Arxiv:2007.08289V2 [Quant-Ph] 5 Nov 2020 Utpoo Ussitrcigwt Utpeeitr Na in Emitters Multiple with Interacting Pulses Multiphoton ∗ Lsoi Aoieculn Oqatmeitr [28]

Home , Fock state

arXiv:2007.08289v2 [quant-ph] 5 Nov 2020 utpoo ussitrcigwt utpeeitr na in emitters multiple with interacting pulses Multiphoton ∗ lsoi aoieculn oqatmeitr [28]. the emitters and quantum 27], to [26, coupling alli- lines nanowire plasmonic transmission the microwave to along the interacting coupling trapped qubits with superconducting defects atoms [25], waveguide cold line of gator [24], photonic number dots the a quantum to as applied such be systems The can [17–23]. is optics theory which quantum waveguide-QED chiral [16] studying momentum for a de- suitable angular direction in longitudinal intrinsic modes have pendent photon also can the the waveguide field, to quasi-1D Due transverse 15]. of unique [14, confinement a transfer long- and information provides 13] quantum This [12, range physics many-body long-range. studying for be system waveg- can (1D) modes one-dimensional uide and in- the emitter-emitter by [7], The mediated detection [8–11]. teraction photon cavity atom single of [6], formation sin- sources the efficient have photon highly by can gle producing This collected in be [5]. applications efficiency can important unit emitter nearly an with dimensions of waveguide of reduced emission the field the interaction in the and enhanced emitter-photon in significantly The decades be two can . past [2–4] the studies QED realizing of in waveguide for number made large system been A have good network. quantum a scale emitters is large between certain photons, interaction waveguide via the and network studies quantum which (QED), scale electrodynamics quantum Waveguide large with channels. rela- quantum sys- system a is small quantum into these it small interconnect tems contrast, a and controls In con- build precision challenge first which high qubits. big to of system easy a number quantum tively remains large quan- single a it single a tains However from manufacture network to [1]. quantum system scale tum large a ing [email protected] h hoyo igepoo rnpr rvdsteba- the provides transport photon single of theory The build- in interest great a been has there years, recent In 2 nttt o unu cec n niern IS)adDe and (IQSE) Engineering and Science Quantum for Institute ASnmes 25.n 25.t 32.70.Jz 42.50.Ct, 42.50.Nn, numbers: theor PACS Our systems. shapes. quantum superpo pulse waveguide-based photon their of output study or the the state calculate Fock reflec also the a calculate can driv state, to emitters relations coherent input-output interacting a generalized Based of vacuum, array considered. be be an can also of can dynamics modes real-time non-guided of em the effects where the system electrodynamics waveguide-quantum a in 1 edrv eeaie atreuto o utpoo pul multiphoton for equation master generalized a derive We colo hsc,SnYtsnUiest,Gaghu51027 Guangzhou University, Yat-sen Sun Physics, of School .INTRODUCTION I. ea A Texas eagLiao Zeyang & nvriy olg tto,T 74-22 USA 77843-4242, TX Station, College University, M 1 ∗ unn Lu Yunning , 1 n .Shi Zubairy Suhail M. and , eapaewv n h eltm yaiso h emit- the of dynamics to stationary real-time assumed these the and usually In wave are plane a photons proposed. be the been formalism calculations, also SLH state have the and 62] [55–58], 60], [61, formalism [59, diagrams input-output Feynman the scat- the [54], particle multiple matrix of the tering Lehmann- decomposition 53], as function [52, than such Green method more methods reduction Alternative to Symanzik-Zimmermann generalized cumber- [48–51]. extremely is How- some becomes calculation method the [44–47]. photons, this two where occur when problem can ever, states scattering extended bound few-photon be photon-photon the can approach calculate Bethe-ansatz com- to more The much interesting becomes plicated. calculation more However provide can physics. problem transport photon single a of Bethe- dynamics [43]. on time problem based transport the photon study method to analytical approach an ansatz Dinc developed and Recently, [41] al. in- [42]. preparation et quantum sensing state quantum and quantum waveguide-based physics as applications such many-body interesting formation collective many the study to in emitters. us [40] non-identical allowed and This We situations [39] identical the [38]. multiple to emitter of approach single wavefunction a the generalized single with the a interacting applied of pulse dynamics al. photon infor- the et study quantum Chen to interesting for approach wavefunction storage. very units and important also processing are are mation emitters system dy- the real-time emitter as the ad- the and In of spectrum [37]. namics stationary switch the photon to single dition pho- and single [36] of transistor realization as ton [35], such (EIT) then occur transparency induced electromagnetic was can sys- [32–34], effects method conversion multi-level frequency interesting This and photon many 31] where . [30, tems [29] multi-emitter cou- waveguide to emitter 1D quantum extended a sin- single to a of by pled properties scattering the stationary photon with the gle together study Hamiltonian to and real-space Shen Bethe-ansatz the system. used waveguide-QED Fan the studying for sics oprdt h igepoo rbe,temulti- the problem, photon single the to Compared iiyadtasisvt fti ytm We system. this of transmissivity and tivity nb nicdn htnplewhich pulse photon incident an by en te rqec a emdltdand modulated be can frequency itter e neatn ihmlil emitters multiple with interacting ses a n motn plctosin applications important find can y nti hoy ecncluaethe calculate can we theory, this on ateto hsc n Astronomy, and Physics of partment ,tePol’ eulco China of Republic People’s the 5, iin.Mroe,w loderive also we Moreover, sitions. n-iesoa waveguide one-dimensional 2 2 ters are usually ignored. Based on Heisenberg-Langevin approach, Domokos et al. studied the coherent photon pulse scattering by a single quantum emitter in a 1D waveguide [63]. Chumak and Stolyarov investigated the propagation of few-photon pulses interacting with a sin- FIG. 1: Multiphoton wavepacket interacting with multiple gle two-level system by the method of distribution func- emitters in a 1D waveguide. tions in coordinate-momentum space [64, 65]. Kony and Gea-Banacloche generalized the wavefunction approach to study the one- and two-photon scattering by two emit- II. MULTIPHOTON SCATTERING THEORY ters coupled to a 1D waveguide [66]. In 2015, Caneva et al. used the effective Hamiltonian approach to de- In this section, we first derive a generalized master rive a master equation to calculate the emitter dynamics equation for general multiphoton pulses interacting with driven by a coherent photon pulse [67], and this method multiple emitters coupled to a 1D waveguide. Then we can be generalized to various systems [68, 69]. In Refs. derive a generalized input-output theory to calculate the [68, 69], although they mentioned that the Fock states reflection and transmission properties of the photon field. can be expressed as derivations of coherent state, but it usually requires that the system dynamics has analytical solutions. Otherwise, it is very difficult to directly A. Generalized master equation for arbitrary calculate the dynamics with the Fock state input. For photon input the continuous-mode Fock state inputs, Gheri et al. derived a mater equation to study the dynamics of a single The model we study in this paper is shown in Fig. 1. emitter driven by a single and two photon wavepackets A light pulse which may contain multiple photons is in- [70]. Baragiola et. al generalized this method to the gen- jected into a 1D waveguide and it can interact with Na eral N-photon case based on the Ito-Langevin approach emitters with arbitrary spatial distributions. The emitter [71] where they mainly focused on the scattering of a sin- positions are denoted as ~rj and their z-components are zj gle emitter system. By modeling the input pulse as the where j =1, 2, ,Na. Here, we consider a general case output of a virtual cavity with time-dependent coupling where the emitters··· can have time-modulated transition strength, the probability to generate a specific prechosen frequencies and they can couple to both the waveguide output field can be calculated from the master equation and non-waveguide photon modes. It is convenient to of a virtual cascaded system [72]. In 2018, we derived a work in the rotating frame with the original emitter fre- master equation to study the dynamics of multiple emit- quency ωa. The total Hamiltonian of the system and ters driven by continuous squeezed vacuum field in 1D reservoir fields in the rotating frame is given by waveguide [73] and found that steady-state population inversion of multiple Ξ-type emitters can occur in this ~ Na z ~ ~ system [74]. H(t)= εj (t)σj + ∆ωka† ak + ∆ω~qλ a† a~qλ 2 k ~qλ Xj=1 Xk X~qλ

Na In this article, we explicitly derive a generalized master ~ j ikzj + + (gke σj ak + H.c.) equation to study the dynamics of multiphoton pulses in Xj=1 Xk a 1D waveguide interacting with multiple emitters whose Na transition frequencies can be modulated and the effects j i~q ~rj + + ~ (g e · σ a~q + H.c.). (1) of non-guided modes are also considered. Moreover, we ~qλ j λ Xj=1 X~q also derive a generalized input-output theory to calcu- λ late the output photon pulse shapes and the reflectiv- The physical meaning of each term in the Hamiltonian ity/transmissivity of this system for various input pho- is as follows. The first term is the emitter Hamiltonian ton fields. In particular, our theory allows to study the z with time-dependent modulating frequency εj(t). σj and collective dynamics of multi-emitter systems driven by + σ (σ−) are the zth component and the raising (lowering) non-classical photon pulses which is seldom studied be- j j Pauli operators of the jth emitter. The second term is the fore. The theory developed here can thus find important Hamiltonian of the waveguide photons with the detuning applications in the researches of waveguide-based quan- frequency ∆ω = ω ω and a (a† ) is the annihilation tum system. k k a k k (creation) operator of− the waveguide photon mode with frequency ωk. When the emitter transition frequency is This article is arranged as follows. In Sec. II, we derive not very close to the photonic band edge, linear disper- a generalized master equation for the emitter dynamics sion approximation can be applied where ∆ωk = δkvg with δk = k k and v is the group velocity of the and present a generalized input-output theory to study − a g the scattering field properties. In Sec. III, we apply this guided photon pulse. The third term is the non-guided reservoir field Hamiltonian where a (a† ) is the annihi- theory to the cases of coherent state, single and general N ~qλ ~qλ photon state inputs. Finally, we summarize our results. lation (creation) operator of the non-guided photon mode 3

with frequency ω~qλ (~q is the wavevector and λ denotes Formally integrating Eqs. (3-6), we can obtain the polarization) and ∆ω~q = ω~q ωa. The fourth term λ λ i∆ωkt is the emitter-waveguide photon interaction− Hamiltonian ak(t)=ak(0)e− j ~ ~ Na t with gk = ~µj Ek(~rj )/ being the coupling strength where ′ j ikzj i∆ωk(t t) ~µ is the transition· dipole moment of the jth emitter, and i g ∗e− σ−(t′)e − dt′, (7) j − k Z j Xj=1 0 E~k(~rj ) is the strength of the electric ﬁeld with wavevec- i∆ωkt tor k at position ~rj . The last term describes the interac- ak† (t)=ak† (0)e tion between the emitters and the non-guided reservoir Na t j ′ ~ ~ ~ j ikzj + i∆ωk(t t) ﬁeld with coupling strength g~q = ~µj E~qλ (~rj )/ . is + i g e σ (t′)e− − dt′, (8) λ · k Z j the Planck constant. We should note that rotating wave Xj=1 0 approximation is applied in the Hamiltonian shown in i∆ω t a (t)=a (0)e ~qλ Eq. (1). Thus, our theory develop here is valid for the ~qλ ~qλ − Na weak and strong coupling regime, while it is invalid in t ′ j i~q ~rj i∆ω (t t) the ultrastrong coupling regime where the rotating wave i g ∗e− · σ−(t′)e ~qλ − dt′, (9) − ~qλ Z j approximation may break down. Xj=1 0 i∆ω t a† (t)=a† (0)e ~qλ ~qλ ~qλ

