Abstract
Two hundred years ago, Ampère discovered that electric loops in which currents of electrons are generated by a penetrating magnetic field can mutually interact. Here we show that Ampère’s observation can be transferred to the quantum realm of interactions between triangular plaquettes of spins on a lattice, where the electrical currents at the atomic scale are associated with the orbital motion of electrons in response to the noncoplanarity of neighbouring spins playing the role of a magnetic field. The resulting topological orbital moment underlies the relation of the orbital dynamics with the topology of the spin structure. We demonstrate that the interactions of the topological orbital moments with each other and with the spins form a new class of magnetic interactions \(\) topological–chiral interactions \(\) which can dominate over the Dzyaloshinskii–Moriya interaction, thus opening a path for realizing new classes of chiral magnetic materials with threedimensional magnetization textures such as hopfions.
Introduction
Exotic magnetic textures with particlelike properties^{1,2,3,4,5,6} offer great potential for innovative spintronic applications^{7} and braininspired computing^{8,9}. Magnetic skyrmions, twodimensional (2D) localized solitons, are a prominent realization of chiral spin structures, first observed in the material class of noncentrosymmetric B20 bulk compounds^{1}. The potential of spintronic applications would change fundamentally if the line of thought could be continued to the emergence of threedimensional (3D) localized magnetic solitons, e.g., hopfions^{10,11,12}. Recently, a 3D lattice of 3D magnetic textures on the nanometer scale was observed in the B20type cubic chiral magnets MnGe^{13,14}. Despite the strong interest in this magnet, a complete theoretical model for the underlying magnetic interactions is remarkably elusive until now. While, for instance, the basic magnetic properties of the 2D skyrmions are determined by an intricate competition involving the Heisenberg exchange and the chiral relativistic Dzyaloshinskii–Moriya interaction^{15,16} (DMI), such models fail to explain the 3Dmagnetic texture observed in MnGe^{17}.
The 3D magnetization textures of 2D skyrmions gives rise to a scalar spin chirality, a driving force behind a plethora of macroscopic phenomena. Examples are the topological Hall effect^{18,19} or a finite topological orbital moment (TOM)^{20,21,22,23,24,25}, which can both serve as experimental fingerprints of skyrmions. Textureinduced contributions to these macroscopic phenomena were also predicted in frustrated magnets^{26,27}, where they originate from the nontrivial spin topology associated with the realspace configuration of magnetic moments \({{\bf{S}}}_{i}\) as reflected by the scalar spin chirality \({\chi }_{ijk}={{\bf{S}}}_{i}\cdot ({{\bf{S}}}_{j}\times {{\bf{S}}}_{k})\). Although the net spin magnetization might vanish, the symmetry of these chiral systems allows for lowering the energy by preferring orbital currents of specific rotational sense^{26,28}. As a consequence, the motion of the electron in the complex magnetic background manifests itself in the finite TOM without any reference to typical relativistic mechanisms. Instead, this response is usually ascribed to an emergent magnetic field \({{\bf{B}}}^{{\rm{eff}}}\) that roots in the noncoplanar spin texture, giving rise to spontaneous orbital currents, see Fig. 1. While these nonrelativistic currents have been so far largely overlooked, only lately, the perception that they could contribute to the energetics of spin systems is nascent.
Here, based on microscopic arguments and a systematic totalenergy expansion, we discover a conceptually new class of chiral interactions between spins on triangular plaquettes, which originates from the TOM of electrons. We refer to these interactions as topological–chiral interactions, favouring the emergence of noncoplanar magnetic structures with scalar spin chirality of specific sign even without an external magnetic field, either in the ground state or as a result of thermal fluctuations^{29}. The first type of topological–chiral interactions is the rotationinvariant chiral–chiral interaction (CCI), which in its general form corresponds to the interaction between pairs of topological orbital currents in a magnet, just in analogy to Ampère’s observation that the force between wires can be described by the effective interaction of currents. The second type of topological–chiral interactions is the rotationally anisotropic spin–chiral interaction (SCI), which arises as a result of a direct coupling between the TOM and local spins, mediated by the spinorbit interaction. We uncover the importance of the discovered topologicalchiral interactions for the energetics of spin systems by explicit firstprinciples calculations in B20 magnets. Finally, if the emerging magnetic textures can be represented by continuous magnetization fields, we show that systems described by the conventional Heisenberg and the chiral–chiral interactions form a physical realization of the Faddeev model^{30} with hopfion solutions^{10,11,12}. This signifies the key role of the topologicalchiral interactions in triggering the formation of 3D magnetic solitons without the assistance of an external magnetic field.
