Abstract
Topological insulators are fascinating states of matter exhibiting protected edge states and robust quantized features in their bulk. Here we propose and validate experimentally a method to detect topological properties in the bulk of onedimensional chiral systems. We first introduce the mean chiral displacement, an observable that rapidly approaches a value proportional to the Zak phase during the free evolution of the system. Then we measure the Zak phase in a photonic quantum walk of twisted photons, by observing the mean chiral displacement in its bulk. Next, we measure the Zak phase in an alternative, inequivalent timeframe and combine the two windings to characterize the full phase diagram of this Floquet system. Finally, we prove the robustness of the measure by introducing dynamical disorder in the system. This detection method is extremely general and readily applicable to all present onedimensional platforms simulating static or Floquet chiral systems.
Introduction
Topological phases of matter escape the canonical characterization of states dictated by the Ginzburg–Landau theory of phase transitions. These phases emerge without breaking symmetries and are not characterized by a longrange order nor a local order parameter but rather by a global topological order. Historically, topology was first proven to have a key role in explaining algebraically decaying order, transport and coherence of twodimensional Bose liquids, XY models and crystals^{1}. Shortly after, the quantization of Hall conductance^{2} was shown to be rooted in currentcarrying edge states, protected by the topology of the bulk^{3,4,5}. Being associated with a global order, these phases are robust against local perturbations and promise important applications in metrology, spintronics and quantum computation (see, for example, refs 6, 7, 8).
Intense studies^{9} followed the early discoveries, and topological insulators have by now been engineered in a variety of physical architectures, such as superconducting^{10}, mechanical^{11}, optomechanical^{12}, photonic^{13}, atomic^{14} and acoustic platforms^{15}. Such diverse systems have been exposed to either real or synthetic magnetic fields, and their topological properties have been studied by scattering at the interface between different domains^{15,16} or imaging edge states^{17,18,19,20,21,22,23,24,25,26}. Direct detection of topological invariants in the bulk of the system (with no need of edges) has been reported so far by very few experiments^{27,28,29}.
Topological insulators are classified in terms of dimensionality and discrete symmetries^{30}. Onedimensional (1D) systems with chiral symmetry are characterized by the Zak phase, that is, the Berry phase accumulated by an eigenstate during its parallel transport through the whole Brillouin zone^{31}. The Zak phase is closely related to the electric polarization in solids and plays a key role in the modern theory of insulators^{32,33}.
Periodically driven (Floquet) systems are attracting an increasing interest, as these show richer topological features than their static counterparts^{17,24,25,26,34,35,36,37,38,39,40,41,42,43,44}. Particularly promising Floquet topological systems are discretetime quantum walks (QWs)^{16,17,29,45,46,47}, and recent works have reported the observation of topological invariants^{16,29}, quantum phase transitions^{46} and edge states^{17} in these systems. In its simplest version, a QW is the discrete time evolution of a particle (the walker) on a 1D lattice^{48}. At each step, the walker moves to neighbouring sites, with the direction of the shift depending on the state of an internal twolevel degree of freedom (the coin). Between consecutive steps, a rotation modifies the coin state, univoquely determining the following evolution.
Here we demonstrate that, in chiral 1D static and Floquet systems with spin 1/2 (that is, a twostate coin), the mean chiral displacement of a particle’s wavepacket becomes quantized and proportional to the Zak phase in the longtime limit. Remarkably, this occurs during the free evolution of the system, in absence of any external force or loss mechanism, with the only requirement that the initial wavefunction is localized. We validate experimentally this finding in a photonic discretetime QW based on the orbital angular momentum (OAM) of a light beam. We implement the same QW in a shifted inequivalent timeframe and measure a second Zak phase. Combining the two windings, we extract the complete set of topological invariants characterizing the system. Finally, we prove the robustness of our detection by adding dynamical disorder. These measurements provide therefore a bulk measurement of the Zak phases and complete topological invariants of a 1D chiral QW. Our proposal may be straightforwardly applied to general driven Floquet systems.