Na According to the Heisenberg equation, the dynamics t ′ j i~q ~rj + i∆ω (t t) of an arbitrary emitter operator O is given by + i g e · σ (t′)e− ~qλ − dt′, s ~qλ Z j Xj=1 0 (10) from which we can see that the field at time t is the interference between the incident field and the emitted N i a field by the emitters. Inserting Eqs. (7-10) into Eq. (2) O˙ (t)= ε (t)[σz(t), O (t)] S 2 j j S and using the Weisskopf-Wigner approximation we can Xj=1 obtain (see Appendix A) Na j ikzj + N + i g e [σ (t), OS (t)]ak a k j ˙ i z Xj=1 Xk OS (t)= εj [σ (t), OS (t)] 2 j Xj=1 j ikzj + g ∗e− a† [σ−(t), OS (t)] k k j Na Γj + Na + i [σj (t), OS (t)][aj (t)+ bj (t)] j i~q ~rj + r 2 + i g e · [σ (t), O (t)]a Xj=1 ~qλ j S ~qλ Xj=1 X~qλ Na Γj j i~q ~rj + i [aj†(t)+ bj†(t)][σj−(t), OS (t)] + g~q∗e− · a~q† [σj−(t), OS (t)] , (2) r 2 λ λ Xj=1 + + Λjl[σj (t), OS (t)]σl−(t) Xjl + Λ∗ σ (t)[σ−(t), O (t)], (11) − jl l j S and the dynamics of the field operators are given by Xjl

v g ∞ ikzj iδωkt where aj (t) = 2π e ak(0)e− dk describes the absorption ofq theR−∞ incident waveguide photons and vg i~qλ ~rj iδω~qλ t 3 bj(t) = 2π e · a~qλ (0)e− d ~qλ is the ab- Na q j ikzj sorption of theR incident R R nonguided photons. The collec- a˙ (t)= i∆ω a i g ∗e− σ−(t), (3) k − k k − k j Xj=1 tive interaction between the emitters can be calculated as [40] Na j ikz + † j a˙ k(t)= i∆ωk + i gke σj (t), (4) Γ Γ 3 γj γl j l ika zjl √ 2 i Xj=1 Λjl = e | | + sin φ − p 2 4 h karjl Na j i~q ~rj 2 1 i ika rjl a˙ ~q (t)= i∆ω~q a~q i g ∗e− · σ−(t), (5) + (1 3cos φ)( + ) e | |, λ − λ λ − ~qλ j 2 2 Xj=1 − (karjl) (karjl) i (12) Na j i~q ~rj + a˙ † (t)= i∆ω a† + i g e · σ (t). (6) ~qλ ~qλ ~qλ ~qλ j where the first term is the effective interaction mediated Xj=1 by the waveguide photons and the second term is the 4 usual dipole-dipole interaction induced by the non-guided be much less than the decay time of the emitter (i.e., reservoir fields. rjl = ~rj ~rl is the distance between Max(zij )/vg 1/Γi,j). Indeed, this is the usual case. | | | − | ≪ 8 8 the jth and lth emitters and zjl = ~zj ~zl is the dis- For example, if vg 10 m/s and Γ 10 Hz, we require | | | j −2 | ∼ ∼ tance in the zth direction. Γj = 4π g /vg is the de- that the largest distance between the emitters is much | ka | cay rate due to the waveguide vacuum field and γj is less than 1m which is the usual case. the spontaneous decay rate due to the nonguided photon modes. φ is the angle between the direction of the tran- From Eq. (11), we can derive a corresponding mas- sition dipole moment and the waveguide direction. In ter equation for the emitters. Since T rS+R[OS(t)ρ] = deriving Eq. (11), we have neglected the time-retarded T rS[OS ρS(t)] where ρS(t)= T rR[ρ(t)] is the emitter sys- effects which is a good approximation when the largest tem density operator, by time derivation on both sides emitter separation is not very large. To be more specific, we have T rS[OSρ˙S(t)] = T rS+R[O˙ S(t)ρ] and from Eq. the photon propagation time through the system should (11) we can obtain (see Appendix A)

Na Na Na i z Γj + Γj + ρ˙ (t)= ε (t)[σ ,ρ (t)] i [σ ,ρ′ (t)] i [σ−,ρ′†(t)] i Im(Λ )[σ σ−,ρ (t)] S − 2 j j S − r 2 j j − r 2 j j − jl j l S Xj=1 Xj=1 Xj=1 Xjl + + + Re(Λ )[σ σ−ρ (t)+ ρ (t)σ σ− 2σ−ρ (t)σ ]. (13) − jl j l S S j l − l S j Xjl

where ρj′ (t) = T rR U(t)[aj (t) + bj (t)]ρ(0)U †(t) is a repeat the above procedures to derive equations for ρj′ (t) new operator appearing{ in the equation. Here,} we until all the equations are closed. For the Fock state in- consider the case that the incident phontons are from put, Eq. (13) is reduced to the results shown in Ref. [70] the waveguide photons while the nonguided reservoir when there is only a single emitter. field is initially in the vacuum. Since a 0 = 0, ~qλ | i we have bj (t)ρ(0) = 0. Then we have ρj′ (t) = B. The generalized input-output theory T rR[U(t)aj (t)ρ(0)U †(t)] which accounts for the driving of the incident waveguide photons. This is the main equation of this section. In the previous subsections, we derive the master equations for the emitter system which allows to calculate The first term in Eq. (13) describes the modulation the real dynamics of the emitters for an arbitrary photon of the emitter transition frequencies. The second and wavepacket input. In this subsection, we derive the gen- third terms describe the excitation and deexcitation due eralized input-output relations of this system by express- to the incident photon field. The forth term describes ing the output field operators as the function of input the dipole-dipole interactions between the emitters in- operators and the system operators. Together with the duced by the guided and nonguided vacuum field. The master equations derived in the previous subsection, we last term is the collective dissipation due to the guided can then study the reflection and transmission properties and nonguided vacuum fluctuation. However, we should of this system. note that Eq. (13) itself is in general not closed because If we integrate Eq. (3) from t to tf where tf > t, we can obtain we have the new operators like ρj′ (t) and ρ′j†(t). In some special cases, Eq. (13) is closed. For example, if there is i∆ωk(tf t) not external driving field, the second and third terms dis- ak(t)=ak(tf )e − Na appear and the equation is closed from which the emitter tf ′ j ikzj i∆ωk(t t) excitation transport can be studied. Another example is + i g ∗e− σ−(t′)e − dt′. (14) k Z j that if the incident field is a coherent field or superpo- Xj=1 t sition of coherent fields, the ρ (t) and ρ †(t) terms can j′ ′j Comparing Eq. (7) with Eq. (14) it is readily to obtain then be reduced to a complex number multiplying ρS(t) and Eq. (13) becomes closed again from which the full that dynamics of the emitters driven by a coherent field can a (t )ei∆ωk(tf t) be calculated. For the coherent state input, the mas- k f − Na t ter equation shown in Eq. (13) is reduced to the results f ′ i∆ωkt j ikzj + i∆ωk(t t) shown in Ref. [67] when there is not frequency modu- =ak(0)e− i gke− σj (t′)e − dt′. − Z0 lation and the effects of non-guided modes are ignored. Xj=1 In general cases such as the Fock state input, we have to (15) 5