Results
Topological orbital moment
We briefly elucidate the concept of emergent orbital currents due to the motion of electrons through a noncollinear magnetic structure^{26,28}. In the absence of the spinorbit interaction (SOI), hopping in a noncoplanar spin texture with finite chirality \({\chi }_{ijk}={{\bf{S}}}_{i}\cdot \left({{\bf{S}}}_{j}\times {{\bf{S}}}_{k}\right)\), where \({\bf{S}}\) are direction vectors of classical spin moments located at sites \(i,j,k\), is equivalent to the electron dynamics in a coplanar spin background but in the presence of a fictitious magnetic field \({B}^{{\rm{eff}}}\ \propto \ {\chi }_{ijk}\) directed along \({{\boldsymbol{\tau }}}_{ijk}\)^{31}, which gives rise to a Berry phase^{26}. Here, the vector \({{\boldsymbol{\tau }}}_{ijk}\ \propto \ ({{\bf{R}}}_{j}{{\bf{R}}}_{i})\times ({{\bf{R}}}_{k}{{\bf{R}}}_{i})\) is the surface normal of the oriented triangle spanned by the lattice sites \({{\bf{R}}}_{i}\), \({{\bf{R}}}_{j}\), and \({{\bf{R}}}_{k}\) at which spin moments are placed. Acting as an effective magnetic field, the complex spin topology of chiral systems thereby allows for groundstate currents of specific rotational sense that manifest in spontaneous orbital properties of the electrons. This microscopic mechanism induces a topological orbital moment (TOM) that stems from the scalar spin chirality without any reference to relativistic origins, as predicted for several situations ranging from noncollinear \(3Q\)states to largescale skyrmions^{20,21,22,23,24,25}. In the limit of vanishing effective magnetic field, i.e., in a situation of small spin chirality, the spontaneous TOM is directly proportional to \({B}^{{\rm{eff}}}\) with the proportionality factor given by the topological orbital susceptibility of the system^{25}.
As illustrated in Fig. 1, the TOM which roots in emergent orbital currents around a given spin triangle can be linked phenomenologically to the triangle normal and the corresponding scalar spin chirality via \({{\bf{L}}}_{ijk}^{{\rm{TO}}}={\kappa }_{ijk}^{{\rm{TO}}}{\chi }_{ijk}{{\boldsymbol{\tau }}}_{ijk}\), where \({\kappa }_{ijk}^{{\rm{TO}}}\) is the local topological orbital susceptibility. Using these individual orbital moments, we quantify the local TOM at the \(i\)th atomic site by taking into account that each spin participates in several triangles:
where the sum over \((jk)\) is restricted to triangles including the \(i\)th magnetic atom in the unit cell. In the following, we demonstrate that these orbital properties are key entities of a novel class of chiral exchange interactions, which we refer to as topological–chiral interactions imprinting on the groundstate energetics of spin systems. They appear in two distinct realizations.
Chiral–chiral interaction
When following the rationale of Ampère’s law down to the quantum scale, it is intuitively clear that the mutual interaction between emergent orbital currents due to the scalar spin chirality contributes to the total energy of chiral systems. While we derive this nonrelativistic interaction explicitly in Supplementary Note 1 from a systematic totalenergy expansion based on multiple scattering theory (see Methods), here, we motivate its final form in terms of macroscopic physical properties. As depicted in Fig. 1, in the absence of SOI, the resulting chiral–chiral interaction (CCI) correlates emergent orbital currents occurring on spin triangles, just in analogy to Ampère’s force law correlating electrical currents. The most dominant contribution to the total energy is expected to originate from currents around the same plaquette such that the local part of the CCI can be interpreted as the orbital Zeeman interaction \({{\bf{L}}}^{{\rm{TO}}}\cdot {{\bf{B}}}^{{\rm{eff}}}\) of the spontaneous TOM with the emergent magnetic field of the noncoplanar spin background:
where the coefficient \({\varkappa }_{ijk}^{{\rm{CC}}}\) describes the strength of the CCI within the triangle formed by \(i\), \(j\), and \(k\).
Remarkably, this nonrelativistic exchange interaction, which is quadratic in \({\chi }_{ijk}\), emerges without any external magnetic field as it is intrinsic to the ground state of complex spin textures. Symmetry allows for such type of interaction since time reversal inverts all spins and changes the sign of the chirality, but it keeps the energy of the spin system invariant. Adopting the quantummechanical viewpoint of ref. ^{32}, the form of CCI can be also interpreted as a 6spin–3site interaction, which requires quantum mechanically speaking at least a spin1 system.
