## Abstract

We investigate coherent control of a single atom interacting with an optical bipartite lattice via a combined high-frequency modulation. Our analytical results show that the quantum tunneling and dynamical localization can depend on phase difference between the modulation components, which leads to a different route for the coherent destruction of tunneling and a convenient phase-control method for stabilizing the system to implement the directed transport of atom. The similar directed transport and the phase-controlled quantum transition are revealed for the corresponding many-particle system. The results can be referable for experimentally manipulating quantum transport and transition of cold atoms in the tilted and shaken optical bipartite lattice or of analogical optical two-mode quantum beam splitter, and also can be extended to other optical and solid-state systems.

© 2014 Optical Society of America

## 1. Introduction

Quantum control of tunneling processes of particles plays a major role in different areas of physics, optics and chemistry [1–4]. As early as 1986, Dunlap and Kenkre studied theoretically the quantum motion of a charged particle on a discrete lattice driven by an ac field [5], and found the surprising result that particle transport can be completely suppressed when ratio of the strength and the frequency of the ac field takes some special values. This effect of extreme localization was later found to be associated with the coherent destruction of tunneling (CDT) [2, 6, 7] at a collapse point of the Floquet quasienergy spectrum [8], and has also been observed in different systems [9–12]. Generally, the dynamical localization (DL) refers to the phenomenon wherein a particle initially localized in a lattice can transport within a fi-nite distance and periodically return to its original state. There has been growing interest in the quantum control of electrons in semiconductor superlattices or arrays of coupled quantum dots from both theoretical and experimental sides [2, 13]. Most of the DL and CDT of the electronic systems are generic and also can occur in atomic [8, 9, 14, 15] and optical [10, 16–20] systems.

Recently, different routes to CDT were found by considering, respectively, the priori prescribed number of bosons of a many-boson system [21, 22], the distinguishable intersite separations of a bipartite lattice [23, 24], the variable driving symmetry of a two-frequency driven particle in a double-well [11, 25], and the different combined modulations to the different systems [26, 27]. The CDT mechanism has been applied to various physical fields such as the qubit control [28, 29], the quantum tunneling switch [30], and the directed transport in a bipartite lattice via the selective CDT to the two different barriers [23, 24]. It is worth noting that the CDT and DL mechanisms can also be applied to coherently control instability of the periodically driven double-well system [31], optical lattice system [32, 33] and fiber system [34]. In the sense of Lyapunov, by the instability we mean that the small initial deviation from a given solution grows without upper limit, which could lead to destruction of the solution behavior. It was found that instability of the bipartite lattice systems depends on different signs of the effective tunneling rates of two nearest-neighbor barriers [32, 33]. Therefore, one can stabilize the systems by tuning the effective tunneling rates. For a single particle held in a simple lattice with a single intersite separation, the nearest-neighbor barriers are the same such that the instability cannot be shown. For a bipartite lattice system, such an instability may be induced and suppressed alternately by adjusting the driving parameters, resulting in the directed transport of particles. Here our aim is finding a new route of CDT and supplying a simple stabilization method to realize the directed transport and quantum transition of cold atoms in a driven bipartite lattice.

The coherent control of an ac driven particle in a single-band lattice with a single intersite separation has been investigated widely in the nearest-neighbor tight binding (NNTB) approximation [5, 26]. More recently, a bipartite optical lattice or double-well train with two different intersite separations has been realized experimentally by superimposing two laser beams with two different wavelengths [35, 36]. Such a system was applied to induce the ratchet-like effect [23, 24], to transport quantum information [37] and to realize two-qubit quantum gates [38]. The periodic modulation is usually applied to the potential tilt (bias) between the lattice sites [5, 39, 40] or the tunnel coupling [41–44]. The combined modulations between both have also been adopted to produce the exact solutions for a phase controllable lattice system [26] or for an analytically solvable two-level system [27, 45]. The adjustments of driving parameters can be performed in a nonadiabatic [23, 37] or adiabatic manner [46, 47]. On the other hand, the directed transports have been investigated for a single classical particle in a spatially periodic potential [48] and for a mean-field treated Bose-Einstein condensate loaded in an optical superlattice [49, 50].

In this work, we firstly consider a single atom held in an optical bipartite lattice with two different separations *a* and *b* and driven by a combined modulation of two resonant external fields with a phase difference between the bias and coupling. Such a system can also be regarded as an atomic analog of the optical two-mode quantum beam splitter [35]. In the high-frequency regime and NNTB approximation, we derive an analytical general solution for the probability amplitude of the particle in any localized state in which the quantum tunneling and stability can depend on the phase difference between the two modulation components. A new route of CDT and a simple method for stabilizing the system to perform the directed transport are found by adjusting the phase difference nonadiabatically. Such a phase-adjustment may be more convenient in experiments compared to the usual amplitude- and interaction-modulations. Finally, we suggest a scheme for extending the results to the phase-controlled directed transport and quantum transition between the superfluid and Mott insulator for the corresponding many-particle system. The results can be readily amenable with existing experimental setups [39, 40, 42–44] on the periodically tilted and shaken optical lattices [26] and could be applied to simulating the different optical systems [16, 17] and solid-state systems [13].