Results
Zak phase detection in the bulk of a QW
In one dimension, discretetime QWs with chiral symmetry display a quantized Zak phase and have been extensively studied in the past years. Among these implementations, we focus on the photonic platform proposed in ref. 46. Here the walk takes place on a lattice whose sites are associated with photonics states , corresponding to light beams that carry mħ units of OAM per photon along the propagation axis and show a twisted wavefront^{49}. The two coin states are instead mapped onto the left and right circular polarizations of the beam, carrying ±ħ units of spin angular momentum per photon along the propagation axis. Once the system is prepared in an initial state , its state after t timesteps is given by
where the singlestep operator is obtained by cascading suitable combinations of quarterwave plates and qplates^{46,50,51}. In Fig. 1a, we show a pictorial representation of our setup that realizes a sevenstep QW with implemented specifically as U ≡ Q·W^{46}. The action of a quarterwave plate oriented at 90° with respect to the horizontal direction is described by the local operator W, rotating the polarization states as
Here creates a particle on site m with polarization L/R and σ_{i} are Pauli matrices acting in the coin (polarization) space. The translation operator Q is implemented by a qplate, a liquid crystal device that yields an effective spin–orbit interaction in the light beam. This couples neighbouring sites and polarization states as
where are the operators that flip the coin states and , δ is the optical retardation of the qplate and h.c. stands for Hermitian conjugate. Further details on the qplates and on the complete experimental setup are provided in the Methods section and Supplementary Fig. 1.
Very generally, QWs are generated by the repeated application of a unitary operator , and therefore the system can be described in the framework of Floquet theory. As a consequence of translational invariance in space, the effective Hamiltonian associated with a full period is diagonal in momentum space and may be written as
with E(k) the quasienergy dispersion, and we have set the period T and ħ to unity. The point on the Bloch sphere identified by the unit vector n(k) represents the coin part of the system eigenstates, while their spatial part is a plane wave with quasimomentum k^{46}. The function ln(x) denotes the principal branch of the natural (matrix) logarithm, so that the quasienergy is a periodic function, with −π and +π identified.
The class of QWs we are considering features chiral symmetry, since there exists a unitary operator Γ such that Γ^{2}=I and . These conditions imply that Γ is Hermitian and that , with v_{Γ} a vector labelling a point on the Bloch sphere. In this case, the unit vector n is bound to rotate around the origin in a plane orthogonal to v_{Γ}, and the Zak phase equals
The winding number γ/π assumes strictly integer values and counts the number of times the unit vector n rotates around the unit vector v_{Γ} as k traverses the whole Brillouin zone. In Fig. 1b, we show the winding of the vector n of the operator U, for two values of δ in different topological sectors. The Zak phase is therefore a bulk property; although it has strong influences in properties of systems where it arises, its detection in current experimental architectures remains challenging.
In the following, we show that information on such topological invariant is hidden in the subleading terms of the mean displacement , when the initial wavepacket is localized on a single site. This extends a previous result showing that, in the same conditions, the ballistic terms of higher moments of the walker’s displacement feature discontinuities at topological phase transitions^{46}. Let us consider the evolution of a wavepacket initially localized at site m=0, and whose polarization is characterized by the expectation values of the three Pauli matrices, . The mean displacement of the wavepacket after t timesteps is given by (see Supplementary Note 1 for details)
The term in square brackets in equation (6) is proportional to , the projection of the initial polarization on a direction orthogonal to v_{Γ}, and contains a ballistic term L(t) (which grows linearly with t) and a subleading part S(t).
The vector identifying the specific direction of in the plane orthogonal to v_{Γ}, and the explicit functional forms of L(t) and S(t), are nonuniversal features that depend on the specific protocol (or timeframe) and have no particular relevance for our discussion. The second term in equation (6), which is weighted by (the projection of the initial polarization along v_{Γ}), is the subleading chiral term S_{Γ} that may be written as (see Supplementary Note 1 for details)
In the limit t→∞, S_{Γ} becomes proportional to the Zak phase, as the oscillatory correction quickly averages to zero (see Fig. 1c).