Spinchiral interaction
A valence electron experiencing the emerging orbital moment, \({{\bf{L}}}_{i}^{{\rm{TO}}}\), connected to its orbital motion around the \(i\)th ion in a noncoplanar texture is also exposed to the nuclear electric field acting as a magnetic field and coupling the orbital moment to its spin. Thus, this spinchiral interaction (SCI) is a second element of the topologicalchiral interaction, and it is a consequence of the relativistic SOI coupling the TOM to single spin magnetic moments, as illustrated in Fig. 1. The corresponding energy of the SCI can be shown (see Methods and Supplementary Note 1) to assume the form
where \({\varkappa }_{i}^{{\rm{SC}}}\) is the spinchiral coupling strength of magnetic atom \(i\). Being linear in the scalar spin chirality, this relativistic 4thorder interaction is rotationally anisotropic.
Conceptually, the SCI energy as given by Eq. (2) is similar to the magnetocrystalline anisotropy energy (MAE), which prefers a maximal projection of the spinorbit induced orbital moment onto the local spin quantization axis^{33,34}. While the chiralchiral coupling roots solely in the spin configuration, the spinchiral coupling is very sensitive to the lattice structure, just like MAE and antisymmetric DMI exchange as discussed in Supplementary Note 1. In contrast to the MAE, however, the SCI is susceptible to the local chirality, which gives rise to its outstanding feature: the SCI energy favours states with a certain chirality as exemplified for real materials below. The SCI is also distinct from the recentlyuncovered chiral biquadratic interaction, which generalizes the DMI to a fourspin interaction but on two sites^{35}. Thus, the SCI is a novel type of magnetic interaction that can replace the DMI in stabilizing chiral ground states.
The form of Eqs. (1)–(3) is adapted to the case of cubic crystals for which the local constants \({\varkappa }_{ijk}^{{\rm{CC}}}\), \({\kappa }_{ijk}^{{\rm{TO}}}\), and \({\varkappa }_{ijk}^{{\rm{SC}}}\) are scalar quantities. In general, however, these constants are tensor quantities, and we provide generalized expressions for the proposed interactions as well as the extended spin Hamiltonian in Supplementary Notes 1, 2, and 3.
Topologicalchiral interactions in real materials
Next, we substantiate the importance of the topological–chiral interactions taking the intensively scrutinized B20 magnet MnGe as one specific example. In contrast to FeGe that forms wellunderstood twodimensional skyrmions with a radius of 70 nm, for MnGe puzzling threedimensional skyrmion lattices with lattice constants of a few nanometers were observed experimentally^{13,18}, but could not be reproduced by any theoretical calculation so far^{36,37}. Employing electronicstructure theory (see Methods), we show below that while the chiralchiral and spinchiral interactions are suppressed in FeGe, where chiral physics is dominated by the Dzyaloshinskii–Moriya interaction (DMI)^{15,16}, the topologicalchiral interactions are very prominent in MnGe.
The chiral structure of the cubic B20 compounds consists of four sublattices that are structurally identical, but their local threefold rotation axes point along different cube diagonals \({{\bf{n}}}_{i}\), see Fig. 2a. We describe these systems by an effective spinlattice Hamiltonian that includes Heisenberg and 4thorder exchange, DMI, magnetic anisotropy, and the proposed topologicalchiral interactions, see Eq. (8) in the Methods section. To disentangle the different types of magnetic interactions, we consider a set of spin configurations which corresponds to one of the possible irreducible representations of the symmetry group of the paramagnetic phase (see Methods), and can be parameterized by
where the unit vector \({{\bf{S}}}_{i}\) refers to the direction of the classical spin moment of atom type \(i\in \{1,\ldots ,4\}\). \(\theta\) and \(\phi\) are the polar and azimuthal angles, respectively, and \({{\bf{n}}}_{i}\) points along a cube diagonal, as illustrated in Fig. 2a, b.
This choice of magnetic structures greatly simplifies the analysis as the energy contribution of the classical Heisenberg interaction \(({{\bf{S}}}_{i}\cdot {{\bf{S}}}_{j})\) among spins following Eq. (4) is independent of \(\theta\) and \(\phi\). Consequently, any nontrivial angular dependence of the total energy will indicate the presence of higherorder magnetic interactions. In fact, our firstprinciples calculations of the total energy in absence of spinorbit coupling, see Fig. 3, show drastic energy variations with respect to \(\theta\) and \(\phi\) on the order of several tens of meV in both FeGe and MnGe, which discloses the absolute significance of beyondHeisenberg terms in these materials. However, while 4thorder exchange interactions of the form \(({{\bf{S}}}_{i}\cdot {{\bf{S}}}_{j})({{\bf{S}}}_{k}\cdot {{\bf{S}}}_{l})\)^{32} indeed describe the energy variation with \(\theta\) well in FeGe, their contribution deviates strongly from the calculated curve in MnGe if the spins are not coplanar, \(\phi ={0}^{\circ }\), but noncoplanar, e.g., for \(\phi =4{5}^{\circ }\) (see Supplementary Note 3 and Supplementary Table 1 for further details).
