Abstract
Nonconservative physical systems admit a special kind of spectral degeneracy, known as exceptional point (EP), at which eigenvalues and eigenvectors of the corresponding nonHermitian Hamiltonian coalesce. Dynamical parametric encircling of the EP can lead to nonadiabatic evolution associated with a state flip, a sharp transition between the resonant modes. Physical consequences of the dynamical encircling of EPs in open dissipative systems have been explored in optics and photonics. Building on the recent progress in understanding the paritytime (\({\mathscr{P}}{\mathscr{T}}\))symmetric dynamics in spin systems, we use topological properties of EPs to implement chiral nonreciprocal transmission of a spin through the material with nonuniform magnetization, like helical magnet. We consider an exemplary system, spintorquedriven single spin described by the timedependent nonHermitian Hamiltonian. We show that encircling individual EPs in a parameter space results in nonreciprocal spin dynamics and find the range of optimal protocol parameters for highefficiency asymmetric spin filter based on this effect. Our findings offer a platform for nonreciprocal spin devices for spintronics and magnonics.
Introduction
In magnetic structures, nonreciprocal spin transport is usually mediated by either surface modes^{1} or asymmetric exchange interaction^{2,3}, giving rise to asymmetric spinwave dispersion with respect to the propagation direction^{4,5,6}. Here we present a general method for setting nonreciprocal spin transport based on the nonHermitian dynamics around exceptional points (EPs), spectral degeneracies arising in nonconservative systems endowed with simultaneous paritytime (\({\mathscr{P}}{\mathscr{T}}\)) invariance^{7,8}. The emerging effect of nonreciprocal transport is fundamentally universal and stems from the adiabaticity breaking unique nonHermitian systems. The resultant nonreciprocity is highly tunable and robust with respect to noise.
NonHermitian extension of Hamiltonian formalism provides a quantitative description of nonconservative dynamics in outofequilibrium systems^{9}. NonHermitian Hamiltonians admit a special class of degeneracy points, called exceptional points, where some of their eigenvalues and the corresponding eigenvectors coalesce. This property is unique to nonHermitian operators and must be distinguished from Hermitian spectrum degeneracies, like diabolical points, where only eigenvalues coalesce. NonHermitian degeneracies are stable against both Hermitian or nonHermitian perturbations, and have far reaching physical implications due to their topological structure that have been explored both theoretically^{10,11,12,13} and experimentally^{14,15,16}.
Appearance of EPs in the eigenspectrum of nonHermitian Hamiltonians is a direct consequence of the \({\mathscr{P}}{\mathscr{T}}\) symmetrybreaking^{7,8}. Hamiltonians invariant under the simultaneous operations of parity (\({\mathscr{P}}\)) and timereversal (\({\mathscr{T}}\)) form a special class of nonHermitian Hamiltonians that admit a fully real eigenspectrum, provided that characteristic amplitudes of their antiHermitian parts are below certain threshold values. The \({\mathscr{P}}{\mathscr{T}}\) symmetrybreaking marks the transition from the real to complex eigenspectrum of \({\mathscr{P}}{\mathscr{T}}\)symmetric Hamiltonians. The physical interpretation of \({\mathscr{P}}{\mathscr{T}}\) symmetry as “balanced gain and loss”^{17} triggered an increasing number of experimental realizations of \({\mathscr{P}}{\mathscr{T}}\)symmetric systems and established \({\mathscr{P}}{\mathscr{T}}\) symmetrybreaking as a nonequilibrium phase transitions between stationary and nonstationary dynamics in driven dissipative systems.
We demonstrate that dynamical encircling a secondorder EP in the parameter space of a classical magnetic system can lead to nontrivial spin transport properties never seen before. We consider a linear nonHermitian Hamiltonian describing dynamics of a single classical spin in the external timedependent magnetic field. This is a generic model for ballistic spin transport experiments, in which spin current is generated by passing nonpolarized electrons through a material with nonuniform magnetization, e.g. helical magnet. We show that the combination of applied spintransfer torque (STT) and timedependent magnetic field, coordinated in the form of a protocol that takes the system around an EP of the corresponding nonHermitian Hamiltonian, enables nonreciprocal spin dynamics and transport with a high degree of spinpolarization efficiency.