We can define the following input-output operators [75] the right in the whole scattering process are then given ∞ ∞ by IR = 0 r(t)dt and IL = 0 l(t)dt. Supposing that v ∞ R g i∆ωkt the photon pulse is initially propagating to the right, the ain(t)= ak(0)e− dk, (16) R R r2π Z0 reflectivity of the pulse is then given by 0 L vg i∆ωkt ain(t)= ak(0)e− dk, (17) r2π Z IL −∞ R = , (24) IR + IL R vg ∞ iδkzN i∆ωk(t tf ) aout(t)= ak(tf )e e− − dk, (18) r2π Z0 0 and the transmissivity T =1 R. L vg iδkz1 i∆ωk(t tf ) − a (t)= ak(tf )e− e− − dk, (19) The scattering power spectrum can be usually obtained out r2π Z −∞ from the two-time correlation function of the output field where z1 is the position of the left most emitter and zN operator is the position of the right most emitter. Since the right output field propagates freely after scattering by th right ∞ ∞ + iω(t1 t2) most emitter and the left output field propagates freely S(ω)= aout(t1)aout(t2) e − dt1dt2, (25) Z Z h i after scattering by the first emitter, phase factors eiδkzN 0 0 iδkz and e− 1 are added in the definitions of the right and left output operators, respectively [67]. From Eq. (15) where the average is over the initial state of the whole sys- we can obtain the generalized input-output relations (see tem. According to the generalized input-output relation Appendix B) shown in Eqs. (20) and (21), aout(t) can be expressed as the summation of the input field operator ain(t) and Na the emitter operators σj−(t). The results when ain(t) R R Γj ikazj a (t)= a (t z /v ) i e− σ−(t), out in − N g − r 2 j operator acts on the initial state can be readily worked Xj=1 out. Usually, the two-time average of the emitter opera- + (20) tors σ (t)σ−(t + τ) can be calculated from the master h j l i Na equation according to the quantum regression theorem L L Γj ikazj a (t)= a (t + z /v ) i e σ−(t), (21) [76]. However, to apply the quantum regression theorem out in 1 g − r 2 j Xj=1 to calculate the two-time correlation function, it usually requires that the reservoir field does not change signifi- where z = z z . From these two generalized Nj N − j cantly. This condition may not be very well satisfied in input-output relations we can calculate the properties the waveguide-QED system because the waveguide pho- of the scattering field of this system. We can define ton can be significantly absorbed by the emitters espe- the instant field intensity propagating to the right and cially near the resonance frequency. Therefore, direct to the left at time t by r(t) = aR+(t)aR (t) and h out out i use of the quantum regression theorem to numerically l(t)= aL+(t)aL (t) , respectively, which are given by h out out i calculate the spectrum here may cause some errors and need to be treated carefully. However, at the plane wave r(t)= aR+(t z /v )aR (t z /v ) h in − N g in − N g i limit, an alternative strategy can be used to calculate the N a Γ scattering property of the system. If the incident pho- 2 j Im[eikazj σ+(t)aR (t z /v )] − r 2 h j in − N g ton pulse has a very narrow bandwidth, we can calculate Xj=1 its reflectivity and transmissivity from the above discus- √ sions. Repeating these procedures for each incident fre- ΓiΓl ika(zj zl) + + e − σ (t)σ−(t) , (22) 2 h j l i quency, we can then obtain the scattering property of the Xjl waveguide-QED system at the plane wave limit. How- l(t)= aL+(t + z /v )aL (t + z /v ) ever, we should note that the general power spectrum h in 1 g in 1 g i Na of this system driven by a pulse with finite bandwidth Γj ikazj + L 2 Im[e− σ (t)a (t + z /v )] can not be completely captured by this frequency-sweep − r 2 h j in 1 g Xj=1 strategy and it should be calculated from Eq. (25). √ ΓiΓl ika(zl zj ) + + e − σ (t)σ−(t) . (23) 2 h j l i Xjl On the right hand side of Eqs. (22) and (23), the first III. APPLICATION TO DIFFERENT PHOTON WAVEPACKETS terms are the incident field intensities, the second terms are the absorption and stimulated emission of the system, and the last terms are the spontaneous emission of the In this section, we take the coherent states and the system. From r(t) and l(t), we can obtain the pulse shape Fock states as example to show how to apply the theory propagating to the right and to the left after the scatter- we developed in the previous section to study the emitter ing process. The field intensity reflected to the left and dynamics and the field scattering property. 6

A. Coherent state wavepacket For the right propagating incident pulse (i.e., k0 > 0), we have

We first consider the case when the incident field is a 2 − 2 n¯∆vg ∆ (zj0 vg t) coherent photon pulse. Suppose that the incident field R 2 ikazj0 i∆k(zj0 vg t) αj (t)= e− e e − . (30) is a continuous-mode coherent state describing by wave- pπ1/4 function Ψcs = Πk αk where | i | i For the left propagating incident pulse (i.e., k0 < 0), we n then have 2 ∞ k αk /2 (αk) αk = e−| | nk . (26) 2 2 | i √ | i n¯∆v ∆ (zj0+vg t) nk! L g ikazj i∆k(zj +vgt) nXk =0 α (t)= e− 2 e− 0 e− 0 , j pπ1/4 2 (31) The average photon for the kth moden ¯k = αk . | | where z = z z and ∆ = k k is the detuning Since ak Ψcs = αk Ψcs , the operator ρ′ (t) = j0 j 0 k 0 a | i | i j between the center− frequency of the| |− pulse and the emitter T r [U(t)a (t)ρ(0)U (t)] = α (t)ρ (t) where R j † j S transition frequency. The numerical results for the coherent state input are vg ∞ ikzj iδωkt αj (t)= e e− αkdk (27) shown in Fig. 2 where the coherent state is scattered r2π Z −∞ by two emitters. We assume that the distance between these two emitters is 0.125λ where λ = 2π/k . The describes the real-time evolution of the incident coherent a a a excitations of the two emitters as a function of time for photon pulse. Therefore, the operator ρ (t) is reduced to j′ two different incident average photon number (¯n = 1 and a number multiplying the system density operator ρ (t). S n¯ = 30) are shown in Fig. 2(a). When the average inci- The master equation shown in Eq. (13) then becomes dent photon number is small, e.g.n ¯ = 1, both emitters N are first excited and then deexcited as the coherent pulse i a ρ˙ (t)= ε (t)[σz ,ρ (t)] passing through. However, when the average incident S − 2 j j S Xj=1 photon number is large, e.g.n ¯ = 20, the excitations of both emitters can have multiple peaks which is the sig- Na Γj + nature of Rabi oscillations. i [α (t)σ + α∗(t)σ−,ρ (t)] − r 2 j j j j S The corresponding reflected and transmitted photon Xj=1 pulse shapes after the scattering are shown in Fig. 2(b). + + i Im(Λjl)[ρS (t), σ σ−]+ [ρS(t)], (28) When the average photon number is small, the reflected j l L Xjl pulse (thinner orange solid line) has a single peak and the transmitted pulse (thinner olive dashed line) has two + where [ρS (t)] = jl Re(Λjl)[σj σl−ρS(t) + peaks due to the interference between the incident photon + L +− ρ (t)σ σ− 2σ−ρ (t)σ ]P describes the collective wavefunction and the reemitted photon wavefunction. S j l − l S j dissipation process. Equation (28) is a general master When the average photon number is large, most pho- equation of the waveguide-QED system when the inci- tons are transmitted (thicker olive dashed line) and only dent photon pulse is in a coherent state and it reduces a very small part of the photons are reflected (thicker or- to the results shown in Ref. [69] when there is not ange solid line). This is because the pulse with large pho- frequency modulation. ton number can saturate the emitter excitation quickly The master equation shown in Eq. (28) is itself a closed and only a very small part of photons can be absorbed. equation from which we can calculate the real-time dy- Here, the reflection photon pulse can have two peaks in- namics of the emitters for arbitrary coherent pulse input. stead of one peak due to the Rabi oscillations which does Our theory can be applied to calculate the dynamics of not occur when the photon number is small. the system with arbitrary photon pulse shapes. Here, For a coherent pulse with finite time duration, the av- without loss of generality we assume that the photon erage photon number reflected by the emitters may be pulse has a Gaussian shape throughout this paper. Sup- saturated. Here, we also study the average reflected pho- posing that the incident coherent field has a Gaussian ton numbern ¯R as a function of average incident photon pulse shape with average photon numbern ¯, its spectrum numbern ¯in for two fixed pulse spectrum widths (∆ = Γ can be written as and ∆ = Γ/5 ) and the results are shown in Fig. 2(c) when the distance between the two emitters is 0.125λa. √ 2 2 n¯ (k k0) /2∆ ikz0 When the pulse width is about Γ, the average reflected αk = e− − e− , (29) π1/4√∆ photon number increases quickly first asn ¯in increases but then it increases extremely slowly whenn ¯in is large where z0 is the initial central peak position of the pulse due to the saturation effect (red line with open circles in and k0 is the wavevector corresponding to the central Fig. 2(c)). It is also noted that when the incident pho- frequency of the photon pulse. When k0 > 0 (k0 < 0) ton number is large, the average reflected photon number the pulse is propagating to the right (left). The average can be larger than two despite that there are only two photon number is given byn ¯ = n¯ = ∞ α 2dk. emitters. This is because the incident pulse is not short k k | k| P R−∞ 7

(c)

(a) 3.0

1.0 (b) 0.010 two atom: / =1 one atom: / =1

atom 1 two atom: / =5 one atom: / =5

0.10 n20 2.5

atom 2

0.8 0.008

n20 2.0 0.05

0.6 0.006 n

R 1.5

0.00

0 2 4 6

0.4 0.004

1.0

R pulse

0.5 T pulse 0.2 0.002 Excitation probability Pulse shape (arb. units)

0.0

0.000 0 50 100 150 200 0.0

0 2 4 6 8 10 0 2 4 6 8 10

t t in

FIG. 2: Coherent state pulse interacting with two emitters. (a) emitter excitation as a function of time for two different average photon number (¯n = 1 andn ¯ = 20). ∆vg /Γ = 1. (b) Reflected and transmitted pulse shapes after the scattering for the same parameters as (a). (c) The average reflected photon number as a function of average incident photon number for two different pulse width (∆/Γ=1 and ∆/Γ = 5). For comparison, the results for one emitter are also shown as the black lines with solid (∆/Γ = 5) and open circles (∆/Γ = 1) . The emitter distance d = 0.125λa for all three figures.

enough to saturate the emitters immediately. When the where αj (t) is given by Eq. (27) and we define a new S + incident pulse duration is shorter, i.e. the incident pulse operator ρ01(t)= T rR[U(t)ρS (0) 0 ΨF U (t)]. If we S S ⊗ | ih | has a broader spectrum (e.g., ∆ = 5Γ),n ¯ first increases define ρ (t)= T r [U(t)ρ (0) Ψ Ψ U †(t)], we can R 11 R ⊗ | F ih F | and then oscillates asn ¯in increases (red line with solid obtain from Eq. (13) that circles in Fig. 2(c)) due to the stimulated emission effects. The average reflected photon number is obviously Na S i z S less than 2 because the shorter pulse can saturate the ρ˙ (t)= εj (t)[σ ,ρ (t)] 11 − 2 j 11 emitters quickly. For comparison, we also plot the re- Xj=1 sults when there is only a single emitter in the system Na Γ (black lines with open and closed circles). We can see i j [α (t)σ+(t),ρS (t)] − r 2 j j 01 that their behaviors are similar but the average reflected Xj=1 photon number for two emitters is larger than that of the Na Γj S single emitter. When the pulse duration is much smaller i [α∗(t)σ−(t),ρ †(t)] − r 2 j j 01 than the decay time of the emitter, the average photon Xj=1 number being reflected by a single emitter is always less + S S i Im(Λ )[σ σ−,ρ (t)] [ρ (t)], (34) than one which can be exploited to produce single photon − jl j l 11 − L 11 sources [77]. Xjl