To uncover the distinct nature of these large qualitative discrepancies between 4thorder exchange and the computed totalenergy variation in noncoplanar MnGe, we consider the physically motivated CCI, Eq. (2), as given by the coupling of emergent orbital currents due to the complex spin topology. Taking into account the crystal symmetries of the B20 compounds and the spin configurations Eq. (4), we find that \({E}^{{\rm{CC}}}=2{\varkappa }^{{\rm{CC}}}{[\sin \theta \sin 2\theta \sin 2\phi ]}^{2}\), with a single materialspecific coupling constant \({\varkappa }^{{\rm{CC}}}\). Fitting this angular dependence to our ab initio results significantly improves the description of the totalenergy variation, capturing now all essential features also for MnGe, Fig. 3b. Modelling additionally changes of the spinmoment length with \(\theta\) enhances the agreement with our calculations even further (see Supplementary Note 4 and Supplementary Fig. 1). The coupling constant \({\varkappa }^{{\rm{CC}}}\) of the CCI amounts to \(1.2\) meV and \(59.9\) meV in FeGe and MnGe, respectively, which underlines the absolute relevance of the proposed interaction in the latter system. Since the spin moment is \(0.78\ {\mu }_{{\rm{B}}}\) in FeGe but amounts to \(1.96\ {\mu }_{{\rm{B}}}\) for MnGe, this finding of a significant energy contribution due to the chiral–chiral coupling is consistent with the quantummechanical mechanism for the CCI, which necessitates at least a spin1 system.
By computing explicitly the orbital moment without SOI on each magnetic atom, see Fig. 4b, we further substantiate the distinct microscopic origin of the chiral–chiral coupling in MnGe. Based on the angular dependence of \({\chi }_{ijk}\) for the choice (4) of magnetic structures, Eq. (1) leads to \({{\bf{L}}}_{i}^{{\rm{TO}}}={\kappa }^{{\rm{TO}}}\sin \theta \sin 2\theta \sin 2\phi \ {{\bf{n}}}_{i}\) such that the TOM at the \(i\)th atom is collinear to the cube diagonal \({{\bf{n}}}_{i}\), and the magnitude is proportional to both the spin chirality and the single effective topological orbital susceptibility constant \({\kappa }^{{\rm{TO}}}\). This analytic expression is in perfect agreement with the ab initio calculations of the local orbital moment in absence of SOI as shown in Fig. 4b, revealing that the topological orbital susceptibility is of opposite sign in the two compounds, namely, \({\kappa }_{{\rm{FeGe}}}^{{\rm{TO}}}=0.02\ {\mu }_{{\rm{B}}}\) and \({\kappa }_{{\rm{MnGe}}}^{{\rm{TO}}}=0.13\ {\mu }_{{\rm{B}}}\). Our results clearly demonstrate that the TOM in MnGe, and specifically its variation with \(\theta\), exceeds the effect in FeGe by roughly an order of magnitude, which underpins the importance of the predicted CCI in the former material.
While our predictions of isotropic 4thorder and chiral–chiral contributions to the exchange energy are independent of the spin–orbit coupling, we elucidate now the role of relativistic SOI for the angular dependence of the total energy. Although the magnetocrystalline anisotropy is uniaxial to lowest order in the noncollinear B20 magnets (see Supplementary Note 5), we explicitly verified that the corresponding energy contribution is negligibly small as compared to all other terms considered here.
Figure 4c displays the projection \(({{\bf{L}}}_{i}^{{\rm{TO}}}\ \cdot {{\bf{S}}}_{i})\) computed without consideration of the SOI, which enters the expression (3) for the spin–chiral coupling. The projection is generally much larger in MnGe than in FeGe, with a very pronounced dependence on the chiral spin texture. In sharp contrast, the additional corrections to the total orbital moment due to SOI exhibit only minor modulations with the angle \(\theta\).
As a result of the small projection in case of FeGe, the spinorbit correction of the total energy with respect to \(\theta\) is primarily due to the antisymmetric exchange \(({{\bf{S}}}_{i}\ \times \ {{\bf{S}}}_{j})\) as illustrated in Fig. 5a, whereas the substantial angular dependence in MnGe cannot be described solely by the DMI energy (see also Supplementary Note 3 and Supplementary Table 1). For MnGe, using Eq. (3), spin configuration (4) and a single constant \({\varkappa }^{{\rm{SC}}}\), we obtain the form \({E}^{{\rm{SC}}}=2{\varkappa }^{{\rm{SC}}}{\kappa }^{{\rm{TO}}}{\sin }^{2}2\theta \sin 2\phi \ [1+\tan \theta \ (\cos \phi +\sin \phi )]\) for the energy contribution of the proposed SCI that substantially improves the modelling of the totalenergy variations based on the firstprinciples results, see Fig. 5b. By directly fitting Eq. (3) to our ab initio data, we find that the constant \({\varkappa }^{{\rm{SC}}}{\kappa }^{{\rm{TO}}}\) is 0.2 meV in FeGe, and \(3.2\) meV in MnGe. Thus, we conclude from our analysis that the novel SCI is by far the dominant spinorbit effect in MnGe – contrary to FeGe, where physical properties are hardly affected by the TOM but driven rather by the prominent DMI.
