 023801 (2018).
  • (64) M. Hafezi and P. Rabl, Opt. Express 20, 7672–7684 (2012).
  • (65) I. M. Mirza, W. Ge, and H. Jing, Opt. Express 27, 25515–25530 (2019).
  • (66) H. Lira, Z. Yu, S. Fan, and M. Lipson, Phys. Rev. Lett. 109, 033901 (2012).
  • (67) J. D. Thompson, B. M. Zwickl, A. M. Jayich, F. Marquardt, S. M. Girvin, and J. G. E. Harris, Nature (London) 452, 72 (2008).
  • (68) A. M. Jayich, J. C. Sankey, B. M. Zwickl, C. Yang, J. D. Thompson, S. M. Girvin, A. A. Clerk, F. Marquardt, and J. G. E. Harris, New J. Phys. 10, 095008 (2008).
  • (69) J. C. Sankey, C. Yang, B. M. Zwickl, A. M. Jayich, and J. G. E. Harris, Nat. Phys. 6, 707 (2010).
  • (70) J. H. Liu, Y. F. Yu, and Z. M. Zhang, Opt. Express 27, 15382–15390 (2019).
  • (71) R. Huang, A. Miranowicz, J. Q. Liao, F. Nori, and H. Jing, Phys. Rev. Lett. 121, 153601 (2018).
  • (72) H. Wang, X. Gu, Y. X. Liu, A. Miranowicz, and F. Nori, Phys. Rev. A 90, 023817 (2014).
  • (73) Y. Xing, L. Qi, J. Cao, D. Y. Wang, C. H. Bai, W. X. Cui, H. F. Wang, A. D. Zhu, and S. Zhang, Opt. Express 26, 16250–16264 (2018).
  • (74) L. L. Zheng, T. S. Yin, Q. Bin, X. Y. Lü, and Y. Wu, Phys. Rev. A 99, 013804 (2019).
  • (75) A. H. S. Naeini and O. Painter, New J. Phys. 13, 013017 (2011).
  • (76) D. Y. Wang, C. H. Bai, H. F. Wang, A. D. Zhu, and S. Zhang, Sci. Rep. 6, 24421 (2016).
  • (77) X. Chen, Y. C. Liu, P. Peng, Y. Zhi and Y. F. Xiao, Phys. Rev. A 92, 033841 (2015).
  • (78) E. X. DeJesus and C. Kaufman, Phys. Rev. A 35, 5288–5290 (1987).
  • (79) M. J. Weaver, F. Buters, F. Luna, H. Eerkens, K. Heeck, S. D. Man and D. Bouwmeester, Nat Commun 8, 824(2017).
  • (80) X. Y. Lu¨\ddot{\rm{u}}, Y. Wu, J. R. Johansson, H. Jing, J. Zhang and F. Nori, Phys. Rev. Lett. 114, 093602(2015).
  • (81) K. Stannigel, P. Rabl, A. S. Sørensen, P. Zoller and M. Lukin, Phys. Rev. Lett. 105, 220501(2010).