S + S where [ρ11(t)] = jl Re(Λjl)[σj σl−ρ11(t) + B. Single photon wavepacket S + L S + ρ11(t)σj σl− 2σl−ρ11(t)σj P] is the collective dissi- −S pation term. ρ01(t) is not a valid density matrix because Compared with the coherent state input, the calcula- S S it is traceless but it satisﬁes ρ01† = ρ10. Since a new tion of Fock state input is more involved mostly because S operator ρ01(t) appears, Eq. (34) is itself not a closed of its quantum nature. The theory developed in Sec. II equation and we need to derive an extra equation for can also be applied for the arbitrary Fock state inputs. S ρ01(t). In this subsection, we consider the simplest case where The dynamical equation for ρ (t) can be derived using the pulse only contains a single photon. Actually, for the 01 similar procedures as deriving ρS(t) shown in Sec. II and single photon pulse case, we have developed a dynamical it is given by (see Appendix C) transport theory for calculating the real-time evolution of the system based on the wavefunction approach [39, 40]. N i a Here, we instead use the master equation developed in ρ˙S (t)= ε (t)[σz,ρS (t)] 01 − 2 j j 01 Sec. II to calculate the system dynamics. Xj=1 The single photon wavepacket can be described by the Na wavefunction Γj S i α∗(t)[σ−,ρ (t)] − r 2 j j 00 ∞ Xj=1 ΨF = α(k)ak† 0 dk, (32) | i Z | i + S S −∞ i Im(Λjl)[σj σl−,ρ01(t)] [ρ01(t)], (35) 2 − − L where ∞ α(k) dk = 1. Since ak ΨF = αk 0 , the Xjl −∞ | | | i | i ρ′ (t) termR in Eq. (13) is then given by j S where ρ00(t)= T rR[U(t)ρS 0 0 U †(t)] is another den- + S ⊗ | ih | ρj′ (t)= T rR[U(t)aj (t)ρ(0)U (t)] = αj (t)ρ01(t), (33) sity matrix describing the evolution of the system when 8

(c)

10 (a) 1.0 0.5 (b) 2.0

atom 1

R pulse

atom 2 0.4 0.8

T pulse

1.5

0.3 0.6 R

1.0

0.2 0.4

0.5 0.1 0.2 Excitation probability Pulse shape (arb. units)

0.0

0.0 Reflectivity and transmissivity

0 5 10 15 20 -4 -2 0 2 4

0 5 10 15 20

t t

FIG. 3: A single photon wavepacket interacting with two emitters. (a) emitter excitation as a function of time. ∆/Γ=1. (b) Reflected and transmitted pulse shapes after the scattering for the same parameters as (a). (c) The reflectivity and transmissivity as a function of photon frequency where the symbols are the numerical results and the solid lines are the analytical results. In all three figures, d = 0.125λ. the field is initially in the vacuum. Using similar proce- bols are the numerical results and the solid lines are the dure, it is not difficult to obtain that analytical results calculated by the stationary scattering N theory such as the Bethe-ansatz approach [29]. It clearly i a ρ˙S (t)= ε (t)[σz ,ρS (t)] shows that for single photon input the results calculated 00 − 2 j j 00 Xj=1 by our theory here are consistent with those shown in the previous literatures. The spectrum shown in Fig. 3(c) + S S i Im(Λ )[σ σ−,ρ (t)] [ρ (t)], (36) − jl j l 00 − L 00 clearly shows an asymmetric Fano-like structure. This Xjl is caused by the interference between the two collective where we see that no new density operator appears. emission channels, i.e., the emission from the two col- 1 Hence, the master equation for the single photon state lective excited states = ( eg ge ) which have |±i √2 | i ± | i input consists of three cascaded equations as given by different energy shifts and decay rates. The results shown Eqs. (34-36) while only a single equation is needed in in this subsection confirm the validity of our theory de- the coherent state input. The dynamics of the emitters veloped here. for arbitrary single photon pulse input can then be calculated from these three equations. The time evolution of the average value of an arbitrary emitter operator O(t) C. N-photon wavepacket can be calculated as O(t) = T r [OρS (t)]. h i S 11 One numerical example is shown in Fig. 3 where In addition to the single photon Fock state, we can we consider a single photon wavepacket interacting with also derive generalized master equations for the multi- two emitters. Here, we assume that the single photon photon Fock state inputs. Compared with the single- wavepacket has a Gaussian spectrum as shown in Eq. photon input, the calculation of multi-photon Fock state (29) and the distance between emitters is λa/8. The input is more complicated. We first consider a relative emitter excitation as a function of time is shown in Fig. simple subset which is the direct generalization of the 3(a). Due to the collective interaction, the first emitter single photon wavepacket, i.e., can have much higher excitation probability than that of the second one and the excitation of the second emitter 1 ∞ N Nα = dkα(k)a† 0 , (37) has a Rabi-like oscillations which does not occur when the | i √N! Z k | i h −∞ i incident photon pulse is in a coherent state withn ¯in = 1. This is due to the interference between the two excitation where we have the normalization condition 2 channels, i.e., the excitation of the incident photon and ∞ αk dk = 1 [71]. A general N-photon wavepacket −∞ | | the excitation by the first excited emitter. In the coher- Rcan be always decomposed into the superposition of the ent state input, this interference is however concealed. wavefunction shown in Eq. (37) and we have The reflected and transmitted photon pulse shapes are a N = √Nα(k) N 1 . (38) shown in Fig. 3(b) from which we can see that the trans- k| αi | − αi mitted pulse has multiple peaks due to the quantum in- In general, we have the relation a m = √mα(k) m terference between the incident photon and the reemitted k α 1 and therefore | i | − photons by the two emitters. The visibility of the oscilla- αi tion is much larger than that in the coherent state input. akρs(0) mα nα = √mα(k)ρs(0) m 1α nα . (39) The emitter dynamics and the scattering pulse shapes ⊗| ih | ⊗| − ih | shown in Figs. 3(a) and 3(b) are the same as those cal- Using the similar procedures to derive Eq. (13), we culated by the wavefunction approach [39]. The reflected can derive a ladder set of dynamical equations for the and transmitted spectra when the incident single photon N-photon wavepacket input which is given by is a plane wave are shown in Fig. 3(c) where the sym- 9

Na Na S i z S Γj + S S ρ˙mn(t)= εj (t)[σj ,ρmn(t)] i √mαj (t)[σj ,ρm 1,n(t)] + √nαj∗(t)[σj−(t),ρmn 1(t)] − 2 − r 2 { − − } Xj=1 Xj=1 + S S i Im(Λ )[σ σ−,ρ (t)] [ρ (t)], (40) − jl j l mn − L mn Xjl

S + where ρmn(t) = T rR[U(t)ρS (0) m n U (t)] and 0 photon input which is shown in the previous subsection, S⊗ | ihS | ≤ m,n N. Considering that ρmn = ρnm† , (N + 1)(N + while for two-photon input we need six cascaded master 2)/2 master≤ equations are required to make the equations equations. closed where N is the total incident photon number. For Taking the two-photon input as an example, the mas- example, three master equations are needed for single- ter equations are given by

Na Na S i z S Γj + S S ρ˙ (t)= ε (t)[σ ,ρ (t)] i √2α (t)[σ ,ρ (t)] + √2α∗(t)[σ−(t),ρ (t)] 22 − 2 j j 22 − r 2 { j j 12 j j 21 } Xj=1 Xj=1 + S S i Im(Λ )[σ σ−,ρ (t)] [ρ (t)], (41) − jl j l 22 − L 22 Xjl

Na Na S i z S Γj + S S ρ˙ (t)= ε (t)[σ ,ρ (t)] i α (t)[σ ,ρ (t)] + √2α∗(t)[σ−(t),ρ (t)] 12 − 2 j j 12 − r 2 { j j 02 j j 11 } Xj=1 Xj=1 + S S i Im(Λ )[σ σ−,ρ (t)] [ρ (t)], (42) − jl j l 12 − L 12 Xjl

Na Na S i z S Γj + S S ρ˙ (t)= ε (t)[σ ,ρ (t)] i α (t)[σ ,ρ (t)] + α∗(t)[σ−(t),ρ (t)] 11 − 2 j j 11 − r 2 { j j 01 j j 10 } Xj=1 Xj=1 + S S i Im(Λ )[σ σ−,ρ (t)] [ρ (t)], (43) − jl j l 11 − L 11 Xjl

Na Na S i z S Γj S + S S ρ˙ (t)= ε (t)[σ ,ρ (t)] i √2α∗(t)[σ−(t),ρ (t)] i Im(Λ )[σ σ−,ρ (t)] [ρ (t)], (44) 02 − 2 j j 02 − r 2 { j j 01 }− jl j l 02 − L 02 Xj=1 Xj=1 Xjl

Na Na S i z S Γj S + S S ρ˙ (t)= ε (t)[σ ,ρ (t)] i α∗(t)[σ−(t),ρ (t)] i Im(Λ )[σ σ−,ρ (t)] [ρ (t)], (45) 01 − 2 j j 01 − r 2 { j j 00 }− jl j l 01 − L 01 Xj=1 Xj=1 Xjl

Na S i z S + S S ρ˙ (t)= ε (t)[σ ,ρ (t)] i Im(Λ )[σ σ−,ρ (t)] [ρ (t)], (46) 00 − 2 j j 00 − jl j l 00 − L 00 Xj=1 Xjl

S S and we have ρnm = ρmn† . Hence, for two-photon decrease which is somewhat similar to the coherent state wavepacket, six cascaded master equations are required input (Fig. 4(a)). Different from the single photon case, to calculate the dynamics of the system. the emitter 2 does not have Rabi-oscillation like structure. This is mainly because the double excited state A numerical example is shown in Fig. 4 where we con- ee can also be populated in the two photon cases (red sider two-photon interacting with two emitters. Similar dashed-dotted| i line in Fig. 4(b)) and it can cover the in- to the single-photon case, we also assume that the two- terference effect which occurs in the single photon case. photon pulse has a Gaussian spectrum and the distance From Fig. 4(b), we can also see that the subradiant state between the emitters is λa/8. Compared with the single can be populated and it can last for extended period photon case, the emitter excitation in the two-photon of|−i time (green short dashed line). The symmetric state case is larger and both excitations increase first and then 10

(b) (a) (c) 5

0.8

1.0

R pulse

atom 1

T pulse

atom 2 gg 0.8

0.6

0.4

0.2

1 0.2 Excitation probability

Excitation probability ee Pulse shape (arb. units)

0.0 0 0.0

0 2 4 6 8 10 0 2 4 6 8 10 0 3 6 9 12 15

t t t

FIG. 4: Two photon wavepacket interacting with two emitters. (a) Emitter excitation as a function of time. (b) The excitation of different eigenstates as a function of time. (c) Reflected and transmitted pulse shapes after the scattering for the same parameters as (a). In both figures, λa/8 and ∆vg /Γ = 1.