To derive the propagation of individual spins through a medium with spatially nonuniform magnetic field, we employ the model of a stationary spin in timedependent magnetic field. We introduce an external imaginary magnetic field to account for the effects of nonconservative forces on spin dynamics, e.g. spin torque. Indeed, it was shown that the action of Slonczewski STT^{18} is equivalent to that of applied imaginary magnetic field in the nonHermitian Hamiltonian formalism^{19}. We propose a design of a classical singlespin system with the inherent \({\mathscr{P}}{\mathscr{T}}\) symmetry, such that the energy spectrum of the corresponding nonHermitian Hamiltonian admits a pair of secondorder EPs. We then consider a cyclic variation of the applied magnetic field, such that a single EP is encircled in the twodimensional parameter space \(({h}_{x},{h}_{y})\) of the model, where \({h}_{x}\) and \({h}_{y}\) are projections of applied magnetic field. Encircling protocols can be constructed in such a way that only a clockwise or anticlockwise direction, depending on the chirality of the EP, yields the final spin state that is different from the initial one at the beginning of the protocol, despite identical initial and final system parameters. The effect relies on the breakdown of the adiabatic theorem in nonHermitian systems, and results in nonreciprocal time evolution of spin.
We perform numerical simulations to verify the robustness of the nonreciprocal effect and to identify optimal protocol parameters for achieving maximum asymmetric spinpolarization efficiency in the proposed scheme. In particular, for encircling protocols centered around an isolated EP, we calculate deviations from perfect nonreciprocity as a function of protocol frequency, radius, and initial conditions. Our results show that the highest degree of spinfilter asymmetry between the clockwise and anticlockwise circular protocols is achieved for slowfrequency parametric trajectories starting in a wide region near the line of \({\mathscr{P}}{\mathscr{T}}\)symmetry and with a large enough radii to encircle both EPs, as opposed to encircling an individual EP.
Experimental realization of the proposed spin evolution protocols enables a novel type of highefficiency asymmetric spinfilter devices possessing a number of unique properties. First, high efficiency of the spinfilter effect is guaranteed by the topology of trajectories encircling the EP. Second, the proposed spinfilter device, Fig. 1, is asymmetric, i.e. breaks timereversal symmetry. As a result, the polarization directions of spins entering the device from the left and from the right are different. Third, both spinpolarization directions are controlled by the amplitudes of the bias STT and external magnetic field, rendering a highly tunable and efficient system.
Results
The most general form of a linear spin Hamiltonian contains three independent parameters^{20}, which, without any loss of generality, can be chosen as two components of the real magnetic field along the \(x\) and \(y\) axes, and a single component of the imaginary field along the \(y\) axis:
where β is the dimensionless amplitude of applied spin torque^{19}.
To describe singlespin dynamics governed by the nonHermitian Hamiltonian (1), we employ SU(2) spincoherent states^{21,22,23}:
where \({S}_{+}\equiv {S}_{x}+i{S}_{y}\), and \(z\in {\mathbb{C}}\) is the standard stereographic projection of the spin direction on a unit sphere, \(z=({s}_{x}+i{s}_{y})/(1{s}_{z})\), with \({s}_{i}\equiv {S}_{i}/S\). For the purpose of this work, we focus on the classical limit of large spin, \(S\to \infty \), in which case one can write the generalized spin equation of motion as follows^{19}:
where \({ {\mathcal H} }_{0}(z,\bar{z})\) is the expectation value in spincoherent states:
For the spin Hamiltonian (1), this yields the following evolution equation:
The eigenvalue problem for the spin Hamiltonian (1) in the classical limit reduces to evaluating the expectation values, Eq. (3), at the fixed points of the corresponding classical spin dynamics on a unit sphere, Eqs (2)–(4). For the linear nonHermitian Hamiltonian (1), this is also equivalent to the eigenvalue problem for a twolevel system described by the Hamiltonian:
where σ_{i} are Pauli matrices, which yields the following complex eigenspectrum:
The two fixed points of classical spin dynamics governed by the Hamiltonian (1), referred to as A and B in the text, are defined by the following expression in stereographic projection coordinates:
To derive nonreciprocal time evolution of classical spin, we consider the eigenspectrum of the Hamiltonian (5) in twodimensional parameter space of mutually orthogonal magnetic fields, \(({h}_{x},{h}_{y})\), at fixed β. Figure 2(a–c) show the real and imaginary parts of the eigenspectrum (6) at (a) \(\beta =0\) and (b), (c) \(\beta =1\). In the first (Hermitian) case, the spectral surfaces meet at a single diabolical point (DP) \({h}_{x}={h}_{y}=\beta =0\). It is a spectral degeneracy that occurs due to simple symmetries and disappears under generic perturbations. In the Hermitian case, states on both spectral surfaces are stable, with two fixed points of classical spin dynamics classified as centers. Distinct surface colors in Fig. 2(b–f) correspond to opposite signs of imaginary parts of the eigenvalues, with the states on the blue (orange) energy surface being dynamically (un)stable, i.e. corresponding to spin directions of (un)stable equilibria. At \(\beta \ne 0\), i.e. in the nonHermitian case, the diabolical point splits into two chiral EPs located at \(({h}_{x}=\pm \,\beta ,{h}_{y}=0)\). The black solid line \({h}_{y}=0\) in Fig. 2(b–f) is the line of \({\mathscr{P}}{\mathscr{T}}\) symmetry, where the Hamiltonian (1) is \({\mathscr{P}}{\mathscr{T}}\)symmetric^{19,20}. The interval of this line between the EPs corresponds to the regime of broken \({\mathscr{P}}{\mathscr{T}}\) symmetry, with the eigenvalues forming complex conjugate pairs, while on the part of the line outside of this interval the eigenvalues are purely real. Note that in the general case, \({h}_{y}\ne 0\), the Hamiltonian (1) is not \({\mathscr{P}}{\mathscr{T}}\)symmetric and has a complex eignespectrum. It is the reducibility of the nonHermitian Hamiltonian to the \({\mathscr{P}}{\mathscr{T}}\)symmetric form that produces EPs in the eigenspectrum.
Figure 2(b,c) show the naïve picture of encircling a single EP in 2dimensional parameter space, with the state evolution following a 4πperiodic closed trajectory (red line) by visiting both, top and bottom, spectral surfaces before returning to the starting point. This picture is generally incorrect in the case of dynamic encircling of an EP with control parameters varied cyclically at finite speed. Instead of adiabatic evolution, the system experiences sudden jumps between the two spectral surfaces, in the direction from the unstable to the stable one, see Fig. 2(d–f). The jumps occur not immediately upon crossing the line of \({\mathscr{P}}{\mathscr{T}}\) symmetry into the unstable region, but after a certain time interval, illustrating the effect of stability loss delay^{15}.
Let us consider the physics of encircling an isolated EP in parameter space \(({h}_{x},{h}_{y})\) of the model (1) by applying timedependent magnetic field. Encircling an EP reveals its chiral nature and provides access to topological nonreciprocal physical effects unique to nonHermitian systems. In Fig. 2(d), we present parametric trajectories starting on different energy surfaces and going around the EP in opposite directions. For fixed model parameters \(({h}_{x},{h}_{y},\beta )\), the zcomponent of the starting point of each trajectory, i.e. the value of complex energy at \(t=0\), is determined by the initial spin orientations s(0), which can take any direction on the unit sphere. For the purpose of Fig. 2(b–f), we only consider the initial spin orientations corresponding to the two fixed points of spin dynamics A and B, which confines the start and end points of the trajectories in Fig. 2(b–f) to the spectral surfaces and best illustrates nonreciprocal dynamics. Numerical simulations of spin evolution show that the end states for parametric trajectories starting on the top and bottom spectral surfaces in Fig. 2(d) and encircling the EP in closed loops of the same orientation always coincide. On the contrary, protocols with opposite encircling directions lead to mutually different end states, either on the top or bottom energy surfaces, and, therefore, different final spin orientations after any number of complete revolutions around the EP. It has been pointed out in the literature that in order to observe such nonreciprocal behaviour, the start/end points of closed loop trajectories much lie on the \({\mathscr{P}}{\mathscr{T}}\)symmetry line, where both eigenstates are neither stable nor unstable^{15,16,24,25,26,27,28}. In fact, the effect of nonreciprocal dynamics stems from the process of crossing the line of \({\mathscr{P}}{\mathscr{T}}\)symmetry in opposite directions and not strictly from the topology of parametric trajectories. It turns out that to achieve nonreciprocity it is sufficient to realize a closed loop protocol in the vicinity of a single EP without actually encircling it^{29}, see Fig. 2(e), or to encircle both EPs, as shown in Fig. 2(f).