## 2. General solution in the high-frequency regime

We consider a driven and tilted bipartite lattice in the form of a train of double wells formed by the tilted laser standing wave

*k*, the short lattice of wave vector 2

_{L}*k*and the linear potential. Here

_{L}*θ*denotes the laser phase [43], the potential tilt between the lattice sites takes the form [39, 40, 51]

*ε*(

*t*) = −

*ε*

_{0}cos(

*ωt*) with amplitude

*ε*

_{0}and frequency

*ω*, and the time-periodic lattice depths reads [37, 43, 44]

*V*(

_{i}*t*) =

*V*

_{i}_{0}+

*δV*cos(

_{i}*mωt*−

*ϕ*) for

*m*= 0, 1, 2,..., and with constants

*V*

_{i}_{0}and

*δV*. The nonzero

_{i}*m*means the frequency resonance between the modulation components. Such a lattice can be realized experimentally by a periodically shaken optical lattice [37, 43, 44], and by moving the position of a retroreflecting mirror which is mounted on a piezoelectric actuator [39] or by imposing a phase modulation to one of the standing wave component fields [51]. A single particle of mass

*M*is initially placed near the lattice center, as shown in Fig. 1 [35], where we have adopted the spatial coordinate normalized by ${k}_{L}^{-1}$ and the phase

*θ*= 4.6, and selected the suitable driving parameters and initial time

*t*

_{0}=

*π*/(2

*ω*) to make

*ε*(

*t*

_{0}) = 0 and

*V*

_{1}(

*t*

_{0}) = 1,

*V*

_{2}(

*t*

_{0}) = 2. The different separations

*a*and

*b*can be adjusted by changing the laser wave vector

*k*and amplitudes

_{L}*V*(

_{i}*t*) [37, 43]. Quantum dynamics of such a system is governed by the Hamiltonian [23, 26, 33]

*i*,

*j*) means the nearest-neighbor site pairs. Signs ${b}_{j}^{\u2020}$ and

*b*are, respectively, the particle creation and annihilation operators in the site

_{j}*j*. The spatial location of the

*n*th lattice site reads [5]

*w*(

*x*−

*x*) is the Wannier function. Expressing the lattice depths in terms of the recoil energy

_{i}*E*= (

_{r}*h̄k*)

_{L}^{2}/(2

*M*), the tunnel coupling is calculated by the formula [5, 43]

*J*(

_{ij}*t*) for an even

*i*has only a small time-independent difference from that for an odd

*i*and it will henceforth be renormalized, so we can take [23, 43]

*J*(

_{ij}*t*) =

*J*(

*t*) =

*J*

_{0}+

*δJ*cos(

*mωt*−

*ϕ*), where constant

*J*

_{0}is from the terms of kinetic energy and of

*V*

_{i}_{0}, the shaking intensity

*δJ*is proportional to the driving amplitudes [43]

*δV*. To simplify, we have set

_{i}*h̄*= 1 and normalized energy and time by

*E*and ${\omega}_{0}^{-1}=10(\overline{h})/{E}_{r}$, which are determined by the laser wave vector and atomic mass. The parameters

_{r}*J*

_{0},

*δJ*and (

*ε*

_{0}

*x*) are in units of

_{n}*ω*

_{0}=

*E*/10 with

_{r}*x*being normalized by the wave length

_{n}*λ*=

_{s}*π/k*of the short lattice. Thus all the parameters are dimensionless throughout this paper. The experimentally achievable parameter regions may be selected as [42–44]

_{L}*λ*∼ 800nm,

_{s}*J*

_{0}∼

*ω*

_{0},

*δJ*<

*J*

_{0},

*ε*

_{0}

*λ*∼

_{s}*ω*∈ [0, 100](

*ω*

_{0}), and

*a*,

*b*∼

*λ*.

_{s}Letting |*n*〉 be the localized state at the site *n*, we expand the quantum state |*ψ*(*t*)〉 as the linear superposition |*ψ*(*t*)〉 = ∑_{n}*c _{n}*(

*t*)|

*n*〉. Combining this with Eq. (1), from the time-dependent Schrödinger equation $i\frac{\partial}{\partial t}|\psi (t)\u3009=H(t)|\psi (t)\u3009$ we derive the coupled equations of the probability amplitudes [5, 33, 47]

*c*(

_{n}*t*) =

*A*(

_{n}*t*) exp(

*iε*

_{0}

*ω*

^{−1}

*x*sin

_{n}*ωt*) which leads Eq. (2) to the form

*n*, and ${\mathrm{\Delta}}_{n}=\frac{{\epsilon}_{0}}{\omega}b$, ${\mathrm{\Delta}}_{n-1}=\frac{{\epsilon}_{0}}{\omega}a$ for odd

*n*.

We focus our attention on the situation of high-frequency regime with *ω* ≫ 1. The selective CDT has been illustrated analytically and numerically under this limit for an amplitude modulation [23]. Here we shall give a general analytical solution of the system, which reveals the phase-controlled CDT and directed transport for the considered combined modulation, then extend the results to a many-particle system. Note that in Eq. (3), *A _{n}*(

*t*) may be treated as a set of slowly varying functions of time, and the coupling

*𝒥*(Δ

_{m}*) = (−1)*

_{n}

^{m}𝒥_{−}

*(Δ*

_{m}*) = (−1)*

_{n}*(−Δ*

^{m}𝒥_{m}*) being the*

_{n}*m*th Bessel function of the first kind [33]. Similarly, the time-average of

*F*(

*t*,

*ϕ*, −Δ

_{n}_{−1}) =

*J*(

*t*)