The above analysis shows that information on the Zak phase is contained in the mean displacement of the walker, and it may be extracted by fitting at long times, isolating in turn the second term of equation (6). A related result for the case of a nonHermitian QW initialized on a chiral eigenstate (that is, an initial condition such that ) was demonstrated theoretically in ref. 52 and verified experimentally in ref. 29. However, this measurement would not be robust. Indeed, even if one prepared the initial polarization in an eigenstate of the chiral operator Γ, so that =0, disorder during the propagation of the beam would introduce polarization components orthogonal to v_{Γ}. These would give rise to ballistic contributions, which in the longtime limit would dramatically affect the result.
An alternative and more convenient approach consists in measuring the mean chiral displacement
which quantifies the relative shift between the two projections of the state onto the eigenstates of the chiral operator (see Supplementary Note 1 for a concise derivation of this equality). Importantly, the result contained in equation (8) is (i) independent of the initial polarization and (ii) robust against disorder. We probe the chiral displacement in our photonic platform by performing a sevenstep QW of the protocol U=Q·W, as depicted in Fig. 1a. The chiral eigenstates correspond to two specific orthogonal polarization states, which depend explicitly on the protocol, and which we detect at the end of the QW (see Methods section). In Fig. 1c, we report the measured values of for two different initial polarization states. Experimental points closely follow the theory curve for seven time steps (blue solid line), and no significant differences can be observed between the two different initial states, proving that this measurement is insensitive to the choice of the polarization of the photons. For completeness, we also show results predicted for 33 steps, and the asymptotic longtime limit, which coincides with the Zak phase (over 2π). We note here that, although both theory and data oscillate, as few as seven steps are enough to have a clear detection of the Zak phase.
Zak phase in a shifted timeframe
In static systems, bulk topological invariants such as the Zak phase or the Chern number are uniquely defined by integrals over the whole Brillouin zone and are in onetoone correspondence with the presence of edge states, thus providing a full classification in terms of the periodic table of topological insulators^{30}. The situation is very different in periodically driven (Floquet) systems in D dimensions, where the integral determining the topological invariants needs to be performed over a D+1 dimensional torus constituted by the Brillouin zone and an extra periodic dimension, the quasienergy^{42}.
Moreover, a gauge freedom is introduced by the choice of the timeframe, that is, the origin of time of the periodic cycle (see Fig. 2a). While the dispersion E(k) is equal in all timeframes, the effective Hamiltonian, its eigenstates and symmetries and the resulting dynamics crucially depend on the timeframe^{40}. As an example, the operator defines a timeframe that is inequivalent to the one introduced by U. In particular, the unit vector (k) defined by may wind twice around the chiral axis as k traverses the Brillouin zone (see Fig. 2c), and its Zak phase (dashed line in Fig. 2d) differs from the Zak phase γ of protocol U (dashed line in Fig. 1c).
We realize experimentally protocol by the setup shown schematically in Fig. 2b. Using the relation , it is straightforward to see that . Hence, we realize the operator by placing qplates yielding an optical retardation δ/2 at the beginning and end of the optical path, while in the bulk of the walk we adopt the same sequence reported in Fig. 1a (with the last qplate removed). Overall, our QW implements seven steps of protocol by means of a total of eight qplates, six with retardation δ and two tuned at δ/2 (first and last plates), separated by quarterwave plates. In Fig. 2d, we report the measure of the mean chiral displacement generated by the singlestep operator . As in the case of protocol U, this quantity accurately follows the theory prediction, providing an unambiguous detection of the Zak phase of the infinite system after just seven steps.