Emergence of threedimensional textures
For magnetization textures with characteristic length scales much larger than the underlying crystal lattice, a continuum representation, \({{\bf{S}}}_{i}\to {\bf{m}}({\bf{r}})\), of the spinlattice Hamiltonians (2) and (3) is able to reflect the essence of the magnetic interactions. Taking this micromagnetic approach, the relevant order parameter of the system is a magnetization density \({\bf{m}}({\bf{r}})=({m}_{x},{m}_{y},{m}_{z})\), \( {\bf{m}}({\bf{r}}) =1\), a unitvector field, defined at any point \({\bf{r}}\in {{\mathbb{R}}}^{3}\). As shown in Supplementary Note 6 the scalar spinchirality \({\chi }_{ijk}\) emerges in the continuum theory as a spinchirality density, which is a dot product of a direction vector of a surface normal and the solenoidal gyrovector field \({\bf{F}}({\bf{r}})\), whose components \({F}_{\alpha }=\sum _{\beta \gamma }{\varepsilon }_{\alpha \beta \gamma }{f}_{\beta \gamma }\), are determined by the unit magnetization field, \({\bf{m}}\), and its spatial derivatives:
where \({\varepsilon }_{\alpha \beta \gamma }\) is the antisymmetric 3D Levi–Civita tensor. The gyrovector field is linearly related to the TOM density. The micromagnetic expression for the energy of the CCI takes then the following form:
where the micromagnetic \({\tilde{\varkappa }}^{{\rm{CC}}}\) is proportional to the microscopic \({\varkappa }^{{\rm{CC}}}\) and it is assumed for simplicity to be a single constant. Together with the Heisenberg exchange, Eq. (6) describes the energy of the magnetization texture as a variant of the highly acclaimed Faddeev model^{30}, the exciting feature of which is that it contains hopfions, stable localized threedimensional knotted topological solitons as solutions. The role of the SCI will be to orient the hopfion relative to the underlying lattice.
Discussion
Opening several exciting vistas in the field of chiral magnetism, our findings raise a number of important fundamental questions. Above all, they call for a review of the relevance of the chiralchiral and spinchiral coupling, uncovered in this study, for the ground state of existing materials that exhibit diverse magnetic orders. This also concerns the systems in which emerging homochiral magnetic structures were previously thought to be the result of the Dzyaloshinskii–Moriya interaction (DMI). In contrast, topological–chiral interactions offer fundamentally different opportunities for imprinting chiral magnetism, as they manifest in the scalar chirality of spin arrangements on triangular plaquettes, as opposed to the vector chirality between pairs of spins in the case of DMI. In the continuum limit, the spinchirality relates to the curvature of the magnetization field and the chiralchiral interaction reverts to the Faddeev model. Thus, magnets with topologicalchiral magnetic interactions offer the first experimental realization of this model with hopfions as the emergent 3D magnetic particles. It could be speculated that the unique 3D magnetic order of MnGe explained in this article by the occurrence of the topological–chiral interaction is a precursor state. Another important aspect to be explored is the influence of the discovered interactions on the dynamical properties of ferromagnetic, chiral, and antiferromagnetic systems. As we show in Supplementary Note 7, the corresponding modifications to the phenomenological model for the free energy of antiferromagnets, brought about by topological–chiral interactions, enable a direct interpretation of magnetic phase transitions at high pressure and finite temperature. This provides a foundation for studying antiferromagnetic dynamics in materials that exhibit the proposed chiralchiral and spinchiral interactions.
Methods
Firstprinciples calculations
We used the FLEUR code (http://www.flapw.de) to calculate the total energy, spin and orbital magnetic moments of MnGe and FeGe in the B20 phase with and without spinorbit interaction for a set of noncollinear magnetic structures within the generalized gradient approximation of Perdew–Burke–Ernzerhof^{38}. The first Brillouin zone was sampled with a 12\(\times\) 12 \(\times\) 12 MonkhorstPack grid. To converge the energy calculations for the singlesite magnetocrystalline anisotropy, the sampling of the full Brillouin zone was increased to 24 \(\times\) 24 \(\times\) 24 \({\bf{k}}\)points and the temperature in the Fermi distribution was reduced from 160 K to 80 K. The planewave cutoff for the basis functions was 4.2 a.u.\(^{1}\). The cubic lattice structure of the B20 compounds^{39} consists of four sublattices, see Fig. 2a, with the ions located at \((u,u,u)\), \((0.5u,1u,0.5+u)\), \((1u,0.5+u,0.5u)\), and \((0.5+u,0.5u,1u)\), where we optimized the structural parameter \(u\) together with the lattice constant \(a\) by minimizing the total energy. The obtained values are \({u}_{{\rm{Fe}}}=0.134,\ {u}_{{\rm{Ge}}}=0.842\) in FeGe with the lattice constant \(a=4.67\) Å, and \({u}_{{\rm{Mn}}}=0.136,\ {u}_{{\rm{Ge}}}=0.843\) in MnGe with \(a=4.76\) Å.
