betaFeedback

WorksPhysics and Astronomy

Treasure Map Toward Skyrmion Evolution in Ambient Conditions: A Perspective from Electronic Instabilities and the Density of Energy

Abstract

Magnetic skyrmions with topologically protected properties are anticipated to shape the future of electronics. Understanding how skyrmions may evolve in ambient conditions presents a key challenge in the pursuit of technologically significant materials. In this perspective, we focus on electronic instabilities and the density of energy of established skyrmion hosts, where a pathway to a skyrmion phase transition is readily available, to identify signposts for the emergence of skyrmions. We value the impressive research efforts in the field that have built the foundation for many more enticing breakthroughs to come. We share a framework that connects the electronic origins of skyrmion formation to the temperature and field requirements, allowing predictions of candidate materials that may host skyrmions at ambient conditions (the treasure).

Perspective manuscript

Introduction

Magnetic skyrmions—nanoscale, vortex-like spin textures with topological protection—display unique physical properties, such as the topological Hall effect, magnetic-field-driven or electric-field-driven skyrmion motion, and non-reciprocal response.[1, 2, 3, 4] These characteristics make them attractive for spintronics.[5, 6, 7] Despite remarkable research progress, realizing skyrmions under practical conditions, stable at room temperature and operable in low magnetic or electric fields, remains a significant challenge. While some systems can host skyrmions, they often require cryogenic conditions or high external fields, hindering their potential value in shaping the future of electronics.

Underlying magnetic exchange interactions

The emergence of skyrmions can be understood from two fundamental mechanisms, each rooted in distinct electronic and structural conditions (Figure 1). On one side, skyrmions evolve from the Dzyaloshinskii–Moriya (DM) interaction, which can be fostered in noncentrosymmetric (NCS) lattices with strong spin–orbit coupling.[8, 9, 10, 11, 12, 13] Depending on the local site symmetry, different types of skyrmions can emerge: polar point-group symmetries favor Néel-type skyrmions; chiral symmetries stabilize Bloch-type skyrmions; and D2dD_{2d} symmetry enables the formation of anti-skyrmions. These three symmetry-derived roots lead to distinct spin textures that collectively populate the “DM branch” of the skyrmion family tree. On the other side, the Ruderman–Kittel–Kasuya–Yosida (RKKY) mechanism operates in centrosymmetric metallic systems, where conduction electrons mediate oscillatory exchange interactions.[14, 15, 16] In geometrically frustrated magnetic lattices—such as square, honeycomb, or kagome networks—this oscillatory coupling bends spin orientations across the lattice, producing Bloch-type skyrmions without requiring inversion symmetry breaking. These frustrated-lattice roots define the “RKKY branch,” providing a parallel path to topological spin textures through electronic frustration rather than structural asymmetry.

The Skyrmion tree showing the formation mechanisms

Figure 1. The Skyrmion tree showing the formation mechanisms. The roots of the tree indicate the underlying driving force. Dzyaloshinskii-Moriya interactions, facilitated by larger spin-orbital coupling, stabilize skyrmions in non-centrosymmetric systems. Ruderman-Kittel-Kasuya-Yosida exchange interactions, assisted by geometrically frustrated lattice systems, stabilize skyrmions in centrosymmetric metallic systems.