Complete topological characterization
It is clear from the previous discussion that the Zak phase associated with a single timeframe does not contain all the topological information of our QW. Indeed in Floquet 1D chiral systems, there exist two independent classes of protected edge states at either 0 and π energies. An example of these edge states is shown in Fig. 3a, where we plot the quasienergies of all eigenstates of an openended lattice. As remarked above, the spectrum is independent of the timeframe. The spectrum contains edge states even for 3π/2<δ<5π/2 where the Zak phase γ of protocol U is zero, explicitly confirming that the Zak phase of a single QW protocol does not contain the complete information about the topological state of the system.
The bulkedge correspondence in these driven systems requires two invariants C_{0} and C_{π}, yielding, respectively, the number of 0 and πenergy edge states. As shown in refs 38, 41, these are simple functions of two Zak phases, measured in two inequivalent timeframes possessing an ‘inversion point’, that is, which may be written, respectively, as and , with F a suitable evolution operator. In the case of our setup, the two special protocols fulfilling this criterion are and . However, it is simple to show that is topologically equivalent to U, as no gap closing happens during the rotation ; therefore the Zak phase of coincides with γ. As such, the complete topological classification of 1D chiral systems may be obtained by means of the two quantities
which converge in the longtime limit, respectively, to the number of 0 and πenergy edge states. By combining our measurements of the mean chiral displacements measured in the inequivalent timeframes, we are now able to compute the invariants C_{0} and C_{π} and detect the complete phase diagram of this system: the result is shown in Fig. 3b. Once again, our measurements show a remarkably fast convergence towards the asymptotic limit.
Robustness to dynamical disorder
Finally, we test the stability of the quantization of the mean chiral displacement against disorder. In particular, we choose protocol U and introduce dynamical disorder by offsetting the optical retardation δ_{j} (1≤j≤7) of each qplate by a small random amount around their common mean value . In our experiment, we set Δ=π/10 and π/5. We note that this disorder is dynamic, in the sense that it affects independently the various qplates crossed by the beam, but crucially it respects chiral symmetry. This can be simply understood by noting that the vector v_{Γ}, defining the chiral operator, does not depend on δ.
As shown in Fig. 4, in single realizations the mean chiral displacement presents oscillations featuring higher amplitude for increasing disorder, but an ensemble average over independent realizations smoothly converges to the expected theoretical result, which in the infinite time limit gives the bulk value of the Zak phase. Here we performed measurements on protocol U, but similar robustness of the chiral displacement shall hold for every 1D QW chiral protocol, and more generally every 1D chiral system, as long of course as the disorder does not break chiral symmetry and its strength is small compared to the gap size to prevent interband transitions. As an example, in the Supplementary Note 2 and Supplementary Figs 2–5 we show that the mean chiral displacement is an equally robust topological marker for a completely different and static (that is, not driven) system, the celebrated SSH model.
Discussion
Summarizing, here we proposed an efficient method to measure the Zak phase of a chiral system by direct observation of its free bulk dynamics. In particular, we showed that information on the topological phase of the bulk is encoded in the mean chiral displacement, an oscillatory quantity that rapidly converges to the Zak phase, and is robust against (chiralpreserving) disorder in both space and time.
We experimentally verified our findings by performing the first measurement of the Zak phase of a chiral QW. The physical platform we chose is a photonic setup based on the OAM of a light beam, where the mean chiral displacement corresponds to the relative shift of the two chiral polarization components. A precise readout of the Zak phase was obtained after only seven QW steps. We further used the same method to measure the Zak phase in a complementary timeframe, which we realized by swapping few optical components. By combining the two measurements, we extracted the two invariants providing the complete bulkedge correspondence for this driven system, that is, the one associated to the 0energy edge state, and the one connected to the anomalous πenergy edge state. Finally, we proved that the mean chiral displacement is a robust measure of the Zak phase by introducing dynamical but chiralpreserving disorder.
Although here we investigated experimentally a specific QW, our results are not restricted to QWs nor to Floquet systems. Indeed, the mean chiral displacement provides a robust topological characterization of arbitrary spin1/2 1D chiral systems, either static or periodically driven. These may nowadays be realized in a variety of platforms, ranging from ultracold atoms in optical lattices to photonic waveguides and from semiconductor quantum wells to optomechanical systems.