Following this scheme, we computed the energies and orbital moments for different noncollinear antiferromagnetic (AFM) configurations, as described by two angles \(\theta\) and \(\phi\):
Here \({\phi }_{i}=\{\phi ,\phi ,\phi ,\phi \}\), \({\theta }_{i}=\{\theta ,\theta ,\pi +\theta ,\pi \theta \}\), and the vector \({{\bf{n}}}_{i}\) is the local symmetry direction at site \(i\) that points along one of the cube diagonals \((1,1,1)\), \((1,1,1)\), \((1,1,1)\), and \((1,1,1)\), see Fig. 2a. The spin texture given by Eq. (4) forms one of the irreducible representations of possible AFM configurations, which can be classified according to the symmetry properties of the AFM order parameters (see Supplementary Note 7).
Depending on the spin configuration, the magnetic moments of Fe and Mn amount to about \(0.78\ {\mu }_{{\rm{B}}}\) and \(1.96\ {\mu }_{{\rm{B}}}\), respectively, see Supplementary Fig. 1. The tiny induced magnetic moment of Ge (\(<\) \(0.005\ {\mu }_{{\rm{B}}}\)) is given with respect to the local quantization axis, which was chosen to follow the magnetization direction of its nearestneighbour magnetic ion. Based on the ab initio electronic structure, the orbital magnetic moment \({{\bf{L}}}_{i}\) is obtained by locally integrating the angular momentum operator \({\bf{L}}={\bf{r}}\ \times \ {\bf{p}}\) within the muffintin sphere of radius 2.25 and 2.30 a.u. for Fe and Mn, respectively, centered around the \(i\)th nucleus. To extract the topological part \({{\bf{L}}}_{i}^{{\rm{TO}}}\), we evaluate the orbital moment in the absence of spin–orbit coupling.
The changes to the energetics and the orbital moments due to the SOI have been included based on the selfconsistent electronic structure without SOI, using the soobtained vector spin density and including the SOI applying the force theorem^{40}. For phenomenological modelling we applied the Landau theory of phase transitions and introduced antiferromagnetic order parameters based on the symmetry analysis. Further details of modelling are provided in Supplementary Note 7.
Spinlattice Hamiltonian
For our study we considered the following effective spinlattice Hamiltonian:
where \({J}_{ij}\), \({K}_{ijkl}\), \({{\bf{D}}}_{ij}\), \({\varkappa }_{ijk}^{{\rm{CC}}}\), and \({\varkappa }_{ijk}^{{\rm{SC}}}\) mediate Heisenberg, symmetric 4thorder, Dzyaloshinskii–Moriya, chiral–chiral, and spin–chiral couplings, respectively. The tensor \({\underline{\underline{A}}}_{i}\) encodes the singlesite magnetic anisotropy at site \(i\), see Supplementary Note 5, and the unit vector \({{\bf{S}}}_{i}\) refers to the direction of the classical spin moment of atom type \(i\in \{1,\ldots ,4\}\). While the considered terms for CCI, SCI, and magnetic anisotropy in the above expression represent the dominant local contributions to the total energy, a more general form with nonlocal contributions is given in Supplementary Note 2.
Multiple scattering theory
We derived the expressions for the proposed topologicalchiral magnetic interactions by a systematic expansion of the total energy of the manyelectron system with respect to simultaneous infinitesimal rotations of the magnetic moments. By rotating the set of magnetic moments within the solid, their mutual interaction can be obtained from the change of the total energy. To extract the interaction parameters, we invoke a multiple scattering theory realized by the Korringa–Kohn–Rostoker (KKR) Greenfunction formalism. Explicit expressions for the rigorous derivation of the topologicalchiral interactions are provided in Supplementary Note 1.
Data availability
The data that support the findings of this study are available from the corresponding authors upon reasonable request.
References
 1.
Mühlbauer, S. et al. Skyrmion lattice in a chiral magnet. Science 323, 915 (2009).
 2.
Rybakov, F. N., Borisov, A. B., Blügel, S. & Kiselev, N. S. New type of stable particlelike states in chiral magnets. Phys. Rev. Lett. 115, 117201 (2015).