*e*

^{−iΔn−1sinωt}in Eq. (3) reads

*F̄*(

*m*,

*ϕ*, − Δ

_{n}_{−1}), which is evaluated from Eq. (4) by using −Δ

_{n}_{−1}instead of Δ

*. For an even (odd)*

_{n}*n*,

*F̄*(

*m*,

*ϕ*, Δ

*) and*

_{n}*F̄*(

*m*,

*ϕ*, − Δ

_{n}_{−1}) are associated with

*the effective tunneling rates of the lattice separations a*(

*b*)

*and b*(

*a*), respectively. Clearly, they may be real or complex, corresponding to the even or odd

*m*. Given Eq. (4), Eq. (3) is transformed to

*A*and

_{e}*A*being the sums of even terms and odd terms respectively in the Fourier series. From the two first order equations of

_{o}*A*and

_{e}*A*we derive the second order equation

_{o}*Ä*(

*k*,

*t*) = −|

*f̄*(

*k*)|

^{2}

*A*(

*k*,

*t*) with the well-known general solution

*f̄*(

*k*) takes the form

*f̄*(

*k*)| and

*f̄*

^{*}(

*k*) are the corresponding modulus and complex conjugate,

*α*(

*k*) and

*β*(

*k*) are adjusted by the initial conditions. It should be noticed that the parameters

*J*

_{±}in Eq. (7) may be complex, while the same parameters in Ref. [33] are real. Such a difference may result in some different properties of the solutions. Without loss of generality, let the initially occupied state be |

*ψ*(

*t*

_{0}= 0)〉 = |

*N*〉 with a fixed integer

*N*, namely the initial conditions read

*A*(0) = 1,

_{N}*A*

_{n}_{≠}

*(0) = 0. Combining the conditions with the discrete Fourier transformation and general solution of*

_{N}*A*(

*k*,

*t*) produces

*α*(

*k*) and

*β*(

*k*) yields

*ψ*(

*t*

_{0})〉 and the different effective tunneling rates

*F̄*(

*m*,

*ϕ*, Δ

*) and*

_{n}*F̄*(

*m*,

*ϕ*, −Δ

_{n}_{−1}). In the general cases, the particle may be in the expanded states or localized states, depending on the system parameters.

## 3. Unusual transport phenomena

We are interested in the unusual transport phenomena such as the DL, CDT, instability and directed transport. It will be found that such unusual phenomena can be controlled under the initial condition |*ψ*(*t*_{0})〉 and for some special parameter sets with different phases. The routes for implementing the phase-controlled transport are very different from that of the previously considered case *δJ* = 0 with a constant tunneling rate *J*_{0} [23, 33].

*Phase-controlled CDT*. The CDT conditions mean the zero effective tunneling rates *F̄*(*m*, *ϕ*_{0}, Δ* _{n}*) =

*F̄*(

*m*,

*ϕ*

_{0}, −Δ

_{n}_{−1}) = 0 in Eq. (5), and

*f̄*(

*k*) = 0 in Eq. (7). Substituting the latter into Eq. (6), the probability amplitudes becomes some constants

*A*(

_{n}*t*) =

*A*(

_{n}*t*

_{0}) determined by the initial conditions, which means the occurrence of CDT. Applying the CDT conditions to Eq. (4), we get

*m*, and

*m*. Therefore, we can arrive at or deviate from the CDT conditions by fixing the parameters

*m*,

*J*

_{0},

*δJ*, Δ

*, Δ*

_{n}

_{n}_{−1}and adjusting the phase to arrive at or deviate from the phase

*ϕ*

_{0}. In the general case,

*𝒥*

_{0}(Δ

*) ≠*

_{n}*𝒥*

_{0}(Δ

_{n}_{−1}), for an even

*m*the above CDT conditions imply −

*δJ*cos

*ϕ/J*

_{0}=

*𝒥*

_{0}(Δ

*)/*

_{n}*𝒥*(Δ

_{m}*) =*

_{n}*𝒥*

_{0}(Δ

_{n}_{−1})/

*𝒥*(Δ

_{m}

_{n}_{−1}). For example, applying the parameters

*J*

_{0}= 1,

*δJ*= 0.8,

*m*= 2 to the CDT conditions produces the required values Δ

*≈ 2.01717, Δ*

_{n}

_{n}_{−1}≈ 5.37977. Adopting these parameters, we plot the effective tunneling rates as functions of phase, as in Fig. 2. It is shown that the effective tunneling rates are tunable by varying the phase, and the CDT conditions

*F̄*(

*m*,

*ϕ*

_{0}, Δ

*) =*

_{n}*F̄*(

*m*,

*ϕ*

_{0}, −Δ

_{n}_{−1}) = 0 are established at the phase

*ϕ*

_{0}≈ 2.4.

As a simplest example, we can fix the lattice separations *a*, *b* and tune the ratio *ε*_{0}/*ω* to obey *𝒥*_{0}(Δ* _{n}*) =

*𝒥*

_{0}(Δ

_{n}_{−1}) = 0 with ${\mathrm{\Delta}}_{n}=\frac{{\epsilon}_{0}}{\omega}a\approx 2.4048$, ${\mathrm{\Delta}}_{n-1}=\frac{{\epsilon}_{0}}{\omega}b\approx 5.5201$, then Eq. (4) becomes

*ϕ*

_{0}=

*π*/2 for an even

*m*or to

*ϕ*

_{0}= 0 for an odd

*m*in a nonadiabatic manner [23, 37].