Consider spin dynamics governed by the Hamiltonian (5) under the following two protocols of encircling the isolated EP \((\beta =1,{h}_{x}={h}_{0},{h}_{y}=0)\):
For the purpose of this work, we focus on circular trajectories centered at the EP, with \({h}_{0}=1\), see Fig. 2(d). Suppose the system is initialized at \(t=0\) with phase \({\varphi }_{0}=0\) and subjected to the protocol P_{1}, which takes the system clockwise around the EP. Regardless of the initial spin orientation, i.e. location of the starting point with respect to the two spectral surfaces, the spin will eventually saturate in the direction of the stable fixed point and quickly converge to the state on the top energy surface, even if it was initialized in the unstable fixed point (in which case the transition occurs rapidly via a nonadiabatic stateflip process after a finite time delay). The clockwise protocol P_{1} ensures that the system’s final state after a full rotation around the EP is on the bottom spectral surface, as shown in Fig. 2(d). Following a completely analogous argument, one can see that the encircling protocol P_{2}, which drives the system anticlockwise around the EP, results in a different final state  on the top spectral surface, regardless of the initial spin direction at \(t=0\). This illustrates the chiral nature and nonreciprocity of parametrically driven dynamics near the EP.
As we have pointed out, in order to exploit the nonreciprocal properties discussed above, the encircling protocols must start and end near the line of \({\mathscr{P}}{\mathscr{T}}\) symmetry. Only such trajectories in parameter space can take the system to different final states upon crossing the line of \({\mathscr{P}}{\mathscr{T}}\) symmetry in opposite directions, e.g. by completing a single clockwise or anticlockwise closed loop around the EP. The question of tolerance to deviations from this condition, as well as optimization of other protocol parameters to achieve consistent nonreciprocal behaviour, have not been thoroughly studied in the literature. Here we investigate the range of initial conditions and parameters of the encircling protocols \({P}_{1,2}\) optimal for observing the nonreciprocal evolution. We performed numerical simulations of spin dynamics for a range of initial phases \({\varphi }_{0}\), radii \({R}_{0}\), and driving frequencies \(\omega \) of the encircling protocols, Eq. (8). The final state of the system after executing the protocols \({P}_{1,2}\) correspond to the fixed points \({z}_{A,B}\), see Eq. (7). To quantify deviation of the end spin state after each encircling protocol from the corresponding fixed point, we define deviation angles \({\epsilon }_{i}({\varphi }_{0},\omega )\) between the spin direction at the end of a protocol \({P}_{i}\), \(i=1,2\), and the direction corresponding to fixed point \({z}_{A,B}\). As an absolute measure of deviation from perfect nonreciprocity, we take the maximum deviation between the two protocols, \(\epsilon ({\varphi }_{0},\omega )\equiv \,max\,[{\epsilon }_{1}({\varphi }_{0},\omega ),{\epsilon }_{2}({\varphi }_{0},\omega )]\).
In Fig. 3(a) we present the results of numerical simulations of spin dynamics following the protocols \({P}_{\mathrm{1,2}}\), Eq. (8), with \({h}_{0}=1\) and \(\omega =0.1\), centered around the EP, for a range of parameters \({R}_{0}\) and \({\varphi }_{0}\), which determine the starting point of both protocols in the \(({h}_{x},{h}_{y})\)plane. The color map shows the maximum deviation parameter \(\epsilon \), measured in radians, in the logarithmic scale. A wide range of initial protocol parameters \(({h}_{x},{h}_{y})\) that favor highefficiency nonreciprocal spinfilter effect lie in the central blue region around the EP. The interval directly between the EPs exhibits poor nonreciprocity because it is adjacent to the region of broken \({\mathscr{P}}{\mathscr{T}}\)symmetry of the Hamiltonian (1) with \({h}_{y}=0\), where the eigenspectrum is purely imaginary, see Fig. 2(b,c), corresponding to one stable and one unstable fixed points of spin dynamics. As a result, the final states of the system in the vicinity of this region in the \(({h}_{x},{h}_{y})\)plane are almost identical regardless of the protocol encircling direction. The boundary of the low deviation region is set by specific properties of the stability loss delay effect, discussing which is beyond the scope of the present work and will be addressed elsewhere. Note that the highestefficiency nonreciprocal protocols in Fig. 2(a) have a relatively large radius, such that they encircle both EPs. This is an important result that extends the idea of encircling a single EP to achieve nonreciprocity to parametric trajectories that go around both isolated EPs to achieve an even more robust effect.
