Chiral spin currents in a trappedion quantum simulator using Floquet engineering
Abstract
The most typical ingredient of topologically protected quantum states are magnetic fluxes. In a system of spins, complexvalued interaction parameters give rise to a flux, if their phases do not add up to zero along a closed loop. Here we apply periodic driving, a powerful tool for quantum engineering, to a trappedion quantum simulator in order to generate such spinspin interactions. We consider a simple driving scheme, consisting of a repeated series of locally quenched fields, and demonstrate the feasibility of this approach by studying the dynamics of a small system. An emblematic hallmark of the flux, accessible in experiments, is the appearance of chiral spin currents. Strikingly, we find that in parameter regimes where, in the absence of fluxes, phonon excitations dramatically reduce the fidelity of the spin model simulation, the spin dynamics remains widely unaffected by the phonons when fluxes are present.
Magnetic fluxes can profoundly alter the behavior of a quantum system. They are responsible for the AharonovBohm effect (Aharonov and Bohm (1959); Shapere and Wilczek (1989)), fractal energy spectra such as the Hofstadter butterfly (Hofstadter (1976); Janssen et al. (1994)), or intriguing manybody phases exhibiting the fractional quantum Hall effect (Laughlin (1983); Wen (2004)). Most importantly, properties of a quantum system can be topologically protected in the presence of fluxes, as for example is the case for quantized Hall conductances. The underlying mechanism is due to breaking of timereversal symmetry which fixes the chirality of the edge currents. Thereby, backscattering is suppressed, giving rise to robust values for the conductance. An exciting application of topological protection is faulttolerant quantum computing using topologically protected states Nayak et al. (2008). In the present paper we will show that a magnetic flux can enhance the fidelity of a quantum simulation even in the fewbody regime.
Quantum simulation of topological matter has been pursued very actively using cold atoms Struck et al. (2012); Miyake et al. (2013); Aidelsburger et al. (2013); Jotzu et al. (2014); Mancini et al. (2015); Stuhl et al. (2015); Meinert et al. (2016), photons in cavities Hafezi et al. (2013); Schine et al. (2016); P. Roushan et al. (2017); Lodahl et al. (2017), or nitrogenvacancies Kong et al. (2016). Another experimental platform with longstanding history of quantum simulations involves trapped ions: Refs. Mintert and Wunderlich (2001); Porras and Cirac (2004) suggested the implementation of spin models, which have later been realized in the laboratory using linear Paul traps Friedenauer et al. (2008); Kim et al. (2009, 2010); Richerme et al. (2014); Jurcevic et al. (2014), or twodimensional Penning traps Britton et al. (2012); Bohnet et al. (2016). Nowadays trapped ions are the leading platform for quantum simulation of spin models Zhang et al. (2017). They offer highefficiency detection and ion addressability at the singlesite level Nägerl et al. (1999); Smith et al. (2016), enabling full state tomography Roos et al. (2004) and measurement of high order body correlators Jurcevic et al. (2016).
A particular interesting feature of trapped ions is the longrange character of spinspin interactions. This can be exploited to engineer magnetic fluxes even in a 1D architecture Graß et al. (2015), for instance by having complex phases only on secondneighbor interactions. A standard route to artificial fluxes is Floquet engineering, that is the control of a Hamiltonian via periodic driving. Early proposals for modifying the tunneling strength in quantum wells using timeperiodic fields date back to the 1980s and 1990s Dunlap and Kenkre (1986); Grossmann et al. (1991); Holthaus (1992); Holthaus and Hone (1993). Since then the progress in laser cooling and the emergence of cold atoms as highly controllable quantum systems have provided an experimental platform to apply these ideas. Controlling the tunneling of atoms in an optical lattice by periodic modulation was proposed in Refs. Eckardt et al. (2005, 2010) and realized in Refs. Lignier et al. (2007); Struck et al. (2011). The technique then became a widespread tool for implementing artificial gauge potentials in atomic systems Struck et al. (2012); Aidelsburger et al. (2013); Jotzu et al. (2014); Meinert et al. (2016). More generally, Floquet engineering has been recognized as a strategy for producing topological phases of matter such as topological insulators and Majorana fermions Kitagawa et al. (2010); Jiang et al. (2011); Lindner et al. (2011).