*Phase-controlled DL and selective CDT*. The DL conditions mean one of the two effective tunneling rates vanishing, which leads to selective CDT to the two different barriers [23, 24]. When *F̄*(*m*, *ϕ*, −Δ_{n}_{−1}) = 0 and |*ψ*(0)〉 = |*N*〉 are set, from Eqs. (7) and (8) we obtain the constant modulus |*f̄*(*k*)| = |*F̄*(*m*, *ϕ*, Δ* _{n}*)| and the periodic functions

*k*. Inserting these into Eq. (6) produces the probability amplitudes

*N*〉 and |

*N*+ 1〉 with oscillating frequency

*ω*

_{1}= |

*F̄*(

*m*,

*ϕ*, Δ

*)|. Here the selective CDTs between the states |*

_{N}*N*− 1〉 and |

*N*〉, and between the states |

*N*+ 1〉 and |

*N*+ 2〉 occur. Similarly, taking

*F̄*(

*m*,

*ϕ*, Δ

*) = 0 leads to the probability amplitudes*

_{n}*N*〉 and |

*N*− 1〉 with oscillating frequency

*ω*

_{2}= |

*F̄*(

*m*,

*ϕ*, −Δ

_{N}_{−1})

*|*. Here the DL conditions and the different oscillating frequencies are modulated by the phase for a set of other parameters.

*Phase-controlled instability*. Now we prove that the solutions of Eq. (5) are unstable under the condition

*ϕ*=

*ϕ*. In fact, when Eq. (12) is satisfied, Eq. (5) can be written as with variable

_{c}*τ*= 2

*iF̄*(

*m*,

*ϕ*, Δ

*)*

_{N}*t*being proportional to time, whose general solution is well-known as where

*𝒥*(

_{n}*τ*) and

*𝒩*(

_{n}*τ*) are the Bessel and Neuman functions respectively, and

*B*,

_{n}*D*the expansion coefficients determined by means of the initial conditions. For the complex variable

_{n}*τ*with nonzero imaginary part, this general solution has the asymptotic property ${\text{lim}}_{t\to \infty}{A}_{n}(\tau )~{\text{lim}}_{t\to \infty}{e}^{\tau}/\sqrt{2\pi \tau}=\infty $ of the Bessel function, which implies that the initially small deviation

*δA*(

_{n}*τ*) from the given solution

*A*(

_{n}*τ*) can grow exponentially fast, meaning the Lyapunov instability of solution

*A*(

_{n}*τ*). In fact, by using

*δA*(

_{n}*τ*) +

*A*(

_{n}*τ*) instead of

*A*(

_{n}*τ*) in the linear Eq. (5), we can find that the deviated solution possesses the same form as that of the given solution,

*δB*and

_{n}*δD*determined by the initial deviation, and has the same asymptotic property, namely lim

_{n}

_{t}_{→∞}

*δA*(

_{n}*τ*) tends to infinity exponentially fast. Obviously such an instability can be controlled by tuning the phase to arrive at or deviate from

*ϕ*in the condition (12), as shown in Fig. 3(a).

_{c}*Phase-controlled directed transport*. For a set of given parameters *J*_{0}, *δJ*, Δ* _{n}* and Δ

_{n}_{−1}, we define two different phases

*ϕ*

_{1}and

*ϕ*

_{2}to obey

*m*, which result in the selective CDT between the localized states |

*n*〉 and |

*n*− 1〉 or between the localized states |

*n*〉 and |

*n*+ 1〉, respectively. By nonadiabatically tuning the phase to alternately change between

*ϕ*

_{1}and

*ϕ*

_{2}just after the two different time intervals

*T*

_{1}=

*π/ω*

_{1}and

*T*

_{2}=

*π/ω*

_{2}, the original tunneling rate becomes the continuous and piecewise analytic function

*T*=

*T*

_{1}+

*T*

_{2}and

*n′*= 0, 1, 2,.... The instability condition (12) will be reached in each process changing phase from

*ϕ*to

_{i}*ϕ*for

_{j}*i*,

*j*= 1, 2. As an example, this is indicated by

*ϕ*in Fig. 3(a) for the parameter set

_{c}*J*

_{0}= 1,

*δJ*= 0.8,

*m*= 2,

*ω*= 30, Δ

*= 2, Δ*

_{n}

_{n}_{−1}= 2.2, where Eqs. (12) and (13) give

*ϕ*≈ 2.17,

_{c}*ϕ*

_{1}≈ 1.93 and

*ϕ*

_{2}≈ 2.49, and the Rabi frequencies and half-periods of Eqs. (10) and (11) read

*ω*

_{1}=

*F̄*(

*m*,

*ϕ*

_{1}, Δ

*) ≈ 0.124666,*

_{n}*ω*

_{2}= |

*F̄*(

*m*,

*ϕ*

_{2}, −Δ

_{n}_{−1})| ≈ 0.140933 and

*T*

_{1}≈ 25.2001,

*T*

_{2}≈ 22.2914, respectively. The corresponding adjustment to the time-dependent tunneling rate

*J*(

*t*,

*ϕ*) = 1 + 0.8cos(60

*t*+

*ϕ*) is exhibited in Fig. 3(b) for the time interval including

*t*=

*T*

_{1}, which only transforms the phase of

*J*(

*t*,

*ϕ*) from

*ϕ*

_{1}to

*ϕ*

_{2}and does not change its magnitude. Clearly, at the time

*t*=

*T*

_{1}+

*T*

_{2}, the tunneling rate will change from

*J*(

*t*,

*ϕ*

_{2}) to

*J*(

*t*,

*ϕ*

_{1}). By repeatedly using such operations, the initially stable Rabi oscillation is broken under the conditions (12), then the instability is suppressed by the conditions (13) with phase