Current stage of knowledge

The DM and RKKY interactions lay a mechanistic foundation upon which a treasure map for hunting materials hosting skyrmions at low-field and high-temperature towards ambient conditions can be built (Figure 2). The layout of the treasure map is drawn from a combination of (i) the current stage of knowledge of skyrmions enabled by elegant work in the field, [17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35] [36, 37, 38, 39] and (ii) insights gained from our electronic structure and density of energy analysis. In this landscape, the realization of skyrmions at low-field and high-temperature toward ambient conditions can be conceptualized as an adventure through a multidimensional materials design space (Figure 2). Let’s get started. When you get to the first fork in the road, you can choose to take either the dd- or ff-electron pathway. The ff-electron path takes you to a set of skyrmion hosts, such as EuGa4_4, GdRu2_2Si2_2, and EuPtSi. These 4f4f magnetic metals, possessing large magnetic moments, geometrically frustrated lattices, and large exchange stiffness, display small (∼\sim a few nm) Bloch-type RKKY-driven skyrmions at low temperatures and high magnetic fields (>1> 1 T).40-43 One ought to cross the “1 Tesla river” on the far left in the treasure map (low-temperature region) to arrive at this set of 4f4f skyrmion hosts. By contrast, the dd-electron path takes you to more skyrmion host options. This rich behavior can be attributed to different dd-electron fillings, appreciable orbital overlap, and DM exchanges. As a result, skyrmion materials with dd electrons typically feature large skyrmion sizes (hundreds of nm) at lower fields (<1< 1 T) and higher temperatures.44-51 Thus, one can enjoy exploring the DM-driven skyrmion space without crossing the 1 Tesla river. When you take the dd-electron path and get to the next fork in the road, you can decide your next step based on dd-electron fillings and orbital occupancy, which fundamentally shape the band structure, magnetic anisotropy, and competing magnetic exchange interactions of the DM-enabled skyrmion materials. For materials with half-filled d5d^5 shells, e.g., MnSi and MnGe, skyrmions in the 3–20 nm range evolve at 0.12 and 2.4 T, 30 and 170 K, respectively. Cu2_2OSeO3_3 (d9d^9) stabilizes Bloch-type skyrmions (∼\sim60 nm) at 0.035 T and 59 K. Materials with less than half-filled (d1−4d^{1-4}) configurations, such as VOSe2_2O5_5, GaV4_4S8_8, GaMo4_4S8_8 yield large Néeltype skyrmions (∼140\sim140 nm) at 0.0030.003 T ≤μ0H≤0.1\le\mu_0 H \le0.1 T and 77 K ≤T≤18\le T \le18 K—lower fields than those of other aforementioned skyrmion materials but still at low temperatures. However, the temperature condition of skyrmion evolutions is approaching room temperature as one follows the path towards more-than-half-filled shells (d6d^6, d7d^7). 47, 52-55 A further distinction lies between metallic and insulating skyrmion hosts. Insulators, such as α\alpha-Fe2_2O3_3, BaFe12_{12}O19_{19}, 56-58 can stabilize skyrmions at relatively low fields and closer to room temperature (0.050.05 T ≤μ0H≤0.2\le\mu_0 H \le0.2 T and 240240 K ≤T≤270\le T \le270 K), while metals and semimetals typically can display skyrmions at higher temperatures. Examples include FeGe (Bloch-type, ∼70\sim70 nm), Fe3_3GaTe2_2 (Bloch-type, ∼100\sim100 nm), and Mn2_2Rh0.95_{0.95}Ir0.05_{0.05}Sn and Mn1.4_{1.4}Pt0.9_{0.9}Pd0.1_{0.1}Sn (anti-skyrmions >130>130 nm) host the topological nontrivial spin textures close to ambient conditions (0.0150.015 T ≤μ0H≤0.165\le\mu_0 H \le0.165 T and 276276 K ≤T≤400\le T \le400 K). 59-63 Crossing the 1 Tesla river on the far right in the treasure map takes one back to the RKKY land, but with the unique, complex 4qq helical skyrmions host—SrFeO3_3 at a high field (10 T). 37-39, 50 Taken together, all the realizations point to some connections between the DM/RKKY exchange and the topology, size, and formation condition of skyrmions.

Treasure map toward skyrmion evolutions at ambient conditions

Figure 2. A treasure map illustrating pathways toward the realization of skyrmions at low-field and high-temperature towards ambient conditions, guided by key design parameters: dd- or ff-electrons, orbital occupancy, metallic or insulating character, and DM or RKKY interactions. The highlighted route marks the parameter combination leading to the target ambient skyrmion (treasure).

The skyrmion landscape built upon prior elegant studies in the field enables us to have a framework in mind as to how far we have come. But where are we going from here? How can we get to the treasure—materials hosting skyrmions at ambient conditions? Is there any other hint or insight into electronic structures that we can connect to the critical field and temperature conditions at which skyrmions evolve? We attempt to answer these questions by looking into the electronic origins that give rise to the phase transition to skyrmions. 64-69 Drawing from the current set of known skyrmion hosts, we developed a unified framework that emphasizes the role of electronic instabilities, particularly features in density of energy (DOE) and spin density of energy (DOEs_s), as key descriptors of the conditions under which skyrmions emerge. Within this framework, we find that there is a common, distinctive feature in the DOEsDOE_s: a destabilizing peak at the Fermi level followed by a local minimum just above it. When normalized by unit-cell volume, these energy descriptors reveal some correlation with the critical field and temperature conditions for skyrmion evolution. This normalization can be justified since it captures the crystal structure and bonding of the skyrmion materials.

By tying the DOEDOE and DOEsDOE_s to the experimental conditions of skyrmion formation, this framework provides a powerful tool for predicting new skyrmion materials and designing skyrmions under ambient conditions. We validate this semi-quantitative model and discuss its application in collaboration with electron fillings, magnetic exchange interactions, and bonding character. We then propose how these design parameters can be integrated to realize materials hosting skyrmions towards ambient conditions in the treasure map.

Conceptual Framework and Semiquantitative Model

Formulation of the Density of Energy and Spin Density of Energy Approach

One emerging theme in recent studies is the role of electronic instabilities as a potential driving force for skyrmion formation.70,71 In particular, Fermi surface nesting, where large, parallel regions of the Fermi surface enhance susceptibility at specific wavevectors, has been identified as a key indicator of instability in several metallic skyrmion hosts. Examples include metallic systems such as EuAl4\mathrm{EuAl}_4, GdRu2Si2\mathrm{GdRu}_2\mathrm{Si}_2, and Gd2PdSi3\mathrm{Gd}_2\mathrm{PdSi}_3.72–75 In the GdRu2X2\mathrm{GdRu}_2\mathrm{X}_2 (X=Si\mathrm{X} = \mathrm{Si}, Ge\mathrm{Ge}) family, an increasing degree of nesting correlates with skyrmion formation at progressively lower fields and higher temperatures. This suggests that electronic instabilities promote a phase transition to skyrmions. However, many known insulating skyrmion hosts are left out by this approach, since they lack a well-defined Fermi surface or nesting conditions. This raises the need for a broader descriptor that captures electronic instabilities across both metallic and insulating systems.