In this paper we study the feasibility of Floquet engineering in a system of trapped ions Bermudez et al. (2011); Graß et al. (2015, 2016). In particular, we are interested in microscopic signatures of complexvalued spinspin interactions. To this end, we consider a minimal system of three ions in a linear Paul trap. Similar to the scheme proposed in Ref. Graß et al. (2015), a timeperiodic series of local quenches is applied to the spins, that is, local potentials are repeatedly switched on and off. Such shaking protocol is able to approximately equalize all interaction strengthes, and to render interaction parameters complex. With spinspin interactions having approximate XX symmetry (via the presence of a transverse field), the dynamics can be described as a single particle (single hole) hopping along a triangle pierced by a magnetic flux. An immediate consequence of the flux, seen in the spin dynamics, is the chirality of the current.
While the description of the system in terms of effective spin model is rather simple and the effective Floquet Hamiltonian can be evaluated exactly, the real dynamics of three ions is actually much more complex. Indeed, apart from the spin dynamics, there are also vibrational degrees of freedom, which are only approximately decoupled as long as spinphonon interactions are sufficiently fast. On the other hand, also the shaking of the spin model must be fast compared to the spin dynamics in order to get the desired Floquet Hamiltonian, but it should be slow compared to the spinphonon coupling. To verify that this separation of time scales can be achieved, we have performed a simulation of the dynamics in the Dicke model, which includes also vibrational degrees of freedom. Strikingly, this simulation does not only show that Floquet engineering is feasible, but we also find that phonon effects, which reduce the fidelity of the quantum simulation, are suppressed through the presence of magnetic fluxes. This effect may be a signature of topological protection of chiral states in the full Hamiltonian governing phonons and spins.
The starting point of our study is a Dickelike Hamiltonian
(1) 
describing the collective motion of the ions in terms of phonons, created by at frequencies , coupled to two internal states of the ions. Note that we work in the (fastly) rotating frame of the atomic transition, and apply rotatingwave approximation. Raman lasers with Rabi frequency and beatnote frequency induce spin flips, described by a Pauli matrix at an individual ion denoted by , and excites or deexcites phonon modes. The strength of this coupling to the light further depends on the LambDicke parameters . It is wellknown Mintert and Wunderlich (2001); Porras and Cirac (2004) that, via a unitary transformation, the spin degrees of freedom can be decoupled from “dressed” phonon operators. By means of this decoupling, the time evolution of the ionic system for times longer than is captured by an effective Ising spin model Kim et al. (2009); Wang and Freericks (2012), , with couplings , controlled by the detuning .
For the shaking, we drive the system with a transverse magnetic field term . The homogeneous and timeindependent term provides an approximate XX symmetry to the Ising model. The inhomogeneous timedependent one, implements the shaking protocol. Note that on the level of the Dicke model, Eq. (1), a transverse field leads to additional entanglement between spins and phonons Wang and Freericks (2012); Wall et al. (2017).
To derive the effective Hamiltonian, we note that . Thus, we may first switch to the interaction picture of through the transformation . Applying the rotatingwave approximation, we get an effective XX model, , with H.c., where . Then, we switch to the interaction picture of the timedependent part , via , where . The final Hamiltonian again has XX structure but with timedependent couplings. By timeaveraging over a period , that is, in first order of the Floquet expansion, we get a timeindependent XX Hamiltonian with effective couplings given by ^{1}^{1}1Here, for simplicity we have assumed sufficiently larger than the Ising couplings such that the net effect of driving is equivalent to the driving of a XX model by the timedependent magnetic field . More generally, the condition to achieve an effective XX model with the complex coupling as in Eq. (2) is , where is the period of .
(2) 
We now apply this general formalism to a concrete driving scheme, similar to the one proposed in Ref. Graß et al. (2015). Each potential takes piecewise constant values, being integer multiples of some frequency . Here, provides the time unit, with being an integer multiple of . This choice is motivated in the following way:

If two potentials remain constant during an interval , the interaction between the corresponding spins remains unchanged if the two potentials are the same, or is fully suppressed if the potentials are different, . This feature allows for engineering the effective strength of interactions.

If a potential changes within an interval , this gives rise to complex interaction parameters. For concreteness, let us assume that two potentials differ by for a time , with an integer. The potentials shall then drop to zero simultaneously for an interval of length , and finally return to their original values for an interval of length . For such sequence the first order Floquet analysis yields
(3) This feature allows for generating complex phases, needed to simulate artificial magnetic flux.