While formerly known methods for detection of topological properties require a uniform filling of the band of interest, external forces, loss mechanisms or finetuning so that only edge states are populated, the method proposed here quite remarkably achieves this goal by observing the free evolution of a single particle, initially localized on a single site in the bulk. This aspect may be specially beneficial for systems where filling a band is intrinsically challenging, such as bosonic condensates or phononic and photonic ensembles.
Future interesting directions opened by this work include the extension of our results to chiral systems with more than two internal states, a further understanding of the role played by temporal disorder and the topological characterization of systems in higher spatial dimensions.
Methods
Experimental setup
Our apparatus is shown schematically in Figs 1a and 2b, and a more detailed description is given in Supplementary Fig. 1. We produce a TEM_{00} mode by coupling the output of a Ti:Sa laser (λ=800 nm) to a singlemode fibre, thus preparing the beam in an OAM state with m=0. At the exit of the fibre, a specific polarization is selected by means of a sequence of a quarterwave plate and a halfwave plate. Therefore, the initial state of the QW is =, where m is the position in the walker (OAM) and space s its coin state (polarization). In the standard protocol U=Q·W, the single step consists of a quarterwave plate oriented at 90° with respect to the horizontal direction (operator W), followed by a qplate (operator Q), as shown in Fig. 1a. To implement the second protocol, we exploited the fact that equation (3) may be written as , that is, it corresponds to a rotation around a suitable unit vector n_{Q}; as such, a qplate with retardation δ/2 implements the desired operator . To implement the singlestep operator , we then added a δ/2 qplate at the beginning of the sequence, and we halved the retardation of the last qplate, as shown in Fig. 2b.
qplates
Each qplate consists of a thin layer of birefringent liquid crystals, whose optic axes are arranged in a singular pattern characterized by a topological charge q (in our case, q=1/2). The patterned birefringence gives rise to an optical spin–orbit coupling that induces the polarizationdependent shift of OAM. Along with the specific pattern, the action of each device is determined by its optical retardation δ, as reported in equation (3). The optical retardation can be continuously tuned by applying an electric field, allowing in turn for an accurate control of the spin–orbit interaction^{53}.
Detection of the chiral displacement
At the end of the walk, we can select any polarization component of the final state by a combination of a quarterwave and a halfwave plate, followed by a linear polarizer, and we measure its OAM content by diffraction on a spatial light modulator coupled to a singlemode fibre and a power meter, which records the light intensity. Since we are interested in analysing the OAM spectrum of chiral components of the final wavepacket, waveplates orientations are selected so as to implement polarization projections onto chiral states and . The chiral operators for protocols U and are, respectively, and σ_{z}, so it is straightforward to see that and for protocol U, while and for protocol . The combination of polarization and OAM projections allows for determining the probabilities P_{i,m}, with i={↑, ↓}, that the system is in the chiral state and in the OAM state . Given the probability distributions P_{i,m}, the chiral displacement is simply given by .
Data availability
The complete set of raw data supporting the findings of this study is available from the corresponding authors upon request.
Additional information
How to cite this article: Cardano, F. et al. Detection of Zak phases and topological invariants in a chiral quantum walk of twisted photons. Nat. Commun. 8, 15516 doi: 10.1038/ncomms15516 (2017).
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
 1.
Kosterlitz, J. M. Kosterlitz–Thouless physics: a review of key issues. Rep. Prog. Phys. 79, 026001 (2016).
 2.
Klitzing, K. V., Dorda, G. & Pepper, M. New method for highaccuracy determination of the finestructure constant based on quantized Hall resistance. Phys. Rev. Lett. 45, 494–497 (1980).
 3.
Laughlin, R. B. Quantized Hall conductivity in two dimensions. Phys. Rev. B 23, 5632–5633 (1981).
 4.
Thouless, D. J., Kohmoto, M., Nightingale, M. P. & den Nijs, M. Quantized Hall conductance in a twodimensional periodic potential. Phys. Rev. Lett. 49, 405–408 (1982).