In Fig. 3(b), the deviation quantity \(\epsilon \) is plotted for protocols starting on the circle of radius \({R}_{0}=0.5\) around the EP, shown by the black dashed line in Fig. 3(a), for a range of initial phases \({\varphi }_{0}\) and protocol frequencies \(\omega \). This result illustrates that lower encircling frequencies result in smaller deviation parameter \(\epsilon \), i.e. higher efficiency of the nonreciprocal effect. Figure 3(c) shows dependence of the deviation parameter on protocol radius \({R}_{0}\) for the same range of frequencies \(\omega \), calculated at fixed initial phase \({\varphi }_{0}=0\). The initial conditions for the protocols studied in Fig. 3(c) are chosen on the line of \({\mathscr{P}}{\mathscr{T}}\)symmetry, shown as a white dashed line in Fig. 3(a). In the limit of slow encircling speeds, the protocols with larger radius result in smaller deviation \(\epsilon \), even when both EPs are encircled by the protocols (\({R}_{0} > 2\)). The example of nonreciprocal spin dynamics that results from parametric encircling of both EPs is presented in Fig. 2(f). While Fig. 3(c) shows relatively low degree of deviation from nonreciprocity in the case of parametric trajectories with small radius \({R}_{0}\), the practical importance of protocols that lie in the immediate vicinity of the EP is weakened by the fact that eigenstates coalesce in the EP. As a result, the angle between spin orientation given by the fixed points \({z}_{A}\) and \({z}_{B}\) disappears in the limit \({R}_{0}\to 0\).
Note that it is not necessary to encircle an EP to observe nonreciprocity, see Fig. 2(e). Periodic trajectories in the vicinity of an EP can achieve the same effect^{29}. What is crucial for realizing nonreciprocal dynamics is the direction in which the protocols cross the line of \({\mathscr{P}}{\mathscr{T}}\)symmetry, where fixed points of the corresponding dynamics change their stability properties. In fact, parametric trajectories may not enclose any area at all and still achieve a highefficiency nonreciprocal state conversion. The study of various trajectory types and the robustness of resultant nonreciprocal dynamics will be considered elsewhere.
Methods
The numerical simulations were carried out in Wolfram Mathematica^{30}. Each data point in Fig. 3(a–d) was obtained by averaging results of 10 separate simulation runs with arbitrary initial spin orientations uniformly distributed on a unit sphere at protocol initiation.
Discussion
To summarize, there has been a remarkable advance in exploiting nonHermitian systems and their distinct feature, spectral exceptional points. The stage for this nonHermitian physics is however restricted mostly to optical and nanophotonic systems. Here we extended the exploration of EP singularities onto another broadest class of physical systems, spin systems. We demonstrated that spin systems can host EP singularities and thus become a crucial element for realizing devices exhibiting nontrivial topological transport seen so far only in optical systems. This establishes that nonreciprocal effects are ubiquitous in open dissipative systems with gain and loss. Implementation of topological spin transfer enabled us to propose a highefficiency asymmetric spin filter device based on nonreciprocal spin dynamics around an EP of the corresponding nonHermitian Hamiltonian. As a prototypical model, we studied time evolution of a single classical spin in presence of timedependent magnetic field and spintransfer torque. Appearance of EPs in the Hamiltonian’s eigenspectrum allows for encircling protocols that result in different final spin states despite identical initial conditions. We considered two protocols with opposite encircling orientations and showed that such a scheme can be used as a highefficiency spin filter. Optimal protocol parameters are suggested, indicating that encircling trajectories in parameter space of applied magnetic fields must start and end near the line of \({\mathscr{P}}{\mathscr{T}}\) symmetry, with slower encircling frequencies resulting in higher efficiency of the asymmetric spinfilter effect.
References
Damon, R. W. & Eshbach, J. Magnetostatic modes of a ferromagnet slab. J. Phys. Chem. Solids 19, 308 (1961).
Dzyaloshinsky, I. A thermodynamic theory of “weak” ferromagnetism of antiferromagnetics. J. Phys. Chem. Solids 4, 241 (1958).
Moriya, T. Anisotropic superexchange interaction and weak ferromagnetism. Phys. Rev. 120, 91 (1960).
Zakeri, K. et al. Asymmetric spinwave dispersion on Fe(110): Direct evidence of the DzyaloshinskiiMoriya interaction. Phys. Rev. Lett. 104, 137203 (2010).
Iguchi, Y., Uemura, S., Ueno, K. & Onose, Y. Nonreciprocal magnon propagation in a noncentrosymmetric ferromagnet LiFe_{5}O_{8}. Phys. Rev. B 92, 184419 (2015).
Sato, T. J. et al. Magnon dispersion shift in the induced ferromagnetic phase of noncentrosymmetric MnSi. Phys. Rev. B 94, 144420 (2016).