According to the reasoning explained above, the shaking period shown in Fig. 1(a) should (i) generate a complex phase on the link between spin 1 and 3, and (ii) enhance the ratio by a factor 2 compared to the original ratio . This can fully compensate the decay of interaction strength with distance, and lead to approximately equal interactions between all pairs, cf. Fig. 1(b).
The reasoning so far was based on the assumption that timeaveraging in the interaction picture approximates well the effective Hamiltonian. Rigorously, though, the effective Hamiltonian is obtained from Floquet’s theorem. Due to timeperiodicity of the evolution operator , the operator fully determines the evolution at stroboscopic times . Writing , we obtain an effective Hamiltonian , which exactly describes the stroboscopic dynamics. For our protocol, consisting only of quenches between timeconstant Hamiltonians, can straightforwardly be evaluated. As seen in Fig. 1(c), the discrepancy between the couplings in the exact and the couplings approximated according to Eq. (2), regarding both absolute value and phase, decreases as with the shaking strength. Relative errors require a shaking , where denotes the root mean square of the spinspin interactions before shaking. We also find that, in order to establish an approximate XX symmetry, a homogeneous is sufficient.
Fig. 1(d) compares the spin dynamics in the Ising and the XX model in the presence of a flux () after initially preparing the system in a state along . The curves show good quantitative agreement. Small wiggles which are only seen in the dynamics of the Ising model are due to the strong transverse fields, but do not appear in the stroboscopic evolution. Both models exhibit clear evidence of a chiral current, as the upspin moves counterclockwise from ion 1, to ion 3, and finally to ion 2. Notably, the evolution is almost periodic with a period , so comparing the states at time and allows for a practical detection of timereversal symmetry breaking P. Roushan et al. (2017).
These results establish that Floquet engineering works sufficiently well on the level of an effective Ising system, if shaking strength is at least an order of magnitude faster than the spin interactions ( before shaking, after shaking). However, the validity of the spin model description also requires the detuning of the spinphonon coupling to be fast. A realistic choice is , i.e. one order of magnitude above . To explicitly check the role played by phonons we have simulated the dynamics under the Dicke Hamiltonian using Krylov methods. The results are seen in Fig. 2 for different fluxes. We first note that the evolution for , shown in panels (a,b), exhibits timereversal symmetry breaking in the same way as discussed before for the pure spin model. As expected, the direction of the spin current depends on the sign of the flux. For other fractional values of the flux, such as shown in panel (c), the evolution is not periodic on relevant time scales, but breaking of timereversal symmetry can still be infered from obvious differences between clockwise and counterclockwise flow. In contrast, for “integer” fluxes (i.e. ) shown in panel (d), the spin flows, at least initially, simultaneously clockwise and counterclockwise at equal rate.
Strikingly, timereversal symmetry breaking seems to stabilize the quantum simulation. For fractional fluxes [panel (a–c)], the evolution in the Dicke model agrees well with the evolution in the Ising model. In contrast to this, the spin dynamics in the system with integer fluxes [panel (d)] is heavily perturbed by the phonons, although only a single parameter has been changed, see also panel (e) plotting the error vs. time. This effect can not be traced to the number of phonons which is approximately the same for different values of . We also verified that the effect is visible for different shaking protocols expected to lead to the same effective dynamics. For instance, it appears even stronger if the complex phase is engineered between nearest neighbors rather than second neighbors. While we have no microscopic understanding of the origin of this effect, timereversal symmetry breaking seems to provide a protection even in our minimal example involving only three ions. We have also carried out analog simulations for a system of four ions, forming two triangles glued together. Differently from the three ions setting, in this case the spin flip hops on more neighbouring sites at once but the dynamics still exhibits clear signs of chiral flow in the presence of a flux. As for three ions, the flux reduces the deviations between spin model dynamics and Dicke model dynamics, although this effect is found to be less pronounced for four ions.
So far, we have studied the dynamics seen when the system is initially prepared out of equilibrium. Quantum revivals of the initial state have allowed for demonstrating timereversal symmetry breaking. However, to keep revival periods sufficiently short, the eigenenergies of states contributing to the initial configuration have to be commensurate. While this will always be the case for a triangle at (half)integer flux [see the energy spectrum as a function of the flux in Fig. 3(b)], revival periods might get long for other fluxes [cf. flux in Fig. 2(c)] or in larger systems. In such cases, a more general criterion for timereversal symmetry breaking and chirality might be required. One possibility is to consider equilibrium chiral currents, quantified as P. Roushan et al. (2017)
(4) 
for an eigenstate . As seen in Fig. 3(a), equilibrium chiral currents are nonzero for any eigenstate, if the flux is not an integer of . At integer fluxes degenerate eigenstates exist, whose chiral currents add up to zero, and also the current of the nondegenerate eigenstate is zero.