 3.
Nayak, A. K. et al. Magnetic antiskyrmions above room temperature in tetragonal Heusler materials. Nature 548, 561–566 (2017).
 4.
Hoffmann, M. et al. Antiskyrmions stabilized at interfaces by anisotropic DzyaloshinskiiMoriya interactions. Nat. Commun. 8, 308 (2017).
 5.
Zheng, F. et al. Experimental observation of chiral magnetic bobbers in B20type FeGe. Nat. Nanotechnol. 13, 451–455 (2018).
 6.
Ahmed, A. S. et al. Chiral bobbers and skyrmions in epitaxial FeGe/Si(111) films. Phys. Rev. Materials 2, 041401(R) (2018).
 7.
Fert, A., Cros, V. & Sampaio, J. Skyrmions on the track. Nat. Nanotechnol. 8, 152 (2013).
 8.
Bourianoff, G., Pinna, D., Sitte, M. & EverschorSitte, K. Potential implementation of reservoir computing models based on magnetic skyrmions. AIP Adv. 8, 055602 (2018).
 9.
Zázvorka, J. et al. Thermal skyrmion diffusion used in a reshuffler device. Nat. Nanotechnol. 14, 658 (2019).
 10.
Radu, E. & Volkov, M. S. Stationary ring solitons in field theory  Knots and vortons. Phys. Rep. 468, 101 (2008).
 11.
Sutcliffe, P. Skyrmion knots in frustrated magnets. Phys. Rev. Lett. 118, 247203 (2017).
 12.
Rybakov, F. N. et al. Magnetic hopfions in solids. Preprint at https://arXiv.org/abs/1904.00250 (2019).
 13.
Tanigaki, T. et al. Realspace observation of shortperiod cubic lattice of skyrmions in MnGe. Nano Lett. 15, 5438 (2015).
 14.
Fujishiro, Y. et al. Topological transitions among skyrmion and hedgehoglattice states in cubic chiral magnets. Nat. Commun. 10, 1–8 (2019).
 15.
Dzyaloshinskii, I. E. Thermodynamic theory of “weak” ferromagnetism of antiferromagnetics. J. Phys. Chem. Solids 4, 241 (1958).
 16.
Moriya, T. New Mechanism of Anisotropic Superexchange Interaction. Phys. Rev. Lett. 4, 228–230 (1960).
 17.
Bornemann, B. et al. Complex magnetism of B20MnGe: from spinspirals, hedgehogs to monopoles. J. Phys.: Condens. Matter 31, 485801 (2019).
 18.
Kanazawa, N. et al. Large topological Hall effect in a shortperiod helimagnet MnGe. Phys. Rev. Lett. 106, 156603 (2011).
 19.
A. Neubauer, A. et al. Topological Hall effect in the a phase of MnSi. Phys. Rev. Lett. 102, 186602 (2009).
 20.
Hoffmann, M. et al. Topological orbital magnetization and emergent Hall effect of an atomicscale spin lattice at a surface. Phys. Rev. B 92, 020401(R) (2015).
 21.
Hanke, J.P. et al. Role of Berry phase theory for describing orbital magnetism: from magnetic heterostructures to topological orbital ferromagnets. Phys. Rev. B 94, 121114(R) (2016).
 22.
dos Santos Dias, M., Bouaziz, J., Bouhassoune, M., Blügel, S. & Lounis, S. Chiralitydriven orbital magnetic moments as a new probe for topological magnetic structures. Nat. Commun. 7, 13613 (2016).
 23.
Hanke, J.P., Freimuth, F., Blügel, S. & Mokrousov, Y. Prototypical topological orbital ferromagnet \(\gamma\)FeMn. Sci. Rep. 7, 41078 (2017).
 24.
dos Santos Dias, M. & Lounis, S. Insights into the orbital magnetism of noncollinear magnetic systems. SPIE 10357, 103572A (2017).
 25.
Lux, F. R., Freimuth, F., Blügel, S. & Mokrousov, Y. Engineering chiral and topological orbital magnetism of domain walls and skyrmions. Commun. Phys. 1, 60 (2018).
 26.
Taguchi, Y., Oohara, Y., Yoshizawa, H., Nagaosa, N. & Tokura, Y. Spin chirality, Berry phase, and anomalous Hall effect in a frustrated ferromagnet. Science 291, 2573–2576 (2001).
 27.
Shindou, R. & Nagaosa, N. Orbital ferromagnetism and anomalous hall effect in antiferromagnets on the distorted fcc lattice. Phys. Rev. Lett. 87, 116801 (2001).
 28.
Tatara, G. & Kohno, H. Permanent current from noncommutative spin algebra Gen. Phys. Rev. B 67, 113316 (2003).
 29.