*ϕ*

_{1}or

*ϕ*

_{2}such that the particle is forced to transit repeatedly between the two stable oscillation states |

*ψ*(

_{n}*t*)〉 =

*A*|

_{n}*n*〉 +

*A*

_{n}_{+1}|

*n*+ 1〉 and |

*ψ*(

_{n′}*t*)〉 =

*A*|

_{n′}*n′*〉 +

*A*

_{n′}_{+1}|

*n′*+ 1〉 with the amplitudes and frequencies of Eqs. (10) and (11) for (

*n*,

*n′*) = (

*N*,

*N*+1);(

*N*+1,

*N*+2);... that lead to the directed motion toward the right [23,33]. Because the phase-transformation does not change the magnitude of

*J*(

*t*,

*ϕ*) at the operation moment, it supplies a more convenient method to manipulate the directed transport, compared to the previous amplitude-modulation schemes to the coupling [37] or to the tilt [23].

## 4. Extension to a many-particle system

It is worth noting that the method realizing the directed transport can be extended to controlling the transport of a many-particle system in an optical bipartite lattice, where a Bose-Hubbard interaction energy
${H}_{b}=\frac{1}{2}{U}_{0}{\sum}_{n}\widehat{n}(\widehat{n}-1)$ for
$\widehat{n}={b}_{n}^{\u2020}{b}_{n}$ should be added into Eq. (1). It is well known that for an undriven simple lattice system with *J* = *J*_{0}, *ε* = *ε*_{0}, *a* = *b* and the interaction strength *U*_{0} > 0, the characteristic parameter is the ratio [39, 40] *r* = *U*_{0}/*J*_{0}. For *U*_{0} ≪ *J*_{0} the ground state of the system describes a superfluid, whereas it has the properties of a Mott insulator for *U* ≫ *J*_{0}. Quantum transition between the superfluid and Mott insulator can occur at the critical value *r* = *r _{c}*. It has been argued that in the presence of a high-frequency driving the system behaves similar to the undriven system, but with the tunneling rate

*J*

_{0}of the latter being replaced by the effective tunneling rate

*J*=

_{eff}*J*

_{0}

*𝒥*

_{0}(

*ε*

_{0}/

*ω*) for the simple lattice system. Thus one can control the quantum transition and transport by adjusting the driving parameters [39, 40]. Another interesting scheme for controlling the quantum transition has also been suggested in which the periodically modulated interactions were applied [52]. In the Mott insulating state, average atomic current along any direction approximately vanishes.

However, in the case of many particles held in the considered optical bipartite lattice, the effective tunneling rates of the lattice separations *a* and *b* can be different, *F̄*(*m*, *ϕ*, Δ* _{n}*) ≠

*F̄*(

*m*,

*ϕ*, −Δ

_{n}_{−1}). After transformation into the interaction picture by the unitary operator

*= Δ*

_{ij}*and −Δ*

_{n}

_{n}_{−1}alternately. In such a system, we have two parameter ratios

*r*

_{1}=

*U*

_{0}/

*F̄*(

*m*,

*ϕ*, Δ

*) and*

_{n}*r*

_{2}=

*U*

_{0}/

*F̄*(

*m*,

*ϕ*, −Δ

_{n}_{−1}), whose values are associated with the following cases:

*Case 1*. The system becomes a Mott insulator with*r*_{1}≫*r*and_{c}*r*_{2}≫*r*;_{c}*Case 2*. There is a superfluid state with*r*_{1}≪*r*and_{c}*r*_{2}≪*r*;_{c}*Case 3*. There exist two different states for*r*_{1}≫*r*,_{c}*r*_{2}<*r*or_{c}*r*_{1}<*r*,_{c}*r*_{2}≫*r*, respectively._{c}

We call the two states the local superfluid states, which may be gone through in the process of transformations between case 1 and case 2.

When the ratios *r _{i}* are changed between the case 1 and case 2, the system undergoes a quantum transition between the superfluid and Mott insulator. Because the ratios depends on the effective tunneling rates and the latter can be tuned by varying the driving phase, we can control the quantum transition by the phase-modulation. On the other hand, to realize the directed transport of the many-particle system, the possible experiment can begin by loading a Bose-Einstein condensate into the long lattice of wave-vector

*k*, then one can increase the lattice depth to make the atomic sample in the Mott insulating state with a single atom per well [42,43]. Further one can divide every well into a double-well by ramping up the short lattice of wave-vector 2

_{L}*k*and tilting the double-well train, that achieve the load of single atoms into the “left” sides of tilted double-wells [43]. In such a state with a single atom per double-well such that the interatomic interaction could be negligible and the phase-modulation method for the single atom system becomes valid for the many-body system. Therefore, we can tune the driving phase according to Eqs. (13) and (14) such that the parameter ratios change alternately between