A natural extension is to consider chemical bonding, particularly the population of antibonding states near or at the Fermi level (EFE_{\mathrm{F}}). If antibonding dominates near EFE_{\mathrm{F}}, the system may be structurally and electronically unstable, thereby predisposing it to symmetry-breaking transitions such as skyrmion formation. This can be probed using the crystal orbital Hamilton population (COHP) method.76–78 In this formalism, a negative sign in the projected COHP (−pCOHP-p\mathrm{COHP}) indicates antibonding (destabilizing) interactions—destructive interference of the overlap of the interacting atomic wavefunctions.79,80 With spin polarization (non-restricted spin), spontaneous magnetization causes the spin-up (majority-spin) electrons to go down in energy and the spin-down (minority-spin) electrons to go up in energy.81, 82, 83, 84, 85 This electronic symmetry-breaking is simply a result of bonding stabilization, alleviating electronic instabilities and stabilizing the system. While COHP is highly effective for identifying local, pairwise bonding interactions, it does not account for on-site atomic energy contributions. To complement COHP, we employed the concept of density of energy (DOE), which captures both interatomic and on-site atomic energy contributions as it integrates the entire electronic band structure with respect to energy.

The DOE is calculated separately for each spin channel (Table S1). The spin density of energy DOEs=DOEDn−DOEUp\mathrm{DOE}_{\mathrm{s}} = \mathrm{DOE}_{\mathrm{Dn}} - \mathrm{DOE}_{\mathrm{Up}} reveals the difference in energies between the two spin sublattices—the exchange splitting. The corresponding DOEs\mathrm{DOE}_{\mathrm{s}} integral (EspinE_{\mathrm{spin}}, Equation (1)) can provide a semiquantitative guide for the link between the electronic origins and the critical field and temperature of skyrmions formation.

Skyrmion Hostskk-mesh (a×b×ca \times b \times c)Charge Density Cutoff (Ry)Wavefunction Cutoff (Ry)Hubbard Parameter (eV)
Cu2_2OSeO3_34×4×44 \times4 \times449755Cu-3d 9.0 eV
FeGe3×3×33 \times3 \times341145N/A
GaMo4_4S8_84×4×44 \times4 \times430639Mo-4d 5.0 eV
MnGe5×5×55 \times5 \times522525Mn-3d 5.0 eV
MnSi5×5×55 \times5 \times525040Mn-3d 5.0 eV
VOSe2_2O5_52×2×32 \times2 \times364147N/A
EuAl4_46×6×36 \times6 \times31099122Eu-4f 5.7 eV
EuGa4_46×6×36 \times6 \times31099122Eu-4f 5.7 eV
EuGa2_2Al2_26×6×36 \times6 \times31099122Eu-4f 5.7 eV
EuPtSi4×4×44 \times4 \times41099122Eu-4f 4.5 eV
Gd3_3Ru4_4Al12_{12}6×6×46 \times6 \times41049116Gd-4f 6.7 eV
GdRu2_2Ge2_26×6×46 \times6 \times473445Gd-4f 6.7 eV
GdRu2_2Si2_26×6×46 \times6 \times473445Gd-4f 6.7 eV
SrFeO3_36×6×66 \times6 \times622525N/A

Table S1. DFT details: kk-mesh, Hubbard U, and Energy cutoff

Espin=∫−∞EF(DOEDn−DOEUp) dE(1)E_{\mathrm{spin}}=\int_{-\infty}^{E_{\mathrm{F}}}\left(\mathrm{DOE}_{\mathrm{Dn}}-\mathrm{DOE}_{\mathrm{Up}}\right)\,\mathrm{d}E \tag*{(1)}

This measure captures the net imbalance in the spin-up and spin-down channels, providing a key descriptor of electronic instabilities that can give rise to a magnetic phase transition. Within this framework, core electrons, structures, and other non-spin components contribute equally to both spin channels and cancel out upon subtraction, leaving only magnetism-specific contributions in EspinE_{\mathrm{spin}}. In addition, the set of material examples used in this analysis is skyrmion hosts; it means a pathway through which a magnetic phase transition to skyrmions is readily available. Since the magnetic contribution is primarily governed by valence electrons near EFE_{\mathrm{F}}, the energy window for integration is carefully selected based on the valence orbital character of the magnetic atom (e.g., dd- or ff-orbitals). This can be validated from the DOEs plot, where the DOEs should be localized to the energy range of the valence orbitals. A near-zero DOEs at lower energies confirms that non-magnetic states are properly excluded from the magnetic contribution.

Case Study I: GdRu2_2X2_2 (X = Si, Ge)

Plots of density of electronic states and magnetic properties for GdRu$_2$Si$_2$ and GdRu$_2$Ge$_2$

Figure 10.