Zhang, L.C. et al. Orbital Nernst Effect of Magnons. Preprint at https://arXiv.org/abs/1910.03317 (2019).
 30.
Faddeev, L. D. Quantization of solitons. Preprint IAS Print75QS70 (Inst. Advanced Study, Princeton, NJ, 1975), 32 pp.
 31.
Fujita, T., Jalil, M. B. A., Tan, S. G. & Murakami, S. Gauge fields in spintronics. J. Appl. Phys. 110, 121301 (2011).
 32.
Hoffmann, M. & Blügel, S. Systematic derivation of realistic spinmodels for beyondHeisenberg solids. Preprint at https://arXiv.org/abs/1803.01315 (2018).
 33.
Solovyev, I. V., Dederichs, P. H. & Mertig, I. Origin of orbital magnetization and magnetocrystalline anisotropy in TX ordered alloys (where T=Fe,Co and X=Pd,Pt). Phys. Rev. B 52, 13419 (1995).
 34.
Bruno, P. Tightbinding approach to the orbital magnetic moment and magnetocrystalline anisotropy of transitionmetal monolayers. Phys. Rev. B 39, 865(R) (1989).
 35.
Brinker, S., dos Santos Dias, M. & Lounis, S. The chiral biquadratic pair interaction. New J. Phys. 21, 083015 (2019).
 36.
Gayles, J. et al. DzyaloshinskiiMoriya interaction and Hall effects in the skyrmion phase of Mn\(_{1x}\)Fe\(_{1x}\)Ge. Phys. Rev. Lett. 115, 036602 (2015).
 37.
Koretsune, T., Kikuchi, T. & Arita, R. Firstprinciples evaluation of the DzyaloshinskiiMoriya interaction. J. Phys. Soc. Jpn. 87, 041011 (2018).
 38.
Perdew, J. P., Burke, K. & Ernzerhof, M. Generalized gradient approximation made simple. Phys. Rev. Lett. 77, 3865 (1996).
 39.
Grytsiuk, S. et al. Ab initio analysis of magnetic properties of the prototype B20 chiral magnet FeGe. Phys. Rev. B 100, 214406 (2019).
 40.
Mackintosh, A. R. & Andersen, O. K. in Electrons at the Fermi surface Ch. 5 (Cambridge Univ. Press, Cambridge, 1980).
Acknowledgements
We acknowledge fruitful discussions with Christof Melcher. We acknowledge funding from Deutsche Forschungsgemeinschaft (DFG) through SPP 2137 “Skyrmionics”, the Collaborative Research Centers SFB 1238 and SFB/TRR 173, project MO 1731/51 and project SHARP 397322108. M.H., Y.M, and S.B. acknowledge the DARPA TEE program through grant MIPR# HR0011831554 from DOI, and O.G. acknowledges the EU FET Open RIA Grant no. 766566, Humboldt Foundation, and the ERC Synergy Grant SC2 (No. 610115). S.L. and J.B. express gratitude to the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation program for funding (ERCconsolidator grant 681405 – DYNASORE). Simulations were performed with computing resources granted by JARAHPC from RWTH Aachen University and Forschungszentrum Jülich under projects jara0161, jiff40 and jias1a.
Author information
Affiliations
Contributions
S.B. and Y.M. motivated the project. S.G., G.B., Y.M. and S.B. devised the details of the project. S.G. uncovered the role of chirality for the magnetic interactions by performing the DFT calculations with assistance from M.H., G.B. and J.P.H. The expressions for the proposed interactions (2) and (3) and their corresponding physical picture were developed by S.G. and Y.M. The rigorous derivation of the proposed interactions in terms of Green functions was worked out by J.B. and S.L. The phenomenological model was developed by O.G. and S.G. The manuscript and the supplement were written by J.P.H., S.G., M.H., S.L., Y.M. and S.B. All the authors contributed to the analysis and interpretation of the results.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Peer review information Nature Communications thanks the anonymous reviewer(s) for their contribution to the peer review of this work. Peer reviewer reports are available.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
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
Grytsiuk, S., Hanke, JP., Hoffmann, M. et al. Topological–chiral magnetic interactions driven by emergent orbital magnetism. Nat Commun 11, 511 (2020). https://doi.org/10.1038/s41467019140303
Received:
Accepted:
Published:
Further reading

Spin Hamiltonians in Magnets: Theories and Computations
Molecules (2021)

Short period magnetization texture of B20MnGe explained by thermally fluctuating local moments
Physical Review B (2021)

Exchange interactions and magnetic force theorem
Physical Review B (2021)

Role of Magnetic Exchange Interactions in ChiralType Hall Effects of Epitaxial MnxPtSn Films
ACS Applied Electronic Materials (2021)

Magnetic Skyrmion Materials
Chemical Reviews (2021)
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.