_{L}*r*

_{1}<

*r*,

_{c}*r*

_{2}≫

*r*and

_{c}*r*

_{1}≫

*r*,

_{c}*r*

_{2}<

*r*, leading to the nonzero average atomic current along a sin-gle direction. Such a phase-controlled directed transport of many particles will form a stronger particle current compared to the single particle case. Particularly, the dynamics of the super-fluid and Mott insulator is very different from that of the single-atom case. The scheme of the phase-controlled quantum transition and directed transport through the local superfluid states also differs from the previous amplitude-modulation [39, 40] and interaction-modulation [52] proposals on controlling quantum transitions and transports in simple optical lattices. In the latter case, the local superfluid states and the directed motion cannot exist for a symmetric driving. It is interesting for us to compare our results with those of a repulsive Bose-Einstein condensate loaded in the spatially asymmetric optical superlattice without tilt potential and treated in mean-field approximation [50]. Differing from the quantum states governed by the spatially discrete and linear Eq. (2), the governing Schrödinger equation of the latter system are spatially continuous and nonlinear. Therefore, quantum transition and directed transport of the former system can be described by the ratios of the interaction strength and the effective tunneling rates associated with the time-periodic lattice tilt, while the directed atomic current of the latter system is expressed by the time-averaged momentum expectation and this transport phenomenon can occur only if the interaction strength exceeds a critical value for a symmetric driving [50]. Thus one can steer the directed current by adjusting the

_{c}*s*-wave scattering length to vary the interaction strength.

## 5. Conclusions and discussion

We have investigated the coherent control of a single atom held in the optical bipartite lattice with two different separations *a* and *b* and driven by a combined modulation of two resonant external fields with a phase difference between the bias and coupling. In the high-frequency regime and NNTB approximation, we derive an analytical general solution of the time-dependent Schrödinger equation, which quantitatively describes the dependence of the tunneling dynamics on the phase difference between the modulation components. It is demonstrated that a new route of CDT (or DL) can be formed by tuning the phase to make two (or one) of the effective tunneling rates of the lattice separations *a* and *b* vanishing. When the two effective tunneling rates are adjusted to go through the values of the same magnitude and opposite signs, the system loses its stability. The phase-controlled selective CDT enables the system to be stabilized and the directed tunneling of the particle to be coherently manipulated. In the process of control, the appropriate operation times are fixed by the two tunneling half-periods. The theoretical results have also been extended to the phase-controlled directed transport and quantum transition between the superfluid and Mott insulator for the corresponding many-particle system. The analytic results based on the high-frequency approximation should have an advantage over a direct numerical integration of the Schröedinger equation for transparently predicting or explaining the experimental results.

The obtained results can be tested by employing the current accessible experimental setups [39, 40, 42, 43], and may be applicable for investigating atomic transport, quantum transition and quantum information processing [37, 38]. In experiments, the periodic modulations of the tunneling rate [37, 43, 44] and the potential tilt [12, 39, 51] have been realized through different methods. In order to experimentally implement the directed transport of the bipartite lattice systems, the effective tunneling rate must be changed alternately between two different values. The previous schemes [23,37] suggested that one can achieve the modulation by rapidly changing the amplitudes of driving fields. In such a process, the deviation from the probability amplitude of the initial Rabi-oscillation state crossing the separation *a* linearly grows in time, then the system is stabilized to the final oscillation state crossing the separation *b*. This mechanism is similar to the resonance transition in standard quantum mechanics [33]. In contrast, when the phase-modulation is adopted, the deviation from the the initial state can grow exponentially fast, as indicated below Eq. (12), that may result in faster transition to the final state. The difference between both implies that the phase-modulation offers new dynamics of tunneling. In addition, in Fig. 3 we show that the phase-transformation does not change the magnitude of the original tunneling rate at the operation moment. Therefore, the phase-modulation scheme of the periodic shaking may be more convenient and experimentally feasible compared to the previous amplitude modulations. The directed tunneling is related to the ratchetlike effect of quantum particles in the periodically tilted and shaken optical bipartite lattices and could be well suited to simulating the different optical systems [16,17] and solid-state systems [13]. Particularly, the optical analog of such a system may be realized by an optical two-mode quantum beam splitter [35].

## Acknowledgments

This work was supported by the NNSF of China under Grant Nos. 11204027, 11175064 and 11205021, the Construct Program of the National Key Discipline of China, the Hunan Provincial NSF ( 11JJ7001) and the Scientific Research Fund of Hunan Provincial Education Department ( 12B082).

## References and links

**1. **H. Rabitz, R. de Vivie-Riedle, M. Motzkus, and K. Kompa, “Whither the future of controlling quantum phenomena?” Science **288**, 824–828 (2000). [CrossRef] [PubMed]

**2. **M. Grifoni and P. Hänggi, “Driven quantum tunneling,” Phys. Rep. **304**, 229–354 (1998). [CrossRef]

**3. **P. Král, I. Thanopulos, and M. Shapiro, “Coherently controlled adiabatic passage,” Rev. Mod. Phys. **79**, 53–77 (2007). [CrossRef]

**4. **J. Thom, G. Wilpers, E. Riis, and A. G. Sinclair, “Accurate and agile digital control of optical phase, amplitude and frequency for coherent atomic manipulation of atomic systems,” Opt. Express **21**, 18712–18723 (2013). [CrossRef] [PubMed]

**5. **D. H. Dunlap and V. M. Kenkre, “Dynamic localization of a charged particle moving under the influence of an electric field,” Phys. Rev. B **34**, 3625–3633 (1986). [CrossRef]

**6. **F. Grossmann, T. Dittrich, P. Jung, and P. Hänggi, “Coherent destruction of tunneling,” Phys. Rev. Lett. **67**, 516–519 (1991). [CrossRef] [PubMed]