2.0 the two photon case.

Fock-1

Coherent-1

1.5

Fock-2 D. The eﬀects of pulse spectrum width

Coherent-2

n 1.0 R In the stationary scattering theory, the incident field is usually assumed to be a plane wave. In practical ex- periments, the incident light is always a pulse with finite 0.5 time duration and finite spectrum bandwidth. Here, our theory allows us to study the effects of the pulse widths.

0.0 Taking the single emitter as an example, we investi- 0 1 2 3 4 5 gate the average reflected photon number as a function of Pulse spectrum width ( ) pulse spectrum widths for different input photon states. The results are shown in Fig. 5. For all four incident FIG. 5: The average reflected photon number by a single pulses, the average reflected photon numbern ¯ decreases emitter as a function of pulse width for four different inci- R when the pulse spectrum width increases (i.e., the pulse dent photon states. Fock-1: a single photon state; Fock-2: two photon state; Coherent-1: coherent state with 1 average time duration becomes shorter). This is because of the photon number; Coherent-2: coherent state with 2 average saturation effects. When the pulse is short, it can quickly photon number. The orange dashed line is the corresponding saturate the emitter absorption and therefor the average reflectivity for the two photon state. The olive dashed-dotted reflected photon number decreases. When the pulse has line is the corresponding reflectivity of the Coherent-2 state. a white spectrum (i.e, the pulse duration is extremely short), almost no photon will be reflected for both the

(b) (a) coherent state inputs and the Fock state inputs because 1.0 1.0 most photons have frequencies far detuned from the res-

0.8 0.8 onance frequency of the emitter. In contrast, when the

R: Na=2

R(coherent) 0.6 0.6

T: Na=2 pulse spectrum is extremely narrow (i.e., the pulse is at T(coherent)

R: Na=1

R(Fock)

0.4 0.4 R: Na=1

T(Fock) the plane wave limit) and its frequency is in resonance with the emitter transition frequency, almost all of the 0.2 0.2 incident photons will be reﬂected for both the Fock state

Reflectivity and transmissivity 0.0 0.0 Reflectivity and transmissivity

-4 -2 0 2 4 -4 -2 0 2 4 inputs and the coherent state inputs. When the pulse spectrum width is finite, the Fock state input can have FIG. 6: The reflectivity and transmissivity at the plane wave larger reflectivity than that of the coherent state input limit. (a) A single emitter case for coherent state input and with the same average incident photon number. For the Fock state input. (b) Comparison of a single-emitter case and same pulse width, the pulse withn ¯in = 1 has larger re- two-emitter case with separation 0.5λa. flectivity than that of the pulse withn ¯in = 2 due to saturation effects.

+ is excited and deexcited much faster than the subra- |dianti state and it is a superradiant state (blue dashed E. The scattering property at the plane wave limit line). The corresponding|−i reﬂected and transmitted pulse shapes are shown in Fig. 4 (c) from which we can see Our theory allows us to study the scattering property that they are similar to those in the single photon case of the system at the plane wave limit. The reﬂectivity but the transmitted pulse has only a small oscillation in and transmissivity by a single emitter as a function of 11

10 (c) (a) (b) 0.5 2.5

1.0

atom 1

0.4 atom 2 2.0

atom 3

0.8

0.3 atom 4 1.5

atom 5

0.2 1.0 v = 0.6

g Reflectivity

v = /2

0.1 0.5 Excitation probability Pulse shape (arb. units)

0.4

0.0 0.0

0 5 10 15 20 1 2 3 4 5 0 5 10 15 20

t Na t

FIG. 7: (a) Emitter excitation as a function of time for five emitters. The distance between nearest emitters is 0.25λa. The pulse spectrum width ∆vg = Γ. (b) The reflected and transmitted pulse shapes for the same parameters as (a). (c) The reflectivity as a function of number of emitters for two different incident pulse spectrum widths. The emitter separation is assumed to be 0.5λa for (c). detuning frequency for the Fock state input and the co- shown in Fig. 7(a) where we assume that the incident herent state input at the plane wave limit are shown as photon pulse is in a coherent state with average pho- the symbols in Fig. 6(a). It is seen that the reflectivity ton number 1. We can see that the first emitter has the and transmissivity are the same for the Fock state input largest excitation probability, but it is quickly deexcited and the coherent state input when the incident photon and can transfer its energy to the other emitters. The is a plane wave. When the incident frequency is res- other emitters have smaller excitation probabilities, but onant with the emitter transition frequency, it will be they can oscillate and last for a period of time much completely reflected back due to the quantum interfer- longer than the decay time of single emitter and the in- ence. When the photon frequency is large detuned from cident pulse duration. This is a signature of collective the emitter frequency, it can pass through the emitter many-body effects where the collective subradiant states without being scattering. The widths of the reflectivity can be generated due to the emitter-emitter interactions and transmissivity depend on the emitter decay rate. The and these subradiant states can be populated by the in- solid lines are the analytical results calculated by the sta- cident photon pulse. The corresponding reflected and tionary theory [29] from which we can see that our results transmitted photon pulses are shown in Fig. 7(b). Most here match the previous theoretical results very well. We energy is reflected and the reflected pulse has a major also find that the reflectivity and transmissivity for a peak. In contrast, the transmitted pulse has multiple certain frequency is a property of the waveguide-QED peaks due to quantum interference between the incident system, and it does not depend on the photon statistics field and the reemitted fields by the emitters. The re- of the incident photons. However, if an incident photon flectivity as a function of number of emitters is shown with finite spectrum width, the reflectivity and transmis- in Fig. 7(c) for two different coherent pulse spectrum sivity of the pulse can strongly depend on the pulse width widths (i.e, ∆vg = Γ and ∆vg = Γ/2) where we assume and the photon statistics of the incident photons. that the emitter distance is 0.5λa. It is clearly seen that In Fig. 6(b), we compare the reflectivity and transmis- the reflectivity increases when the number of emitter in- sivity when there is a single emitter or two emitters with creases and it can approach almost unit when the number separation a =0.5λa. The symbols are numerical results of emitter is large. This is another indication of collective calculated by our input-output theory and the solid lines effects where superradiant states can be formed. This are theoretical results [39]. We can see that the theoret- phenomena can be utilized for atomic mirrors with large ical results match our numerical results very well which frequency bandwidth. again indicates the validity of our theory. We also find that the reflectivity spectrum when there are two emitters has a broader linewidth than that when there is only G. Emitter frequency modulation a single emitter due to the collective effect. Our theory also allow us to calculate the transport dynamics when the emitter frequencies are externally mod- F. General multiple-emitter case ulated. As an example, we consider a single emitter interacting with a coherent photon pulse. The emitter’s In addition to the one or two emitters, our theory can frequency is modulated such that ε(t) = 10Γsin(10Γt). in principle be applied to calculate the dynamics of an The emitter excitation as a function of time is shown arbitrary number of emitters interacting a multi-photon as the red solid line in Fig. 8(a). For comparison, the pulse. Here, we take five emitters with nearest neighbor result without frequency modulation is also plotted as distance 0.25λa as an example. The excitation proba- the black dashed line. It is seen that the excitation with bilities for the five emitters as a function of time are modulation is smaller and has some small oscillations. 12

(b)

(a) 5 0.4 culate the scattering of multiple emitters with random

R w mod.

w/o modulation

T w mod. 4 distribution and even with external frequency modula-

w modulation R w/o mod. 0.3 T w/o mod. tion. We compare the dynamics of emitters and scat- 3

0.2

tering pulse shapes when the incident photon pulses are 2 coherent state, single photon state or multiple photon

0.1 1 states. With ﬁnite incident pulse width, diﬀerent states Excitation probability Pulse shape (arb. units)

0 0.0 of light can induce diﬀerent system dynamics and diﬀer-

0 2 4 6 8 0 2 4 6 8

t t ent scattering properties. The average reflected photon number by a single emitter decreases when the incident FIG. 8: (a) Emitter excitation as a function of time for a pulse duration is shorter for both the coherent state in- single emitter with frequency modulation (solid red line). For put and the Fock state input, but the Fock state input comparison, the excitation without modulation is also shown can have higher average reflected photon number than as the black dashed line. ε(t) = 10Γ sin(10Γt). The pulse that of the coherent state input with the same average spectrum width ∆vg = Γ. (b) The reflected and transmitted photon number. The results shown here can be useful pulse shapes for the same parameters as (a). for single photon generation. At the plane wave limit, the reflectivity and transmissivity of the waveguide-QED system for a certain frequency are the same and do not The corresponding scattering pulses are shown in Fig. depend on the statistics of the incident photons. Our 8(b) where the thicker orange solid line is the reflected theory also allows to study the scattering properties of pulse and the thicker olive dashed line is the transmitted a photon pulse by emitters with frequency modulations pulse. We can see that the reflected pulse has small mod- which can be used for photon pulse shaping. We also ulations, while the transmitted pulse has very significant show that the reflectivity can significantly increase for a modulations. In comparison, the scattering pulses with- broad spectrum due to the collective interaction between out modulations have smooth shapes (thinner black solid the emitters which may be useful for designing atomic and thinner dashed lines). Hence, by frequency modula- mirror with wide bandwidth. Thus, the theory developed tion we can realize complicated photon pulse shaping. here can become important basics for studying the many- Since photon pulse shape is very important for high ef- body physics and quantum information applications in ficient quantum state preparation and transfer [41], the the waveguide-QED system. theory here can may find important applications in the quantum information.