Bender, C. M. & Boettcher, S. Real spectra in nonHermitian Hamiltonians having PT symmetry. Phys. Rev. Lett. 80, 5243 (1998).
Bender, C. M., Boettcher, S. & Meisinger, P. N. PTsymmetric quantum mechanics. J. Math. Phys. 40, 2201 (1999).
Moiseyev, N. NonHermitian quantum mechanics (Cambridge University Press, 2011).
Heiss, W. Repulsion of resonance states and exceptional points. Phys. Rev. E 61, 929 (2000).
Heiss, W. & Harney, H. The chirality of exceptional points. Eur. Phys. J. D 17, 149 (2001).
Heiss, W. Exceptional points–their universal occurrence and their physical significance. Czech. J. Phys. 54, 1091 (2004).
Berry, M. V. Physics of nonhermitian degeneracies. Czech. J. Phys. 54, 1039 (2004).
Dembowski, C. et al. Experimental observation of the topological structure of exceptional points. Phys. Rev. Lett. 86, 787 (2001).
Doppler, J. et al. Dynamically encircling an exceptional point for asymmetric mode switching. Nature 537, 76 (2016).
Choi, Y., Hahn, C., Yoon, J. W., Song, S. H. & Berini, P. Extremely broadband, onchip optical nonreciprocity enabled by mimicking nonlinear antiadiabatic quantum jumps near exceptional points. Nat. Commun. 8, 14154 (2017).
Ruschhaupt, A., Delgado, F. & Muga, J. Physical realization ofsymmetric potential scattering in a planar slab waveguide. J. Phys. Math. Gen. 38, L171 (2005).
Slonczewski, J. C. Currentdriven excitation of magnetic multilayers. J. Magn. Magn. Mater. 159, L1–L7 (1996).
Galda, A. & Vinokur, V. M. Paritytime symmetry breaking in magnetic systems. Phys. Rev. B 94, 020408 (2016).
Galda, A. & Vinokur, V. Linear dynamics of classical spin as Möbius transformation. Sci. reports 7, 1168 (2017).
Lieb, E. H. The classical limit of quantum spin systems. Commun. Math. Phys. 31, 327–340 (1973).
Stone, M., Park, K.S. & Garg, A. The semiclassical propagator for spin coherent states. Journ. Math. Phys. 41, 8025–8049 (2000).
Radcliffe, J. Some properties of coherent spin states. J. Phys. A: Gen. Phys. 4, 313 (1971).
Uzdin, R., Mailybaev, A. & Moiseyev, N. On the observability and asymmetry of adiabatic state flips generated by exceptional points. J. Phys. Math. Theor. 44, 435302 (2011).
Gilary, I., Mailybaev, A. A. & Moiseyev, N. Timeasymmetric quantumstateexchange mechanism. Phys. Rev. A 88, 010102 (2013).
Milburn, T. J. et al. General description of quasiadiabatic dynamical phenomena near exceptional points. Phys. Rev. A 92, 052124 (2015).
Hassan, A. U., Zhen, B., Soljačić, M., Khajavikhan, M. & Christodoulides, D. N. Dynamically encircling exceptional points: exact evolution and polarization state conversion. Phys. Rev. Lett. 118, 093002 (2017).
Zhang, X.L., Wang, S., Hou, B. & Chan, C. Dynamically encircling exceptional points: in situ control of encircling loops and the role of the starting point. Phys. Rev. X 8, 021066 (2018).
Hassan, A. U. et al. Chiral state conversion without encircling an exceptional point. Phys. Rev. A 96, 052129 (2017).
Wolfram Mathematica, Version 11.0, Wolfram Research, Inc., Champaign, IL, 2016.
Acknowledgements
A.G. and V.M.V. were supported by the U.S. Department of Energy, Office of Science, Basic Energy Sciences, Materials Sciences and Engineering Division.
Author information
Authors and Affiliations
Contributions
A.G. and V.M.V. conceived the work, A.G. carried out the calculations, both authors discussed the results, and wrote the paper.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Galda, A., Vinokur, V.M. Exceptional points in classical spin dynamics. Sci Rep 9, 17484 (2019). https://doi.org/10.1038/s41598019534550
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598019534550
This article is cited by

Nonreciprocal phase transitions
Nature (2021)

Steering magnonic dynamics and permeability at exceptional points in a parity–time symmetric waveguide
Nature Communications (2020)
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.