In the case of GdRu₂X₂ (X = Si and Ge), the integrated DOE (EbandE_{\mathrm{band}}) is positive at EFE_F, indicating a destabilizing contribution and an electronic instability (Figure 3, S1, S2).36, 86–88 This instability can give rise to a magnetic phase transition to skyrmions. Since skyrmions emerge under finite temperature and field, the system must transition into an excited state. Thus, electronic states above EFE_F should give us some hint of skyrmion formation. It is tricky to isolate the magnetic contribution in the DOE curves (Figure 3a-Figure 3b) since there are strongly destabilizing energy contributions (positive DOE values and EbandE_{\mathrm{band}}). However, the DOEs_{\mathrm{s}} curves provide better clarity. The integrated DOEs_{\mathrm{s}} reveal changes in slope and sign at ∼4\sim4 eV, consistent with the anomaly in the DOE at the same energy. Note that the change in sign of the integrated DOEs_{\mathrm{s}} in Figure 3c-Figure 3d is only due to the mathematical operation (Equation (1)). The spin energy (EspinE_{\mathrm{spin}}, Equation (1)) can give us a hint of a spin instability that can induce a phase transition to skyrmions. The integrated DOEs_{\mathrm{s}} also reveals another unique feature—a local minimum right above 4 eV where its first derivative is zero. This local minimum is indicative of the energy required (EreqE_{\mathrm{req}}) to stabilize the system in an excited state. The finite field and temperature necessary for skyrmion formation may be linked to EreqE_{\mathrm{req}}. The system may exhibit its most pronounced magneto-entropy and topological Hall effect, which are hallmarks of the skyrmion phase, at the characteristic temperature TSmaxT_{\mathrm{Smax}} and magnetic field HSmaxH_{\mathrm{Smax}}—conditions related to EreqE_{\mathrm{req}}. GdRu₂Si₂ hosts skyrmions at Tf=38T_f = 38 K, Hf=0.5H_f = 0.5 T, and its highest entropy region is at TSmax=25T_{\mathrm{Smax}} = 25 K, HSmax=2H_{\mathrm{Smax}} = 2 T, while the Ge material displays skyrmions at Tf=25T_f = 25 K, Hf=1.9H_f = 1.9 T, and its highest entropy region is at TSmax=20T_{\mathrm{Smax}} = 20 K, HSmax=1.2H_{\mathrm{Smax}} = 1.2 T (lower temperature and field compared to the Si version.[35] This can be linked to GdRu2_2Ge2_2 possessing a lower EreqE_{\mathrm{req}} (5.23 eV) than GdRu2_2Si2_2 (5.45 eV).

Image not extracted

Figure 3. Density of energy analysis of (a) GdRu₂Si₂, (b) GdRu₂Ge₂; (c, d) Corresponding spin density of energy DOEs; (e) Field and temperature conditions for skyrmion formation (Inset showing the crystal structure)

Case Study II: EuAl4_4, EuAl2_2Ga2_2, and EuGa4_4

Now, let us have a look at the series of EuAl4_4, EuAl2_2Ga2_2, and EuGa4_4—isostructural centrosymmetric skyrmion hosts (Figure 4, S3-5). Both EuAl4_4 and EuAl2_2Ga2_2 display their largest magnetic entropy region at a similar field HSmax⁡≈1H_{S\max} \approx1 T, but at different temperatures TSmax⁡T_{S\max}.[28, 29, 31] Since EuAl4_4 possesses a higher EreqE_{\mathrm{req}} (8.94 eV) than EuAl2_2Ga2_2 (8.84 eV), more energy is required to transition the system to the excited state; thus, EuAl4_4 has a higher TSmax⁡T_{S\max}.

Spin density of energy DOEs and field and temperature conditions for skyrmion formation

Figure 4. Spin density of energy DOEs of (a) EuAl4_4, (b) EuAl2_2Ga2_2, (c) EuGa4_4; (d) Field and temperature conditions for skyrmion formation (Inset showing the crystal structure)

Similarly, EuAl2_2Ga2_2 and EuGa4_4 share a similar TSmax⁡≈2T_{S\max} \approx2 K, but require different fields.[30] EuGa4_4, which has a higher EreqE_{\mathrm{req}} (8.93 eV), requires a larger HSmax⁡H_{S\max} (6.5 T) for its skyrmion formation.

Case Study III: MnSi, MnGe and FeGe

In the first two case studies, we demonstrate that the semiquantitative model is effective in linking the spin density of energy to the skyrmion formation conditions in RKKY-driven systems. To further examine this model, we study skyrmion hosts facilitated by DM interactions. Let us focus on MnSi, MnGe, and FeGe, which crystallize in the noncentrosymmetric B20 crystal structure (Figure 5, S6-8). [23, 24] [89, 90, 91] In these DM-driven skyrmion materials, the DOEs and corresponding integrated DOEs curves are more diffused, consistent with the more directional and spatially diffused nature of dd-orbitals, as opposed to the more localized feature of the ff-electron RKKY systems discussed earlier. Despite these differences, the isostructural DM-driven skyrmion hosts show similar trends: a positive EspinE_{\mathrm{spin}} indicates a spin instability around EFE_{\mathrm{F}}; and a local minimum appears above EFE_{\mathrm{F}}, indicative of EreqE_{\mathrm{req}} required to stabilize the system in an excited state. Among the selected candidates, MnSi has the lowest Ereq=1.74E_{\mathrm{req}} = 1.74 eV, followed by MnGe (Ereq=3.42E_{\mathrm{req}} = 3.42 eV) and FeGe (Ereq=3.88E_{\mathrm{req}} = 3.88 eV). This agrees well with experimental observations that MnGe requires a higher magnetic field than MnSi, and FeGe requires a higher temperature than MnSi, to enter the skyrmion phase.