IV. SUMMARY V. ACKNOWLEDGEMENT

In this article we derive master equations for multi- This research is supported by the Key R&D Program photon interacting with multiple emitters coupled to 1D of Guangdong Province (Grant No. 2018B030329001) waveguide. Our theory can be applied to calculate the and startup grants (No. 74130-18841222 and No. transport of arbitrary incident photon wavepackets with 20lgpy163) from Sun Yat-sen University. The research very general states of light such as coherent state, Fock of MSZ is supported by a grant from King Abdulaziz state and their superpositions. It can also be used to cal- city for Science and Technology (KACST). 13

Appendix A: Derivation of Eq. (13)

On inserting Eqs. (7-10) in the main text into Eq. (2), we can obtain

N i a O˙ (t)= ε (t)[σz (t), O (t)] S 2 j j S Xj=1

Na Na t ′ j ikzj i∆ωkt + l ikzl i∆ωkt + i g e e− [σ (t), OS (t)] ak(0) i g ∗e− σ−(t′)e dt′ k j − k Z l Xj=1 Xk Xl=1 0

Na Na t ′ j ikzj i∆ωkt l ikzl + i∆ωkt + i g ∗e− e a† (0) + i g e σ (t′)e− dt′ [σ−(t), OS (t)] k k k Z l j Xj=1 Xk Xl=1 0

Na Na t ′ j i~q ~rj i∆ω t + l i~q ~rl i∆ω t + i g e · e− ~qλ [σ (t), O (t)] a (0) i g ∗ e− · σ−(t′)e ~qλ dt′ ~qλ j S ~qλ ~qλ l − Z0 Xj=1 X~qλ Xl=1

Na Na t ′ j i~q ~rj i∆ω t l i~q ~rl + i∆ω t + i g ∗e− · e ~qλ a† (0) + i g e · σ (t′)e− ~qλ dt′ [σ−(t), O (t)] ~qλ ~qλ ~qλ l j S Z0 Xj=1 X~qλ Xl=1 N i a = ε (t)[σz (t), O (t)] 2 j j S Xj=1

Na Na j ikzj i∆ωkt + j ikzj i∆ωkt + i gke e− [σj (t), OS (t)]ak(0) + i gk∗e− e ak† (0)[σj−(t), OS (t)] Xj=1 Xk Xj=1 Xk

Na Na j i~q ~rj i∆ω t + j i~q ~rj i∆ω t + i g e · e− ~qλ [σ (t), O (t)]a (0) + i g ∗e− · e ~qλ a† (0)[σ−(t), O (t)] ~qλ j S ~qλ ~qλ ~qλ j S Xj=1 X~qλ Xj=1 X~qλ

Na t ′ j l ik(zj zl) i∆ωkt + i∆ωkt + g g ∗e − e− [σ (t), OS (t)] σ−(t′)e dt′ k k j Z l Xjl Xk 0

Na t j ′ j l ik(zj zl) i∆ω (t)t + i∆ωkt g g ∗e− − e k σ (t′)e− dt′[σ−(t), OS (t)] − k k Z l j Xjl Xk 0

Na t ′ j l i~q (~rj ~rl) i∆ω t + i∆ω t + g g ∗ e · − e− ~qλ [σ (t), O (t)] σ−(t′)e ~qλ dt′ ~qλ ~qλ j S l Z0 Xjl X~qλ

Na t ′ j l i~q (~rj ~rl) i∆ω t + i∆ω t g ∗g e− · − e ~qλ σ (t′)e− ~qλ dt′[σ−(t), O (t)]. (A1) ~qλ ~qλ l j S − Z0 Xjl X~qλ According to the Weisskopf-Wigner approximation, we have [40]

′ Γ Γ ′ ′ j l ik(zj zl) i∆ωkt i∆ωkt j l ika zjl iεj (t)t iεl(t )t zjl gkgk∗e − e− e = e | |e − δ[t′ (t | |)], (A2) p 2 − − vg Xk ′ Γ Γ ′ ′ j l ik(zj zl) i∆ωkt i∆ωkt j l ika zjl iεj (t)t+iεl(t )t zjl gkgk∗e− − e e− = e− | |e− δ[t′ (t | |)], (A3) p 2 − − vg Xk

′ ′ ′ j l i~q (~rj ~rl) i∆ω t i∆ω t iεj (t)t iεl(t )t rjl g g ∗ e · − e− ~qλ e ~qλ =Ω e − δ[t′ (t | |)], (A4) ~qλ ~qλ jl − − vg X~qλ

′ ′ ′ j l i~q (~rj ~rl) i∆ω t i∆ω t iεj (t)t+iεl(t )t rjl g ∗g e− · − e ~qλ e− ~qλ =Ω∗ e− δ[t′ (t )], (A5) ~qλ ~qλ jl | | − − vg X~qλ

2L i 2 i Γivg 3√γj γl 2 i 2 1 i ikarjl where Γi = gk with gk = and Ωjl = [sin φ − + (1 3cos φ)( 2 + 2 )]e with vg | 0 | 0 q 2L 4 karjl − (karjl) (karjl) r = ~r ~r . jl | j − l| 14

zjl To proceed, we assume that the emitters are close such that zij /vg 1/Γ, we can approximate that σ−(t ) ≪ j − vg ≈ 8 8 σ−(t) in the rotating frame. Indeed, this is the usual case. For example, if v 10 m/s and Γ 10 Hz, we require j g ∼ ∼ that the distance between the emitters zij 1m which is the usual case. By doing this approximation, Eq. (13) then becomes ≪

Na Na Na i z Γj + Γj O˙ (t)= ε [σ (t), O (t)] + i [σ (t), O (t)][a (t)+ b (t)] + i [a†(t)+ b†(t)][σ−(t), O (t)] S 2 j j s r 2 j S j j r 2 j j j S Xj=1 Xj=1 Xj=1 + + + Λ [σ (t), O (t)]σ−(t) Λ∗ σ (t)[σ−(t), O (t)], (A6) jl j S l − jl l j S Xjl Xjl

v g ∞ ikzj iδωk t where aj (t) = 2π 0 e ak(0)e− dk is the absorption of the incident waveguide photons and bj (t) = q vg R i~qλ ~rj iδω~qλ t 3 2π e · a~qλ (0)e− d ~qλ is the absorption of the incident nonguided photons. The collective interaction betweenq R Rthe R emitters is given by [40]

Γ Γ 3 γj γl j l ika zjl √ 2 i 2 1 i ika rjl Λjl = e | | + [sin φ − + (1 3cos φ)( 2 + 2 )]e | |. (A7) p 2 4 karjl − (karjl) (karjl)

From Eq. (A6), we can derive a corresponding master equation for the emitters. Since T rS+R[OS (t)ρ] = T rS[OSρS(t)] where ρS(t)= T rR[ρ(t)], we have

T rS[OSρ˙S(t)]

= T rS+R[O˙ S(t)ρ] N i a = ε (t)T r [σz(t), O (t)]ρ 2 j S+R{ j S } Xj=1

Na Na Γj + Γj + i T r [σ (t), O (t)][a (t)+ b (t)]ρ + i T r [a†(t)+ b†(t)][σ−(t), O (t)]ρ r 2 S+R{ j S j j } r 2 S+R{ j j j S } Xj=1 Xj=1 + + + Λ T r [σ (t), O (t)]σ−(t)ρ Λ∗ T r σ (t)[σ−(t), O (t)]ρ jl S+R{ j S l }− jl S+R{ l j S } Xjl Xjl

Na Na Na i z Γj j + Γj j = ε (t)T r O [σ ,ρ (t)] + i T r O [ρ (t), σ ] + i T r O [ρ † (t), σ−] −2 j S { S j S } r 2 S{ S in j } r 2 S{ S in j } Xj=1 Xj=1 Xj=1 + + + + T r O [σ σ−ρ (t) σ−ρ (t)σ ] Λ∗ T r O [ρ (t)σ σ− σ−ρ (t)σ ] , (A8) − S{ S j l S − l S j }− jl S{ S S l j − j S l } Xjl Xjl

j where ρin(t) = T rR U(t)[aj (t)+ bj(t)]ρ(0)U †(t) is the contribution from the incident sources. In this paper, we consider that the incident{ photon is coming from} the waveguide photons and the non-guided modes are initially in j the vacuum. Since a~qλ (0) 0 = 0, we have bj (t)ρ(0) = 0 and therefore ρin(t)= T rR U(t)aj (t)ρ(0)U †(t) is due to the contribution of the incident| i waveguide photons. Comparing both size of Eq. (A8), we{ can obtain the master} equation for the system density matrix given by

Na Na Na i z Γj + j Γj j ρ˙ (t)= ε (t)[σ ,ρ (t)] i [σ ,ρ (t)] i [σ−,ρ † (t)] S − 2 j j S − r 2 j in − r 2 j in Xj=1 Xj=1 Xj=1 + + + + i Im(Λ )[σ σ−,ρ (t)] Re(Λ )[σ σ−ρ (t)+ ρ (t)σ σ− 2σ−ρ (t)σ ], (A9) − jl j l S − jl j l S S j l − l S j Xjl Xjl which is the master equation shown in Eq. (13) in the main text. 15

Appendix B: Derivation of the input-output relations

From Eqs. (15-19) in the main text, we can obtain

Na tf ′ R R Γj vg vg ∞ iδkzN ikzj i∆ωk(t t) a (t)= a (t zN /vg) i σ−(t′)dt′ e e− e − dk out in − − r 4π Z j r2π Z Xj=1 0 0

Na tf ′ R Γj vg vg ikazj ∞ iδk(zN zj ) iδkvg (t t) = a (t zN /vg) i e− σ−(t′)dt′ e − e − dk in − − r 4π r2π Z j Z Xj=1 0 0

Na tf ′ R Γj vg vg ikazj ∞ iδkzNj iδkvg (t t) = ain(t zN /vg) i e− σj−(t′)dt′ e e − dδk − − r 4π r2π Z0 Z k Xj=1 − 0

Na tf ′ R Γj vg vg ikazj ∞ iδkzNj iδkvg (t t) = ain(t zN /vg) i e− σj−(t′)dt′ e e − dδk − − r 4π r2π Z0 Z Xj=1 −∞