 5.
Halperin, B. I. Quantized Hall conductance, currentcarrying edge states, and the existence of extended states in a twodimensional disordered potential. Phys. Rev. B 25, 2185–2190 (1982).
 6.
von Klitzing, K. The quantized Hall effect. Rev. Mod. Phys. 58, 519–531 (1986).
 7.
Fert, A. Nobel Lecture: origin, development, and future of spintronics. Rev. Mod. Phys. 80, 1517–1530 (2008).
 8.
Pachos, J. K. Introduction to Topological Quantum Computation Cambridge University Press (2012).
 9.
Qi, X.L. & Zhang, S.C. Topological insulators and superconductors. Rev. Mod. Phys. 83, 1057–1110 (2011).
 10.
Beenakker, C. & Kouwenhoven, L. A road to reality with topological superconductors. Nat. Phys. 12, 618–621 (2016).
 11.
Huber, S. D. Topological mechanics. Nat. Phys. 12, 621–623 (2016).
 12.
Peano, V., Brendel, C., Schmidt, M. & Marquardt, F. Topological phases of sound and light. Phys. Rev. X 5, 031011 (2015).
 13.
Lu, L., Joannopoulos, J. D. & Soljačić, M. Topological states in photonic systems. Nat. Phys. 12, 626–629 (2016).
 14.
Goldman, N., Budich, J. C. & Zoller, P. Topological quantum matter with ultracold gases in optical lattices. Nat. Phys. 12, 639–645 (2016).
 15.
Xiao, M. et al. Geometric phase and band inversion in periodic acoustic systems. Nat. Phys. 11, 240–244 (2015).
 16.
Barkhofen, S. et al. Measuring topological invariants and protected bound states in disordered discrete time quantum walks. Preprint at http://arxiv.org/abs/1606.00299 (2016).
 17.
Kitagawa, T. et al. Observation of topologically protected bound states in photonic quantum walks. Nat. Commun. 3, 882 (2012).
 18.
Rechtsman, M. C. et al. Photonic Floquet topological insulators. Nature 496, 196–200 (2013).
 19.
Hafezi, M., Mittal, S., Fan, J., Migdall, A. & Taylor, J. M. Imaging topological edge states in silicon photonics. Nat. Photon. 7, 1001–1005 (2013).
 20.
Stuhl, B. K., Lu, H.I., Aycock, L. M., Genkina, D. & Spielman, I. B. Visualizing edge states with an atomic Bose gas in the quantum Hall regime. Science 349, 1514–1518 (2015).
 21.
Mancini, M. et al. Observation of chiral edge states with neutral fermions in synthetic Hall ribbons. Science 349, 1510–1513 (2015).
 22.
Leder, M. et al. Realspace imaging of a topologically protected edge state with ultracold atoms in an amplitudechirped optical lattice. Nat. Commun. 7, 13112 (2016).
 23.
Meier, E. J., An, F. A. & Gadway, B. Observation of the topological soliton state in the SuSchriefferHeeger model. Nat. Commun 7, 13986 (2016).
 24.
Mukherjee, S. et al. Experimental observation of anomalous topological edge modes in a slowly driven photonic lattice. Nat. Commun. 8, 13918 (2017).
 25.
Peng, Y.G. et al. Experimental demonstration of anomalous Floquet topological insulator for sound. Nat. Commun. 7, 13368 (2016).
 26.
Maczewsky, L. J., Zeuner, J. M., Nolte, S. & Szameit, A. Observation of photonic anomalous Floquet topological insulators. Nat. Commun. 8, 13756 (2017).
 27.
Atala, M. et al. Direct measurement of the Zak phase in topological Bloch bands. Nat. Phys. 9, 795–800 (2013).
 28.
Aidelsburger, M. et al. Revealing the topology of Hofstadter bands with ultracold bosonic atoms. Nat. Phys. 11, 162 (2015).
 29.