Spin density of energy DOEs of MnSi, MnGe, and FeGe, and field and temperature conditions for skyrmion formation

Figure 5. Spin density of energy DOEs of (a) MnSi, (b) MnGe, (c) FeGe; and (d) Field and temperature conditions for skyrmion formation (Inset showing the crystal structure)

Case Study IV: VOSe2O5\mathrm{VOSe}_{2}\mathrm{O}_{5}, GaMo4S8\mathrm{GaMo}_{4}\mathrm{S}_{8}, and Cu2OSeO3\mathrm{Cu}_{2}\mathrm{OSeO}_{3}

DM-driven skyrmion materials are not limited to metallic or semimetal systems; skyrmions can also emerge in insulating compounds when the magnetic sublattice adopts a polar or chiral symmetry. Examples include VOSe2O5\mathrm{VOSe}_{2}\mathrm{O}_{5}, GaMo4S8\mathrm{GaMo}_{4}\mathrm{S}_{8}, and Cu2OSeO3\mathrm{Cu}_{2}\mathrm{OSeO}_{3} (Figure 6, S9–11).[17, 18, 19, 26, 27] The DOEs and corresponding integrated DOEs curves for these materials show similar trends to those of the B20 compounds. One exception is Cu2OSeO3\mathrm{Cu}_{2}\mathrm{OSeO}_{3}, which does not exhibit a positive EspinE_{\mathrm{spin}} like the other skyrmion hosts. This can be attributed to the electronic configuration of Cu2+\mathrm{Cu}^{2+} (3d93d^{9}), where most 3d3d orbitals are filled and thus the dd-orbital states are deep down in energy -12 eV. (Figure S11) Nevertheless, changes in slope in the energy region between -2 eV and 3 eV are apparent, similar to the trend of the other skyrmion materials. However, it can be tricky to correlate the EreqE_{\mathrm{req}} values of VOSe2_2O5_5, GaMo4_4S8_8, and Cu2_2OSeO3_3 and their skyrmion formation conditions (Figure 6d, Figure S12-13). This challenge is due to these materials possessing different crystal structures. In order to make a more meaningful comparison, we attempt to figure out plausible ways to normalize EreqE_{\mathrm{req}} with respect to (i) magnetic atoms or (ii) a unit cell volume. Simply normalizing by the number of magnetic atoms is insufficient because it overemphasizes the magnetic ions' contribution to EreqE_{\mathrm{req}}, neglecting the role of nonmagnetic atoms that also participate in exchange pathways and influence collective electronic and magnetic contributions. A more justified physical approach is to normalize EreqE_{\mathrm{req}} by unit cell volume, since the relevant electronic and magnetic contributions scale with the spatial extent of the lattice. With this normalization, the EreqE_{\mathrm{req}} values become 0.022 eV/\AA3^3 for VOSe2_2O5_5, 0.054 eV/\AA3^3 for GaMo4_4S8_8, and 0.082 eV/\AA3^3 for Cu2_2OSeO3_3. This trend explains the experimental observation that VOSe2_2O5_5 requires the lowest field and temperature for skyrmion formation among the series.

Spin density of energy DOEs and crystal structures of skyrmion hosts

Figure 6. Spin density of energy DOEs_{\mathrm{s}} of (a) VOSe2O5\mathrm{VOSe}_{2}\mathrm{O}_{5}, (b) GaMo4S8\mathrm{GaMo}_{4}\mathrm{S}_{8}, (c) Cu2OSeO3\mathrm{Cu}_{2}\mathrm{OSeO}_{3}; (d) Field and temperature conditions for skyrmion formation; (e-g) Crystal structures of the skyrmion hosts.

Magnetic entropy of VOSe₂O₅

Figure S12. Magnetic entropy of VOSe2_2O5_5.

Generalization Across the Diverse Set of Known Skyrmion Materials

We then apply the normalization of EspinE_{\mathrm{spin}} and EreqE_{\mathrm{req}} with respect to a unit cell volume to a larger set of skyrmion hosts (Figure 7). The selected materials span a diverse range, including all DM- and RKKY-driven skyrmion hosts from metallic to insulating magnets with dd- and ff-electrons and various crystal structures. As shown in Figure 7a, although Espin/A˚3E_{\mathrm{spin}}/\AA^3 points to some hints of the evolution conditions of skyrmions, it is not sufficient to capture the trend for all the skyrmion hosts of investigation (Table S2). For example, the Espin/A˚3E_{\mathrm{spin}}/\AA^3 values of MnSi and MnGe are comparable, but the skyrmion phase in MnGe emerges at a higher field.