Na tf R Γj vg vg ikazj 2π = a (t zN /vg) i e− σ−(t′) δ(t′ t + zNj/vg)dt′ in − − r 4π r2π Z j v − Xj=1 0 g

Na R Γj ikazj = a (t z /v ) i e− σ−(t z /v ) in − N g − r 2 j − Nj g Xj=1

Na R Γj ikazj a (t z /v ) i e− σ−(t), (B1) ≈ in − N g − r 2 j Xj=1 where z = z z . Similarly, we have Nj N − j

Na tf 0 ′ L L Γj vg vg iδkz1 ikzj i∆ωk(t t) aout(t)= ain(t + z1/vg) i σj−(t′)dt′ e− e− e − dk − r 4π Z0 r2π Z Xj=1 −∞

Na tf ∞ ′ L Γj vg vg ikazj iδkzj1 iδkvg (t t) = a (t + z1/vg) i e σ−(t′)dt′ e e − d( k) in − r 4π r2π Z j Z − Xj=1 0 0

Na tf ∞ ′ L Γj vg vg ikazj iδkzj1 iδkvg (t t) = ain(t + z1/vg) i e σj−(t′)dt′ e e − dδk − r 4π r2π Z0 Z k Xj=1 − 0

Na tf ∞ ′ L Γj vg vg ikazj iδkzj1 iδkvg (t t) = ain(t + z1/vg) i e σj−(t′)dt′ e e − dδk − r 4π r2π Z0 Z Xj=1 −∞

Na tf L Γj vg vg ikazj 2π = a (t + z1/vg) i e σ−(t′) δ(t′ t + zj1/vg)dt′ in − r 4π r2π Z j v − Xj=1 0 g

Na L Γj ikazj = a (t + z /v ) i e σ−(t z /v ) in 1 g − r 2 j − j1 g Xj=1

Na L Γj ik z a (t + z /v ) i e a j σ−(t). (B2) ≈ in 1 g − r 2 j Xj=1

Eqs. (B1) and (B2) are the input-output relations of the system from which we can calculate the ﬁeld scattering properties of this system. 16

Appendix C: Derivation of Eq. (35)

We can derive a dynamical equation for ρ01(t) using similar method as deriving ρS(t). Since T rS+R[OS ρ01(t)] = T r [O (t)ρ (0)] where ρ (0) = ρ (0) 0 Ψ , we have S+R S 01 01 S ⊗ | ih F |

T rS+R[OS ρ˙01(t)]

=T rS+R[O˙ S (t)ρ01(0)] N i a = ε (t)T r [σz(t), O (t)]ρ (0) 2 j S+R{ j s 01 } Xj=1

Na Na Γj + j Γj j + i T r [σ (t), O (t)]a (t z /v )ρ (0) + i T r a † (t z /v )[σ−(t), O (t)]ρ (0) r 2 S+R{ j S in − j g 01 } r 2 S+R{ in − j g j S 01 } Xj=1 Xj=1 + + + Λ T r [σ (t), O (t)]σ−(t)ρ Λ∗ T r σ (t)[σ−(t), O (t)]ρ jl S+R{ j S l }− jl S+R{ l j S } Xjl Xjl

Na Na i z Γj = ε (t)T r [σ ,ρ (t)]O i α∗(t z /v )T r O [σ−,ρ (t)] − 2 j S+R{ j 01 s}− r 2 j − j g S+R{ S j 00 } Xj=1 Xj=1 + + + + Λ T r O [σ σ−ρ (t) σ−ρ (t)σ ] Λ∗ T r O [ρ (t)σ σ− σ−ρ (t)σ ] , (C1) − jl S+R{ S j l 01 − l 01 j }− jl S+R{ S 01 l j − j 01 l } Xjl Xjl where ρ (t)= U(t)ρ 0 0 U †(t). Comparing both sides, we have 00 S ⊗ | ih |

Na Na S i z S Γj vg S + S S ρ˙ (t)= ε (t)[σ ,ρ (t)] i α∗(t)[σ−,ρ (t)] i Im(Λ )[σ σ−,ρ (t)] [ρ (t)], (C2) 01 −2 j j 01 − r 2L j j 00 − jl j l 01 − L 01 Xj=1 Xj=1 Xjl

S where ρ00(t)= T rR[ρ00(t)].

[1] H. J. Kimble, The quantum internet, Nature 453, 1023 Quantum supercavity with emitter mirrors, Phys. Rev. (2008). A 78, 063827 (2008). [2] H. Zheng, D. J. Gauthier, and H. U. Baranger, [9] H. Dong, Z. Gong, H. Ian, L. Zhou, and C. P. Sun, In- Waveguide-QED-based photonic quantum computation, trinsic cavity QED and emergent quasinormal modes for Phys. Rev. Lett. 111, 090502 (2013). a single photon, Phys. Rev. A 79, 063847 (2009). [3] Z. Liao, X. Zeng, H. Nha, and M. S. Zubairy, Photon [10] D. E. Chang, L. Jiang, A. V. Gorshkov, and H. J. Kim- transport in a one-dimensional nanophotonic waveguide ble, Cavity QED with emitter mirrors, New J. Phys. 14, QED system, Phys. Scr. 91, 063004 (2016). 063003 (2012). [4] D. Roy, C. Wilson, and O. Firstenberg, Strongly inter- [11] Z. Liao, H. Nha, and M. S. Zubairy, Single-photon acting photons in one-dimensional continuum, Rev. Mod. frequency-comb generation in a one-dimensional waveg- Phys. 89, 021001 (2017). uide coupled to two emitter arrays, Phys. Rev. A 93, [5] M. Arcari, I. Söllner, A. Javadi, S. Lindskov Hansen, S. 033851 (2016). Mahmoodian, J. Liu, H. Thyrrestrup, E. H. Lee, J. D. [12] J. S. Douglas, H. Habibian, C.-L. Hung, A. V. Gorshkov, Song, S. Stobbe, and P. Lodahl, Near-Unity Coupling H. J. Kimble, and D. E. Chang, Quantum many-body Efficiency of a Quantum Emitter to a Photonic Crystal models with cold atoms coupled to photonic crystals, Waveguide, Phys. Rev. Lett. 113, 093603 (2014). Nat. Photon. 9, 326 (2015). [6] A. Laucht, S. Pütz, T. Günthner, N. Hauke, R. Saive, S. [13] P. Richerme, Z.-X. Gong, A. Lee, C. Senko, J. Smith, M. Frédérick, M. Bichler, M.-C. Amann, A. W. Holleitner, Foss-Feig, S. Michalakis, A. V. Gorshkov, and C. Mon- M. Kaniber, and J. J. Finley, A waveguide-coupled on- roe, Non-local propagation of correlations in quantum chip single-photon source, Phys. Rev. X 2, 011014 (2012). systems with long-range interactions, Nature 511, 198 [7] C. Schuck, W. H. P. Pernice, and H. X. Tang, Waveg- (2014). uide integrated low noise NbTiN nanowire single-photon [14] Y. Yu, F. Ma, X. Y. Luo, B. Jing, P.-F. Sun, R.-Z. detectors with milli-Hz dark count rate, Sci Rep. 3, 1893 Fang, C.-W. Yang, H. Liu, M.-Y. Zheng, X.-P. Xie, W.- (2013). J. Zhang, L.-X. You, Z. Wang, T.-Y. Chen, Q. Zhang, [8] L. Zhou, H. Dong, Y.-X. Liu, C. P. Sun, and F. Nori, X.-H. Bao, and J.-W. Pan, Entanglement of two quan- 17