In our previous considerations, the chirality of the spin current served as a direct hallmark of the flux. However, this can barely be used to determine the flux also quantitatively. In principle, though, trapped ions allow for state tomography providing the full density matrix of the system. This data could then be used to reveal quantitatively the value of the flux felt by the ions. We will now discuss how, for a slightly simplified configuration, information about a complex phase can be extracted already by measuring a single spinspin correlation function.
Therefore, we consider a shaking protocol of period , with always detuned from and which are both set to zero for , while taking values zero and at other times. This protocol leads to particularly simple dynamics, as it freezes out the center spin. Thus, with () denoting the position of the upspin on the left (right) end, we obtain a twolevel or doublewell system described by Accordingly, as seen in Fig. 4(a), the spin dynamics in this case reduces to Rabi oscillations between the left and the right spin with a period . The complex phase of is reflected in the relative phase between the two levels. An initial state will evolve to , where is the complex phase of . Thus, information about the phase is contained in the phase difference between the two components of the wave function. During the course of time, is expected to jump between constant values , when and have equal signs, and , if signs are opposite.
The relative phase can be extracted (up to a sign) from a measurement. Denoting we have
(5) 
where , and .
In Fig. 4(b), we plot , as measured in three different ways: (i) from the wave function in the Ising model, (ii) from the wave function in the Dicke model, (iii) from the correlation functions in the Dicke model according to Eq. (5). It is seen that all three measures reflect the jumps related to the Rabi oscillation, and are quantitatively close to the expected values ( and for ). It should be noted, though, that in contrast to the flux through the triangle, the relative phase in the doublewell is not a gaugeinvariant quantity, and is fully defined within the first shaking period. That is, shifting the initial time within the shaking period, would lead to a different phase.
In summary, we have simulated a system of three ions which encircle an artificial flux, engineered by periodic driving. Our simulation not only has shown the feasibility of Floquet engineering in trapped ions, but also revealed an unexpected robustness of the spin dynamics when the driving breaks timereversal symmetry in the effective model. This effect could stabilize and further enhance the quantum simulation of topological models also in larger systems. Having shown a path towards complexvalued spinspin interaction, we believe that trapped ions provide an ideal system for studying the role of topology and topological protection in small clusters. Potentially, such clusters will provide robust building blocks for larger quantum devices.
Acknowledgements. We are grateful to Alexey Gorshkov, Bruno JulíaDíaz, Jiehang Zhang, Paul Hess, and Zhexuan Gong for interesting discussions. Financial support from AFOSRMURI, EU grants EQuaM (FP7/20072013 Grant No. 323714), OSYRIS (ERC2013AdG Grant No. 339106), SIQS (FP7ICT20119 No. 600645), QUIC (H2020FETPROACT2014 No. 641122), Spanish MINECO (SEVERO OCHOA Grant SEV20150522, FISICATEAMO FIS201679508P), the Generalitat de Catalunya (SGR 874 and CERCA program), Fundacio Privada Cellex is acknowledged. GP is supported by the IC postdoctoral Research Fellowship program.
References
 Aharonov and Bohm (1959) Y. Aharonov and D. Bohm, Phys. Rev. 115, 485 (1959).
 Shapere and Wilczek (1989) A. Shapere and F. Wilczek, Geometric Phases in Physics, Advanced Series in Mathematical Physics Vol. 5 (World Scientific, 1989).
 Hofstadter (1976) D. R. Hofstadter, Phys. Rev. B 14, 2239 (1976).
 Janssen et al. (1994) M. Janssen, O. Viehweger, U. Fastenrath, and J. Hajdu, Introduction to the Theory of the Integer Qunantum Hall Effect (Wiley, John and Sons, 1994).
 Laughlin (1983) R. B. Laughlin, Phys. Rev. Lett. 50, 1395 (1983).
 Wen (2004) X. Wen, Quantum Field Theory of ManyBody Systems, Oxford Graduate Texts (OUP Oxford, 2004).