Three-dimensional plots showing correlations between characteristic temperature, magnetic field, and extracted electronic energy densities

Figure 7. Three-dimensional plot showing the correlation between characteristic temperature (TSmaxT_{\mathrm{Smax}}), magnetic field (HSmaxH_{\mathrm{Smax}}), and the extracted (a) Espin/A˚3E_{\mathrm{spin}}/\mathrm{\AA}^{3} and (b) Ereq/A˚3E_{\mathrm{req}}/\mathrm{\AA}^{3}. Orange diamonds indicate dd-electron skyrmion materials, while blue circles represent ff-electron skyrmion hosts.

Skyrmion HostsTmax⁡T_{\max} (K)Hmax⁡H_{\max} (T)Volume (Å3^3)EReqE_{\mathrm{Req}} (eV)EReqE_{\mathrm{Req}} per Å3^3EspinE_{\mathrm{spin}} per Å3^3Size (nm)Ref
Cu2_2OSeO3_32.70.075710.932.420.054-0.996216-18
FeGe2760.025103.823.880.1493.357019
GaMo4_4S8_8250.1230.084.690.0821.0520-22
MnGe202.4110.073.420.1241.913-623
MnSi280.1295.211.740.0732.021812, 24
VOSe2_2O5_57.60.003989.482.760.0220.4114025, 26
EuAl4_4101215.578.940.0830.8527, 28
EuGa4_426.5205.718.930.0871.2929
EuGa2_2Al2_221204.028.840.0871.1630
EuPtSi4.31266.599.40.1414.681.815, 31
Gd3_3Ru4_4Al12_{12}51.2641.305.830.0552.172.832
GdRu2_2Ge2_2201176.675.230.0591.3933
GdRu2_2Si2_2102.3166.735.470.0661.4134, 35
SrFeO3_31051057.114.540.0790.621.87-9

Table S2. Skyrmion hosts, formation conditions and calculated energy required and spin energy.

Now, let us turn to Ereq/A˚3E_{\mathrm{req}}/\mathrm{\AA}^{3}, which enables a more effective link to the skyrmion formation conditions. Figure 7b clearly reveals that systems with larger Ereq/A˚3E_{\mathrm{req}}/\mathrm{\AA}^{3} require higher temperatures and/or magnetic fields to transition to a skyrmion state.

From all the case studies and the overall trend of the known skyrmion hosts, we demonstrate the key role of electronic instabilities and density of energy in manifesting the temperature and field conditions of skyrmion evolution. By examining Ereq/A˚3E_{\mathrm{req}}/\mathrm{\AA}^{3}, we can directly compare different skyrmion materials and establish a trend between their electronic origins and their skyrmion formation conditions. While this realization is enticing, it is worthwhile assessing the semiquantitative model to see whether any potential pitfalls or improvements should be taken into account.

Assessment of the Semiquantitative Model

To evaluate the Ereq/A˚3E_{\mathrm{req}}/\AA^3 relationship with skyrmion formation conditions, we developed a simple theoretical model that combines phonons from the Debye model and magnetic energy contribution from applied fields (Equation (5)). This approach captures the proper energy scale for low-temperature, low-field skyrmion hosts, but tends to underestimate phonon contributions and overestimate field effects for high-temperature and high-field systems. (Figure S19)

Difference between the calculated external energy and the extracted excitation energy for 14 skyrmion hosts

Figure S19. Difference between the calculated external energy (EexE_{\mathrm{ex}}) using model 1 and the extracted excitation energy (EreqE_{\mathrm{req}}) for 14 skyrmion hosts.

To improve approximation, we constructed a second model based on experimental observables, including magneto-entropy and magnetization (Equation (6)). This data-driven formulation reproduces the required energy window (10−210^{-2}–10−110^{-1} eV/A˚3\AA^3) more reliably and points to a convergence between this second model and the applied, semiquantitative one (Figure S20).

Difference between the calculated external energy and extracted excitation energy for eight skyrmion hosts

Figure S20. Difference between the calculated external energy (EexE_{\rm ex}) using model 2 and the extracted excitation energy (EreqE_{\rm req}) for 8 skyrmion hosts.

Although the convergence lends some credibility to the semiquantitative model of linking Ereq/A˚3E_{\mathrm{req}}/\AA^3 with skyrmion formation conditions, some small differences remain due to the limitation on collinear ground-state electronic structure calculations. One wants to keep in mind that most skyrmion hosts exhibit noncollinear or helical spin structures at the ground state. More rigorous noncollinear supercell calculations could yield higher accuracy, though at high computational cost. Nevertheless, the presented model provides a practical framework for linking electronic instabilities and density of energy to the temperature and field conditions of skyrmion evolution and for guiding rapid materials screening using electronic structure calculations.

Materials Design Perspectives and Guiding Principles

Within this framework, we now look back at the DOEsDOE_s and corresponding DOEsDOE_s integral plots to gain some further insights. In addition to the case studies mentioned above, we add a few more model systems (Figure 8) that help build a more complete picture for our center question: how to design a material that hosts skyrmions close to or at ambient conditions—room temperature and low-to-zero field.