Zeuner, J. M. et al. Observation of a topological transition in the bulk of a nonHermitian system. Phys. Rev. Lett. 115, 040402 (2015).
 30.
Chiu, C.K., Teo, J. C. Y., Schnyder, A. P. & Ryu, S. Classification of topological quantum matter with symmetries. Rev. Mod. Phys. 88, 035005 (2016).
 31.
Zak, J. Berry’s phase for energy bands in solids. Phys. Rev. Lett. 62, 2747–2750 (1989).
 32.
Resta, R. Macroscopic polarization in crystalline dielectrics: the geometric phase approach. Rev. Mod. Phys. 66, 899–915 (1994).
 33.
Xiao, D., Chang, M.C. & Niu, Q. Berry phase effects on electronic properties. Rev. Mod. Phys. 82, 1959–2007 (2010).
 34.
Oka, T. & Aoki, H. Photovoltaic Hall effect in graphene. Phys. Rev. B 79, 081406 (2009).
 35.
Lindner, N. H., Refael, G. & Galitski, V. Floquet topological insulator in semiconductor quantum wells. Nat. Phys. 7, 490–495 (2011).
 36.
Asbóth, J. K. Symmetries, topological phases, and bound states in the onedimensional quantum walk. Phys. Rev. B 86, 195414 (2012).
 37.
Cayssol, J., Dóra, B., Simon, F. & Moessner, R. Floquet topological insulators. Phys. Status Solidi 7, 101–108 (2013).
 38.
Asbóth, J. K. & Obuse, H. Bulkboundary correspondence for chiral symmetric quantum walks. Phys. Rev. B 88, 121406 (2013).
 39.
Rudner, M. S., Lindner, N. H., Berg, E. & Levin, M. Anomalous edge states and the bulkedge correspondence for periodically driven twodimensional systems. Phys. Rev. X 3, 031005 (2013).
 40.
Goldman, N. & Dalibard, J. Periodically driven quantum systems: effective Hamiltonians and engineered gauge fields. Phys. Rev. X 4, 031027 (2014).
 41.
Asbóth, J. K., Tarasinski, B. & Delplace, P. Chiral symmetry and bulkboundary correspondence in periodically driven onedimensional systems. Phys. Rev. B 90, 125143 (2014).
 42.
Nathan, F. & Rudner, M. S. Topological singularities and the general classification of FloquetBloch systems. New J. Phys. 17, 125014 (2015).
 43.
Dal Lago, V., Atala, M. & Foa Torres, L. E. F. Floquet topological transitions in a driven onedimensional topological insulator. Phys. Rev. A 92, 023624 (2015).
 44.
Fleury, R., Khanikaev, A. B. & Alù, A. Floquet topological insulators for sound. Nat. Commun. 7, 11744 (2016).
 45.
Kitagawa, T., Rudner, M. S., Berg, E. & Demler, E. Exploring topological phases with quantum walks. Phys. Rev. A 82, 033429 (2010).
 46.
Cardano, F. et al. Statistical moments of quantumwalk dynamics reveal topological quantum transitions. Nat. Commun. 7, 11439 (2016).
 47.
Groh, T. et al. Robustness of topologically protected edge states in quantum walk experiments with neutral atoms. Phys. Rev. A 94, 013620 (2016).
 48.
VenegasAndraca, S. E. Quantum walks: a comprehensive review. Quantum Inf. Process. 11, 1015–1106 (2012).
 49.
Yao, A. M. & Padgett, M. J. Orbital angular momentum: origins, behavior and applications. Adv. Opt. Photon. 3, 161–204 (2011).
 50.
Marrucci, L., Manzo, C. & Paparo, D. Optical spintoorbital angular momentum conversion in inhomogeneous anisotropic media. Phys. Rev. Lett. 96, 163905 (2006).
 51.
Cardano, F. et al. Quantum walks and wavepacket dynamics on a lattice with twisted photons. Sci. Adv. 1,, e1500087 (2015).
 52.