 Nayak et al. (2008) C. Nayak, S. H. Simon, A. Stern, M. Freedman, and S. Das Sarma, Rev. Mod. Phys. 80, 1083 (2008).
 Struck et al. (2012) J. Struck, C. Ölschläger, M. Weinberg, P. Hauke, J. Simonet, A. Eckardt, M. Lewenstein, K. Sengstock, and P. Windpassinger, Phys. Rev. Lett. 108, 225304 (2012).
 Miyake et al. (2013) H. Miyake, G. A. Siviloglou, C. J. Kennedy, W. C. Burton, and W. Ketterle, Phys. Rev. Lett. 111, 185302 (2013).
 Aidelsburger et al. (2013) M. Aidelsburger, M. Atala, M. Lohse, J. T. Barreiro, B. Paredes, and I. Bloch, Phys. Rev. Lett. 111, 185301 (2013).
 Jotzu et al. (2014) G. Jotzu, M. Messer, R. Desbuquois, M. Lebrat, T. Uehlinger, D. Greif, and T. Esslinger, Nature 515, 237 (2014).
 Mancini et al. (2015) M. Mancini, G. Pagano, G. Cappellini, L. Livi, M. Rider, J. Catani, C. Sias, P. Zoller, M. Inguscio, M. Dalmonte, and L. Fallani, Science 349, 1510 (2015).
 Stuhl et al. (2015) B. K. Stuhl, H.I. Lu, L. M. Aycock, D. Genkina, and I. B. Spielman, Science 349, 1514 (2015).
 Meinert et al. (2016) F. Meinert, M. J. Mark, K. Lauber, A. J. Daley, and H.C. Nägerl, Phys. Rev. Lett. 116, 205301 (2016).
 Hafezi et al. (2013) M. Hafezi, S. Mittal, J. Fan, A. Migdall, and J. Taylor, Nat. Photon. 7, 1001 (2013).
 Schine et al. (2016) N. Schine, A. Ryou, A. Gromov, A. Sommer, and J. Simon, Nature 534, 671 (2016).
 P. Roushan et al. (2017) P. Roushan et al., Nat. Phys. 13, 146 (2017).
 Lodahl et al. (2017) P. Lodahl, S. Mahmoodian, S. Stobbe, A. Rauschenbeutel, P. Schneeweiss, J. Volz, H. Pichler, and P. Zoller, Nature 541, 473 (2017).
 Kong et al. (2016) F. Kong, C. Ju, Y. Liu, C. Lei, M. Wang, X. Kong, P. Wang, P. Huang, Z. Li, F. Shi, L. Jiang, and J. Du, Phys. Rev. Lett. 117, 060503 (2016).
 Mintert and Wunderlich (2001) F. Mintert and C. Wunderlich, Phys. Rev. Lett. 87, 257904 (2001).
 Porras and Cirac (2004) D. Porras and J. I. Cirac, Phys. Rev. Lett. 92, 207901 (2004).
 Friedenauer et al. (2008) A. Friedenauer, H. Schmitz, J. T. Glueckert, D. Porras, and T. Schaetz, Nat. Phys. 4, 757 (2008).
 Kim et al. (2009) K. Kim, M.S. Chang, R. Islam, S. Korenblit, L.M. Duan, and C. Monroe, Phys. Rev. Lett. 103, 120502 (2009).
 Kim et al. (2010) K. Kim, M. S. Chang, S. Korenblit, R. Islam, E. E. Edwards, J. K. Freericks, G. D. Lin, L. M. Duan, and C. Monroe, Nature 465, 590 (2010).
 Richerme et al. (2014) P. Richerme, Z.X. Gong, A. Lee, C. Senko, J. Smith, M. FossFeig, S. Michalakis, A. V. Gorshkov, and C. Monroe, Nature 511, 198 (2014).
 Jurcevic et al. (2014) P. Jurcevic, B. P. Lanyon, P. Hauke, C. Hempel, P. Zoller, R. Blatt, and C. F. Roos, Nature 511, 202 (2014).
 Britton et al. (2012) J. W. Britton, B. C. Sawyer, A. C. Keith, C.C. J. Wang, J. K. Freericks, H. Uys, M. J. Biercuk, and J. J. Bollinger, Nature 484, 489 (2012).
 Bohnet et al. (2016) J. G. Bohnet, B. C. Sawyer, J. W. Britton, M. L. Wall, A. M. Rey, M. FossFeig, and J. J. Bollinger, Science 352, 1297 (2016).