tum memories via fibres over dozens of kilometres, Na- [33] Z. H. Wang, L. Zhou, Y. Li, and C. P. Sun, Controllable ture 578, 240 (2020). single-photon frequency converter via a one-dimensional [15] T. van Leent, M. Bock, R. Garthoff, K. Redeker, W. waveguide, Phys. Rev. A 89, 053813 (2014). Zhang, T. Bauer, W. Rosenfeld, C. Becher, and H. Wein- [34] Y. Lu, S. Gao, A. Fang, P. Li, F. Li, and M.S. Zubairy, furter, Long-distance distribution of emitter-photon en- Coherent frequency down-conversions and entanglement tanglement at telecom wavelength, Phys. Rev. Lett. 124, generation in a Sagnac interferometer, Opt. Express 25, 010510 (2020). 16151 (2017). [16] K. Y. Bliokh and F. Nori, Transverse and longitudinal [35] D. Roy, Two-photon scattering by a driven three-level angular momenta of light, Phys. Rep. 592, 1–38 (2015). emitter in a one-dimensional waveguide and electromag- [17] H. Pichler, T. Ramos, A. J. Daley, and P. Zoller, Quan- netically induced transparency, Phys. Rev. Lett. 106, tum optics of chiral spin networks, Phys. Rev. A 91, 053601 (2011). 042116 (2015). [36] D. Witthaut and A. S. Sørensen, Photon scattering by a [18] P. Lodahl, S. Mahmoodian, S. Stobbe, A. Rauschenbeu- three-level emitter in a one-dimensional waveguide, New tel, P. Schneeweiss, J. Volz, H. Pichler, and P. Zoller, J. Phys. 12, 043052 (2010). Chiral quantum optics, Nature 541, 473 (2017). [37] I. Shomroni, S. Rosenblum, Y. Lovsky, O. Bechler, G. [19] J. Petersen, J. Volz, and A. Rauschenbeutel, Chiral Guendelman, and B. Dayan, All-optical routing of sin- nanophotonic waveguide interface based on spin-orbit in- gle photons by a one-atom switch controlled by a single teraction of light, Science 346, 67 (2014). photon, Science 345, 903–906 (2014). [20] S. Mahmoodian, P. Lodahl, and A. S. Sørensen, Quan- [38] Y. Chen, M. Wubs, J. Mørk, and A. F. Koenderink, Co- tum networks with chiral-light–matter interaction in herent single-photon absorption by single emitters cou- waveguides, Phys. Rev. Lett. 117, 240501 (2016). pled to one-dimensional nanophotonic waveguides, New [21] M.-T. Cheng, X.-S. Ma, J.-Y. Zhang, and B. Wang, Sin- J. Phys. 13, 103010 (2011). gle photon transport in two waveguides chirally coupled [39] Z. Liao, X. Zeng, S. –Y. Zhu, and M. S. Zubairy, Single- by a quantum emitter, Opt. Express 24, 19988 (2016). photon transport through an atomic chain coupled to a [22] K. Xia, F. Nori, and M. Xiao, Cavity-free optical isola- one-dimensional nanophotonic waveguide, Phys. Rev. A tors and circulators using a chiral cross-Kerr nonlinearity, 92, 023806 (2015). Phys. Rev. Lett. 121, 203602 (2018). [40] Z. Liao, H. Nha, and M. S. Zubairy, Dynamical theory [23] Y. Lu, S. Gao, A. Fang, Z. Liao , F. Li, Micro-scale of single-photon transport in a one-dimensional waveg- Fabry–Perot interferometer with high spectral resolution uide coupled to identical and nonidentical emitters, Phys. and tunable transmission frequency via chiral waveguide- Rev. A 94, 053842 (2016). emitter coupling, Phys. Lett. A 382 1823 (2018). [41] Z. Liao and M. S. Zubairy, Quantum state preparation by [24] J. Q. Grim, A. S. Bracker, M. Zalalutdinov, S. G. Carter, a shaped photon pulse in a one-dimensional continuum, A. C. Kozen, M. Kim, C. S. Kim, J. T. Mlack, M. Yakes, Phys. Rev. A 98, 023815 (2018). B. Lee, and D. Gammon, Scalable in operando strain tun- [42] Z. Liao, M. Al-Amri, and M. S. Zubairy, Measurement of ing in nanophotonic waveguides enabling three-quantum- deep-subwavelength emitter separation in a waveguide- dot superradiance, Nat. Mat. 18, 963 (2019). QED system, Opt. Express 25, 31997 (2017). [25] J. S. Douglas, T. Caneva, and D. E. Chang, Photon [43] F. Dinc, I.˙ Ercan, and A. M. Brańczyk, Exact Markovian molecules in atomic gases trapped near photonic crystal and non-Markovian time dynamics in waveguide QED: waveguides, Phys. Rev. X 6, 031017 (2016). collective interactions, bound states in continuum, su- [26] F. Arute, K. Arya, R. Babbush et al., Quantum perradiance and subradiance, Quantum 3, 213 (2019). supremacy using a programmable superconducting pro- [44] J. T. Shen and S. Fan, Strongly correlated two-photon cessor, Nature 574, 505 (2019). transport in a one-dimensional waveguide coupled to a [27] C. Song, K. Xu, H. Li, Y.-R. Zhang, X. Zhang, W. Liu, two-level system, Phys. Rev. Lett. 98, 153003 (2007). Q. Guo, Z. Wang, W. Ren, J. Hao, H. Feng, H. Fan, D. [45] T. Shi, Y.-H. Wu, A. González-Tudela, and J. I. Cirac, Zheng, D.-W. Wang, H. Wang, S.-Y. Zhu, Generation of Bound states in Boson impurity models, Phys. Rev. X 6, multicomponent atomic Schrödinger cat states of up to 021027 (2016). 20 qubits, Science 365, 574 (2019). [46] P. Facchi, M. S. Kim, S. Pascazio, F. V. Pepe, D. Po- [28] M. S. Tame, K. R. McEnery, S. K. Ozdemir,¨ J. Lee, S. A. marico, and T. Tufarelli, Bound states and entanglement Maier, and M. S. Kim, Quantum plasmonics, Nat. Phys. generation in waveguide quantum electrodynamics, Phys. 9, 329 (2013). Rev. A 94, 043839 (2016). [29] J. -T. Shen and S. Fan, Coherent photon transport [47] G. Calajó, Y.-L. L. Fang, H. U. Baranger, and F. Cic- from spontaneous emission in one-dimensional waveg- carello, Exciting a bound state in the continuum through uides, Opt. Lett. 30, 2001 (2005). multiphoton scattering plus delayed quantum feedback, [30] T. S. Tsoi and C. K. Law, Quantum interference ef- Phys. Rev. Lett. 122, 073601 (2019). fects of a single photon interacting with an atomic chain [48] H. Zheng, D. J. Gauthier, and H. U. Baranger, Waveg- inside a one-dimensional waveguide, Phys. Rev. A 78, uide QED: Many-body bound-state effects in coherent 063832(2008). and Fock-state scattering from a two-level system, Phys. [31] M. -T. Cheng, J. Xu, and G. S. Agarwal, Waveguide Rev. A 82, 063816 (2010). transport mediated by strong coupling with atoms, Phys. [49] Y.-L. L. Fang, H. Zheng, and H. U. Baranger, One- Rev. A 95, 053807 (2017). dimensional waveguide coupled to multiple qubits: [32] M. Bradford, K. C. Obi, and J.-T. Shen, Efficient single- photon-photon correlations, EPJ Quantum Technol. 1, photon frequency conversion using a Sagnac interferom- 3 (2014). eter, Phys. Rev. Lett. 108, 103902 (2012). [50] Y. Shen and J.-T. Shen, Photonic-Fock-state scattering 18

in a waveguide-QED system and their correlation func- [64] O. O. Chumak and E. V. Stolyarov, Phase-space distri- tions, Phys. Rev. A 92, 033803 (2015). bution functions for photon propagation in waveguides [51] Y.-L. L. Fang and H. U. Baranger, Multiple emitters in coupled to a qubit, Phys. Rev. A 88, 013855 (2013). a waveguide: non-reciprocity and correlated photons at [65] O. O. Chumak and E. V. Stolyarov, Photon distri- perfect elastic transmission, Phys. Rev. A 96, 013842 bution function for propagation of two-photon pulses (2017). in waveguide-qubit systems, Phys. Rev. A 90, 063832 [52] T. Shi and C. P. Sun, Lehmann-Symanzik-Zimmermann (2014). reduction approach to multiphoton scattering in coupled- [66] W. Konyk and J. Gea-Banacloche, One- and two-photon resonator arrays, Phys. Rev. B 79, 205111 (2009). scattering by two emitters in a waveguide, Phys. Rev. A [53] T. Shi, S. Fan, and C. P. Sun, Two-photon transport in 96, 063826 (2017). a waveguide coupled to a cavity in a two-level system, [67] T. Caneva, M. T. Manzoni, T. Shi, J. S. Douglas, J. I. Phys. Rev. A 84, 063803 (2011). Cirac, and D. E. Chang, Quantum dynamics of prop- [54] M. Laakso and M. Pletyukhov, Scattering of two photons agating photons with strong interactions: a generalized from two distant qubits: exact solution, Phys. Rev. Lett. input–output formalism, New J. Phys. 17, 113001 (2015). 113, 183601 (2014). [68] G.-Z. Song, E. Munro, W. Nie, L.-C. Kwek, F.-G. Deng, [55] S. Fan, S. E. Kocaba¸s, and J.-T. Shen, Input-output and G.-L. Long, Photon transport mediated by an emit- formalism for few-photon transport in one-dimensional ter chain trapped along a photonic crystal waveguide, nanophotonic waveguides coupled to a qubit, Phys. Rev. Phys. Rev. A 98, 023814 (2018). A 82, 063821 (2010). [69] T. Shi, D. E. Chang, and J. I. Cirac, Multiphoton- [56] K. Lalumière, B. C. Sanders, A. F. van Loo, A. Fedorov, scattering theory and generalized master equations, A. Wallraff, and A. Blais, Input-output theory for waveg- Phys. Rev. A 92, 053834 (2015). uide QED with an ensemble of inhomogeneous emitters, [70] K. M. Gheri, K. Ellinger, T. Pellizzari, and P. Zoller, Phys. Rev. A 88, 043806 (2013). Photon-wavepackets as flying quantum bits, Fortschr. [57] S. Xu and S. Fan, Input-output formalism for few-photon Phys. 46, 401–415 (1998). transport: a systematic treatment beyond two photons, [71] B. Q. Baragiola, R. L. Cook, A. M. Branćzyk, and J. Phys. Rev. A 91, 043845 (2015). Combes, N-photon wave packets interacting with an arbi- [58] J. Combes, J. Kerckhoff , and M. Sarovar, The SLH trary quantum system, Phys. Rev. A 86, 013811 (2012). framework for modeling quantum input-output networks, [72] A. H. Kiilerich and K. Mølmer, Input-output theory with Advances in Physics: X 2, 784–888 (2017). quantum pulses, Phys. Rev. Lett. 123, 123604 (2019). [59] A. Roulet and V. Scarani, Solving the scattering of N [73] J. You, Z. Liao, S.-W. Li, and M. S. Zubairy, Waveguide photons on a two-level emitter without computation, quantum electrodynamics in squeezed vacuum, Phys. New J. Phys. 18, 093035 (2016). Rev. A 97, 023810 (2018). [60] F. Dinc, Diagrammatic approach for analytical non- [74] J. You, Z. Liao, and M. S. Zubairy, Steady-state popula- Markovian time evolution: Fermi’s two-atom problem tion inversion of multiple Ξ-type emitters by the squeezed and causality in waveguide quantum electrodynamics, vacuum in a waveguide, Phys. Rev. A 100, 013843 Phys. Rev. A 102, 013727 (2020). (2019). [61] D. J. Brod, J. Combes, and J. Gea-Banacloche, Two pho- [75] D.F. Walls and G. J. Milburn, Quantum Optics, tons co- and counterpropagating through N cross-Kerr (Springer, Verlag Berlin Heidelberg, 2008). sites. Phys. Rev. A 94, 023833 (2016). [76] M. O. Scully and M. S. Zubairy, Quantum Optics, (Cam- [62] J. Combes and D. J Brod, Two-photon self-Kerr non- bridge University Press, Cambridge, 1997). linearities for quantum computing and quantum optics, [77] P. Senellart, G. Solomon, and A. White, High- Phys. Rev. A 98, 062313 (2018). performance semiconductor quantum-dot single-photon [63] P. Domokos, P. Horak, and H. Ritsch, Quantum descrip- sources, Nat. Nanotech. 12, 1026–1039 (2017). tion of light-pulse scattering on a single emitter in waveguides, Phys. Rev. A, 65, 033832 (2002).