Rudner, M. S. & Levitov, L. S. Topological transition in a nonHermitian quantum walk. Phys. Rev. Lett. 102, 065703 (2009).
 53.
Piccirillo, B., D’Ambrosio, V., Slussarenko, S., Marrucci, L. & Santamato, E. Photon spintoorbital angular momentum conversion via an electrically tunable q plate. Appl. Phys. Lett. 97, 241104 (2010).
Acknowledgements
We acknowledge insightful discussions with János Asbóth, Alessio Celi and MiguelAngel Martin Delgado. We thank Damiano Fiorillo for his help in taking and analysing the experimental data. We acknowledge support from Adv. ERC grants PHOSPhOR and OSYRIS, EU grant QUIC (H2020FETPROACT2014 No. 641122), MINECO (Severo Ochoa grant SEV20150522 and FOQUS FIS201346768), Generalitat de Catalunya (SGR 874 and CERCA) and the Fundació Privada Cellex. A.Da. acknowledges funding from the CellexICFOMPQ fellowship. P.M. acknowledges funding from a ‘Ramón y Cajal’ fellowship.
Author information
Affiliations
Dipartimento di Fisica, Università di Napoli Federico II, Complesso Universitario di Monte Sant’Angelo, Via Cintia, Napoli 80126, Italy
 Filippo Cardano
 , Alessio D’Errico
 , Maria Maffei
 , Bruno Piccirillo
 , Corrado de Lisio
 , Giulio De Filippis
 , Vittorio Cataudella
 , Enrico Santamato
 & Lorenzo Marrucci
ICFOInstitut de Ciencies Fotoniques, The Barcelona Institute of Science and Technology, Av. Carl Friedrich Gauss 3, Castelldefels 08860, Spain
 Alexandre Dauphin
 , Maria Maffei
 , Maciej Lewenstein
 & Pietro Massignan
CNRSPIN, Complesso Universitario di Monte Sant’Angelo, Via Cintia, Napoli 80126, Italy
 Corrado de Lisio
 , Giulio De Filippis
 , Vittorio Cataudella
 & Enrico Santamato
CNRISASI, Institute of Applied Science and Intelligent Systems, Via Campi Flegrei 34, Pozzuoli (NA) 80078, Italy
 Lorenzo Marrucci
ICREA—Institució Catalana de Recerca i Estudis Avançats, Pg. Lluis Companys 23, Barcelona E08010, Spain
 Maciej Lewenstein
Authors
Search for Filippo Cardano in:
Search for Alessio D’Errico in:
Search for Alexandre Dauphin in:
Search for Maria Maffei in:
Search for Bruno Piccirillo in:
Search for Corrado de Lisio in:
Search for Giulio De Filippis in:
Search for Vittorio Cataudella in:
Search for Enrico Santamato in:
Search for Lorenzo Marrucci in:
Search for Maciej Lewenstein in:
Search for Pietro Massignan in:
Contributions
A.Da., M.L. and P.M., helped by F.C., A.D’E. and M.M., developed the complete theoretical framework, building on preliminary studies of the dynamical subleading terms carried out by E.S., G.D.F. and V.C. F.C., A.D’E., M.M. and L.M. designed the experimental methodology. F.C. and A.D’E., with contributions from M.M. and C.d.L., carried out the experiment. B.P. prepared the qplates. F.C., A.D’E., A.Da., M.M., M.L. and P.M. wrote the manuscript. L.M. and M.L. supervised the project. All authors discussed the results and contributed to refining the manuscript.
Competing interests
The authors declare no competing financial interests.
Corresponding authors
Correspondence to Filippo Cardano or Alexandre Dauphin.
Supplementary information
PDF files
 1.
Supplementary Information
Supplementary Figures and Supplementary Notes
 2.
Peer Review File
Rights and permissions
This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/
About this article
Further reading

Twisted photons: new quantum perspectives in high dimensions
Light: Science & Applications (2018)

Staterecycling and timeresolved imaging in topological photonic lattices
Nature Communications (2018)
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.