**7. **F. Grossmann and P. Hänggi, “Localization in a driven two-level dynamics,” Europhys. Lett. **18**, 571–576 (1992). [CrossRef]

**8. **M. Holthaus, “Collapse of minibands in far-infrared irradiated superlattices,” Phys. Rev. Lett. **69**, 351–354 (1992). [CrossRef] [PubMed]

**9. **H. Lignier, C. Sias, D. Ciampini, Y. Singh, A. Zenesini, O. Morsch, and E. Arimondo, “Dynamical control of matter-wave tunneling in periodic potentials,” Phys. Rev. Lett. **99**, 220403, 2007). [CrossRef]

**10. **G. Della Valle, M. Ornigotti, E. Cianci, V. Foglietti, P. Laporta, and S. Longhi, “Visualization of coherent destruction of tunneling in an optical double well system,” Phys. Rev. Lett. **98**, 263601, 2007). [CrossRef] [PubMed]

**11. **E. Kierig, U. Schnorrberger, A. Schietinger, J. Tomkovic, and M. K. Oberthaler, “Single-particle tunneling in strongly driven double-well potentials,” Phys. Rev. Lett. **100**, 190405, 2008). [CrossRef] [PubMed]

**12. **A. Eckardt, M. Holthaus, H. Lignier, A. Zenesini, D. Ciampini, O. Morsch, and E. Arimondo, “Exploring dynamic localization with a Bose-Einstein condensate,” Phys. Rev. A **79**, 013611, 2009). [CrossRef]

**13. **J. M. Villas-Boas, S. E. Ulloa, and N. Studart, “Selective coherent destruction of tunneling in a quantum-dot array,” Phys. Rev. B **70**, 041302(R) (2004). [CrossRef]

**14. **C. Weiss and N. Teichmann, “Differences between mean-field dynamics and N-particle quantum dynamics as a signature of entanglement,” Phys. Rev. Lett. **100**, 140408, 2008). [CrossRef] [PubMed]

**15. **C. Ding, J. Li, R. Yu, X. Hao, and Y. Wu, “High-precision atom localization via controllable spontaneous emission in a cycle-configuration atomic system,” Opt. Express **20**, 7870–7885 (2012). [CrossRef] [PubMed]

**16. **I. L. Garanovich, S. Longhi, A. A. Sukhorukov, and Y. S. Kivshar, “Light propagation and localization in modulated photonic lattices and waveguides,” Phys. Rep. **518**, 1–79 (2012). [CrossRef]

**17. **F. Dreisow, Y. V. Kartashov, M. Heinrich, V. A. Vysloukh, A. Tünnermann, S. Nolte, L. Torner, S. Longhi, and A. Szameit, “Spatial light rectification in an optical waveguide lattice,” Europhys. Lett. **101**, 44002, 2013). [CrossRef]

**18. **R.-G. Wan and T.-Y. Zhang, “Two-dimensional sub-half-wavelength atom localization via controlled spontaneous emission,” Opt. Express **19**, 25823–25832 (2011). [CrossRef]

**19. **Q. T. Xie, X. B. Luo, and B. Wu, “All-optical switch with two periodically modulated nonlinear waveguides,” Opt. Lett. **35**, 321–323 (2010). [CrossRef] [PubMed]

**20. **X. B. Luo, Q. T. Xie, and B. Wu, “Nonlinear coherent destruction of tunneling,” Phys. Rev. A **76**, 051802(R) (2007). [CrossRef]

**21. **J. Gong, L. Morales-Molina, and P. Hänggi, “Many-body coherent destruction of tunneling,” Phys. Rev. Lett. **103**, 133002, 2009). [CrossRef] [PubMed]

**22. **S. Longhi, “Many-body selective destruction of tunneling in a bosonic junction,” Phys. Rev. A **86**, 044102, 2012). [CrossRef]

**23. **C. E. Creffield, “Quantum control and entanglement using periodic driving fields,” Phys. Rev. Lett. **99**, 110501, 2007). [CrossRef] [PubMed]

**24. **K. Hai, W. Hai, and Q. Chen, “Controlling transport and entanglement of two particles in a bipartite lattice,” Phys. Rev. A **82**, 053412, 2010). [CrossRef]

**25. **G. Lu, W. Hai, and H. Zhong, “Quantum control in a double-well with symmetric or asymmetric driving,” Phys. Rev. A **80**, 013411, 2009). [CrossRef]

**26. **N. Singh, “Phase controllable dynamical localization of a quantum particle in a driven optical lattice,” Phys. Lett. A **376**, 1593–1595 (2012). [CrossRef]

**27. **W. Hai, K. Hai, and Q. Chen, “Transparent control of an exactly solvable two-level system via combined modulations,” Phys. Rev. A **87**, 023403, 2013). [CrossRef]

**28. **A. Greilich, S. E. Economou, S. Spatzek, D. R. Yakovlev, D. Reuter, A. D. Wieck, T. L. Reinecke, and M. Bayer, “Ultrafast optical rotations of electron spins in quantum dots,” Nat. Phys. **5**, 262–266 (2009). [CrossRef]

**29. **C. Padurariu and Y. V. Nazarov, “Spin blockade qubit in a superconducting junction,” EPL **100**, 57006, 2012). [CrossRef]

**30. **G. Lu and W. Hai, “Quantum tunneling switch in a planar four-well,” Phys. Rev. A **83**, 053424, 2011). [CrossRef]