 Zhang et al. (2017) J. Zhang, G. Pagano, P. W. Hess, A. Kyprianidis, P. Becker, H. Kaplan, A. V. Gorshkov, Z.X. Gong, and C. Monroe, ArXiv eprints (2017), arXiv:1708.01044 [quantph] .
 Nägerl et al. (1999) H. C. Nägerl, D. Leibfried, H. Rohde, G. Thalhammer, J. Eschner, F. SchmidtKaler, and R. Blatt, Phys. Rev. A 60, 145 (1999).
 Smith et al. (2016) J. Smith, A. Lee, P. Richerme, B. Neyenhuis, P. W. Hess, P. Hauke, M. Heyl, D. A. Huse, and C. Monroe, Nat Phys 12, 907 (2016).
 Roos et al. (2004) C. F. Roos, M. Riebe, H. Häffner, W. Hänsel, J. Benhelm, G. P. T. Lancaster, C. Becher, F. SchmidtKaler, and R. Blatt, Science 304, 1478 (2004).
 Jurcevic et al. (2016) P. Jurcevic, H. Shen, P. Hauke, C. Maier, T. Brydges, C. Hempel, B. P. Lanyon, M. Heyl, R. Blatt, and C. F. Roos, ArXiv eprints (2016), arXiv:1612.06902 [quantph] .
 Graß et al. (2015) T. Graß, C. Muschik, A. Celi, R. W. Chhajlany, and M. Lewenstein, Phys. Rev. A 91, 063612 (2015).
 Dunlap and Kenkre (1986) D. H. Dunlap and V. M. Kenkre, Phys. Rev. B 34, 3625 (1986).
 Grossmann et al. (1991) F. Grossmann, T. Dittrich, P. Jung, and P. Hänggi, Phys. Rev. Lett. 67, 516 (1991).
 Holthaus (1992) M. Holthaus, Phys. Rev. Lett. 69, 351 (1992).
 Holthaus and Hone (1993) M. Holthaus and D. Hone, Phys. Rev. B 47, 6499 (1993).
 Eckardt et al. (2005) A. Eckardt, C. Weiss, and M. Holthaus, Phys. Rev. Lett. 95, 260404 (2005).
 Eckardt et al. (2010) A. Eckardt, P. Hauke, P. SoltanPanahi, C. Becker, K. Sengstock, and M. Lewenstein, EPL (Europhysics Letters) 89, 10010 (2010).
 Lignier et al. (2007) H. Lignier, C. Sias, D. Ciampini, Y. Singh, A. Zenesini, O. Morsch, and E. Arimondo, Phys. Rev. Lett. 99, 220403 (2007).
 Struck et al. (2011) J. Struck, C. Ölschläger, R. Le Targat, P. SoltanPanahi, A. Eckardt, M. Lewenstein, P. Windpassinger, and K. Sengstock, Science 333, 996 (2011).
 Kitagawa et al. (2010) T. Kitagawa, E. Berg, M. Rudner, and E. Demler, Phys. Rev. B 82, 235114 (2010).
 Jiang et al. (2011) L. Jiang, T. Kitagawa, J. Alicea, A. R. Akhmerov, D. Pekker, G. Refael, J. I. Cirac, E. Demler, M. D. Lukin, and P. Zoller, Phys. Rev. Lett. 106, 220402 (2011).
 Lindner et al. (2011) N. H. Lindner, G. Refael, and V. Galitski, Nat. Phys. 7, 490 (2011), 10.1038/nphys1926.
 Bermudez et al. (2011) A. Bermudez, T. Schaetz, and D. Porras, Phys. Rev. Lett. 107, 150501 (2011).
 Graß et al. (2016) T. Graß, M. Lewenstein, and A. Bermudez, New Journal of Physics 18, 033011 (2016).
 Wang and Freericks (2012) C.C. J. Wang and J. K. Freericks, Phys. Rev. A 86, 032329 (2012).
 Wall et al. (2017) M. L. Wall, A. SafaviNaini, and A. M. Rey, Phys. Rev. A 95, 013602 (2017).
 (50) Here, for simplicity we have assumed sufficiently larger than the Ising couplings such that the net effect of driving is equivalent to the driving of a XX model by the timedependent magnetic field . More generally, the condition to achieve an effective XX model with the complex coupling as in Eq. (2\@@italiccorr) is , where is the period of .