Spin density of energy DOEs for (a) FeGe, (b) Mn2RhSn, (c) Cu2OSeO3, (d) GaMo4S8, (e) EuPtSi, (f) Gd3Ru4Al12, (g) GdRu2Si2, and (h) SeFeO3

Figure 8. Spin density of energy DOEs\mathrm{DOE}_{s} of (a) FeGe, (b) Mn2_{2}RhSn, (c) Cu2_{2}OSeO3_{3}, (d) GaMo4_{4}S8_{8}, (e) EuPtSi, (f) Gd3_{3}Ru4_{4}Al12_{12}, (g) GdRu2_{2}Si2_{2}, and (h) SeFeO3_{3}. Orange color denotes dd-electron skyrmion hosts, while blue color signifies ff-electron skyrmion materials.

Distinctive Behaviors of dd- and ff-Electron Systems

As illustrated in Figure 8, dd-electron and ff-electron skyrmion materials exhibit significantly different features in DOEs\mathrm{DOE}_{s}. For dd-electron systems, the DOEs\mathrm{DOE}_{s} curves are more dispersed, consistent with the more effective overlap and diffused character of dd-orbitals. This trend is also reflected in Figure 7, where dd-electron compounds tend to stabilize skyrmions under relatively low magnetic fields and across a wide temperature range. In contrast, ff-electron systems generally require higher fields and lower temperatures to access their skyrmion phase.[33, 34] This difference arises partly from how magnetic moments respond to temperature and magnetic field perturbations. From magnetization measurements, where the moment MM is a function of both temperature and field, M(T,H)M(T, H), it is often observed that the magnetic moment varies significantly with field in M(H)M(H) curves, while remaining nearly constant across temperature in M(T)M(T) data. This suggests that magnetic fields more directly affect spin orientation through dipole interactions, while temperature influences spin behavior more indirectly, such as through entropy-driven fluctuations. In the context of skyrmions, this distinction is crucial: temperature supports skyrmion formation by providing entropic stabilization, while a magnetic field drives spin arrangement. As a result, skyrmions in dd-electron magnets are more likely to emerge under more accessible (low-field and high-temperature) or ambient conditions.

Electron Fillings of Valence Orbitals

In addition to valence electron characters, electron fillings of valence orbitals play a key role in determining whether a system can facilitate skyrmion formation. For instance, when the valence orbitals are less than half-filled—as in VOSe2O5\mathrm{VOSe}_{2}\mathrm{O}_{5} and GaMo4S8\mathrm{GaMo}_{4}\mathrm{S}_{8}—the dd-states lie mostly above EFE_{\mathrm{F}}, resulting in small electronic instabilities.[21, 22] On the other hand, almost all filled orbitals, as in Cu2OSeO3\mathrm{Cu}_{2}\mathrm{OSeO}_{3} (d9d^{9}), most dd-states lie well below EFE_{\mathrm{F}}, contributing little to electronic instabilities (Figure S11).

Density of energy for Cu₂OSeO₃ showing spin-resolved density of energy and spin asymmetry with its integration

Figure S11. Density of energy for Cu2_2OSeO3_3: (a) spin-resolved density of energy (DOE); (b) spin asymmetry (DOEs_s) and its integration.

Exactly half-filled dd- or ff-orbitals tell us a unique story. These systems have large magnetic spin moments, with quenched orbital contributions; thus, they interact more directly with external magnetic fields via dipole interactions, and the influence of temperature is reduced. The external magnetic field couples directly to the spin moment, making the response to magnetic fields more pronounced and predictable. This behavior is evident in Eu- and Gd-based (4f74f^{7}) systems, as well as in MnSi and MnGe (d5d^{5}). When switching from Si (3p3p) to Ge (4p4p) in MnGe, the EspinE_{\mathrm{spin}} remains largely unchanged, but the EreqE_{\mathrm{req}} increases. This suggests that MnGe requires a greater energy input, through the magnetic field, than MnSi to transition to its skyrmion phase. This field sensitivity is expected from systems with predominantly spin-only magnetic moments (half-filled orbitals). In contrast, when the valence orbitals are slightly more than half-filled or spin-orbit coupling (SOC) is sizable, the total magnetic moment is a combination of spin and orbital angular momentum contributions (J≠SJ \ne S). The resulting electronic states are defined by crystal-field splitting and closely spaced JJ-manifolds. These low-lying energy levels are sensitive to thermal excitations, and their population can change at finite temperature. In such systems, temperature plays a more important role in the state population.92 For example, when Mn in MnGe (d5d^{5}) is replaced by Fe in FeGe (d6d^{6}),93 the EspinE_{\mathrm{spin}} and EreqE_{\mathrm{req}} values of FeGe are higher than those of MnGe. This can explain why FeGe hosts skyrmions at higher temperatures than MnGe.