**31. **K. Xiao, W. Hai, and J. Liu, “Coherent control of quantum tunneling in an open double-well system,” Phys. Rev. A **85**, 013410, 2012). [CrossRef]

**32. **C. E. Creffield, “Instability and control of a periodically driven Bose-Einstein condensate,” Phys. Rev. A **79**, 063612, 2009). [CrossRef]

**33. **K. Hai, Q. Chen, and W. Hai, “Instability inducing directed tunnelling of a single particle in a bipartite lattice,” J. Phys. B At. Mol. Opt. Phys. **44**, 035507, 2011). [CrossRef]

**34. **N. Korneev, V. A. Vysloukh, and E. M. Rodríguez, “Propagation dynamics of weakly localized cnoidal waves in dispersion-managed fiber: from stability to chaos,” Opt. Express **11**, 3574–3582 (2003). [CrossRef] [PubMed]

**35. **J. Sebby-Strabley, B. L. Brown, M. Anderlini, P. J. Lee, W. D. Phillips, J. V. Porto, and P. R. Johnson, “Preparing and probing atomic number states with an atom interferometer,” Phys. Rev. Lett. **98**, 200405, 2007). [CrossRef] [PubMed]

**36. **S. Trotzky, P. Cheinet, S. Folling, M. Feld, U. Schnorrberger, A. M. Rey, A. Polkovnikov, E. A. Demler, M. D. Lukin, and I. Bloch, “Time-resolved observation and control of superexchange interactions with ultracold atoms in optical lattices,” Science **319**, 295–299 (2008). [CrossRef]

**37. **O. Romero-Isart and J. J. García-Ripoll, “Quantum ratchets for quantum communication with optical superlat-tices,” Phys. Rev. A **76**, 052304, 2007). [CrossRef]

**38. **G. De Chiara, T. Calarco, M. Anderlini, S. Montangero, P. J. Lee, B. L. Brown, W. D. Phillips, and J. V. Porto, “Optimal control of atom transport for quantum gates in optical lattices,” Phys. Rev. A **77**, 052333, 2008). [CrossRef]

**39. **A. Zenesini, H. Lignier, D. Ciampini, O. Morsch, and E. Arimondo, “Coherent control of sressed matter waves,” Phys. Rev. Lett. **102**, 100403, 2009). [CrossRef]

**40. **A. Eckardt, C. Weiss, and M. Holthaus, “Superfluid-insulator transition in a periodically driven optical lattice,” Phys. Rev. Lett. **95**, 260404, 2005). [CrossRef]

**41. **F. Massel, M. J. Leskinen, and P. Törmä, “Hopping modulation in a one-dimensional Fermi-Hubbard Hamiltonian,” Phys. Rev. Lett. **103**, 066404, 2009). [CrossRef] [PubMed]

**42. **R. Ma, M. E. Tai, P. M. Preiss, W. S. Bakr, J. Simon, and M. Greiner, “Photon-assisted tunneling in a biased strongly correlated Bose gas,” Phys. Rev. Lett. **107**, 095301, 2011). [CrossRef] [PubMed]

**43. **Y.-A. Chen, S. Nascimbéne, M. Aidelsburger, M. Atala, S. Trotzky, and I. Bloch, “Controlling correlated tunneling and superexchange interactions with ac-driven optical lattices,” Phys. Rev. Lett. **107**, 210405, 2011). [CrossRef] [PubMed]

**44. **C. Schori, T. Stöferle, H. Moritz, M. Köhl, and T. Esslinger, “Excitations of a superfluid in a three-dimensional optical lattice,” Phys. Rev. Lett. **93**, 240402, 2004). [CrossRef]

**45. **F. T. Hioe and C. E. Carroll, “Two-state problems involving arbitrary amplitude and frequency modulations,” Phys. Rev. A **32**, 1541–1549 (1985). [CrossRef] [PubMed]

**46. **Y. Qian, M. Gong, and C. Zhang, “Quantum transport of bosonic cold atoms in double-well optical lattices,” Phys. Rev. A **84**, 013608, 2011). [CrossRef]

**47. **S. Longhi and G. D. Valle, “Quantum transport in bipartite lattices via Landau-Zener tunneling,” Phys. Rev. A **86**, 043633, 2012). [CrossRef]

**48. **A. P. Itin and A. I. Neishtadt, “Directed transport in a classical lattice with a high-frequency driving,” Phys. Rev. E **86**, 016206, 2012). [CrossRef]

**49. **D. Poletti, T. J. Alexander, E. A. Ostrovskaya, B. Li, and Yuri S. Kivshar, “Dynamics of matter-wave solitons in a ratchet potential,” Phys. Rev. Lett. **101**, 150403, 2008). [CrossRef] [PubMed]

**50. **D. Poletti, G. Benenti, G. Casati, P. Hänggi, and B. Li, “Steering Bose-Einstein condensates despite time symmetry,” Phys. Rev. Lett. **102**, 130604, 2009). [CrossRef] [PubMed]

**51. **K. W. Madison, M. C. Fischer, R. B. Diener, Q. Niu, and M. G. Raizen, “Dynamical Bloch band suppression in an optical lattice,” Phys. Rev. Lett. **81**, 5093–5096 (1998). [CrossRef]

**52. **Á. Rapp, X. Deng, and L. Santos, “Ultracold lattice gases with periodically modulated interactions,” Phys. Rev. Lett. **109**, 203005, 2012). [CrossRef] [PubMed]