Influence of Electronic Structure in Metals and Insulators

Let us revisit Figure 2—the treasure map. Among the reported skyrmion materials, insulating systems have shown great promise for realizing the skyrmion phase at lower magnetic fields than metals. However, for utilities in spintronic applications, metallic materials offer increased tunability. In metals, skyrmions can be manipulated not only by magnetic fields but also by electric fields, enabling more refined control. Skyrmions in metallic systems may be stabilized by either DM or RKKY exchange. DM interactions arise in noncentrosymmetric lattices, while RKKY interactions require free charge carriers. Thus, metallic systems foster an additional driving force, mediated by conduction electrons, that can enhance or tune skyrmion formation. When targeting skyrmions at ambient conditions, the RKKY interaction alone is insufficient to provide the necessary driving force. In fact, some semi-metallic noncentrosymmetric magnets are able to stabilize DM-driven skyrmions close to room temperature at relatively low fields. Examples include FeGe (Bloch-type, stable up to 276 K at 150 Oe)59, Co8_8Zn9_9Mn3_3 (Bloch-type, 315 K at 1000 Oe)60, 61, Mn2_2Rh0.95_{0.95}Ir0.05_{0.05}Sn (antiskyrmion, 237 K at 350 Oe)62, and Mn1.4_{1.4}Pt0.9_{0.9}Pd0.1_{0.1}Sn (antiskyrmion, 400 K at 1650 Oe)63. These systems illustrate that skyrmions at ambient conditions are most likely to be realized in semi-metals with slightly more than half-filled dd-orbitals (d6d^6, d7d^7). This observation suggests that looking into magnetic semi-metals that feature similar electronic instabilities and density of energy to those of the skyrmion hosts can be a jumping-off point toward the realization of skyrmions under ambient conditions.

DM and RKKY Interactions

Within the aforementioned materials design perspectives, we have discussed the influence of dd- vs ff-electron systems, electron fillings in valence orbitals, and metals vs insulators. Now, let us return to the central Hamiltonians that govern the two underlying magnetic interactions responsible for skyrmion formation: the DM and the RKKY interaction. Understanding the nature of these Hamiltonians provides physical insight into why different skyrmion materials respond differently to external perturbations such as temperature and magnetic field. We constructed normalized magnetic phase diagrams corresponding to each interaction, enabling a direct visual comparison of the characteristic behavior of the two skyrmion classes (Figure 9).[20], 43, 48, 94-101

Normalized phase diagrams of DM- and RKKY-driven skyrmions

Figure 9. Normalized phase diagrams of DM- and RKKY-driven skyrmions. PM: paramagnetic; FP: field-polarized; FM: ferromagnetic; NC-AMF: non-collinear antiferromagnetic; SkL: skyrmions.

The DM interaction originates from spin–orbit coupling in a non-centrosymmetric crystal structure. Its Hamiltonian is given by:

HDMI=∑⟨i,j⟩Dij⋅(Si×Sj)(2)\mathcal{H}_{DMI} = \sum_{\langle i,j \rangle} \mathbf{D}_{ij} \cdot\left(\mathbf{S}_i \times\mathbf{S}_j\right) \tag*{(2)}

where Dij\mathbf{D}_{ij} is the DM vector, and Si\mathbf{S}_i, Sj\mathbf{S}_j are neighboring spins. This interaction favors non-collinear, chiral spin textures such as spin spirals and skyrmions. Figure 9a shows that as temperature decreases at zero field, a DM-driven skyrmion host goes through a paramagnetic state, then a field-polarized state and evolves into a helical ground state. While the DM interaction is necessary to stabilize these spin textures, its strength is typically small compared to the conventional Heisenberg exchange. In the presence of applied fields, the system also experiences the Zeeman effect. At modest fields, the Zeeman term induces spin alignment without disrupting the chiral winding, thereby stabilizing the conical spin configuration at low temperature or skyrmions at elevated temperature (Figure 9a). However, as the field increases beyond a critical threshold, the Zeeman energy overcomes the twisting of DM interaction, leading to a collapse of the skyrmion phase into a field-polarized state or a ferromagnetic state. It is worth noting that DM-driven skyrmions typically evolve at low fields across a broad range of temperatures up to 400 K (Figure 2, Figure 9a). Overall, the formation of DM-induced skyrmions can be described as a delicate balance among the Heisenberg exchange, the Zeeman effect, and the DM interaction, consistent with phase diagrams in which the skyrmion phase occupies only a narrow region between other conventional magnetic states.

In contrast, the RKKY interaction, which arises from indirect exchange coupling mediated by conduction electrons, is strongly dependent on the electronic structure. Its simplified form is:

HRKKY∝JRKKY(r)Si⋅Sj(3)\mathcal{H}_{RKKY} \propto J_{RKKY}(r)\mathbf{S}_i \cdot\mathbf{S}_j \tag*{(3)}
JRKKY(r)∼cos⁡(2kFr)r3(4)J_{RKKY}(r) \sim\frac{\cos(2k_F r)}{r^3} \tag*{(4)}

The coupling strength JRKKYJ_{RKKY} is an oscillatory function of the interatomic distance rr and the Fermi wavevector kFk_F. RKKY exchange—a first-order effect in exchange coupling—can exceed the magnitude of the DM interaction (a second-order effect). This difference in interaction strength is reflected in the distinct ground states of skyrmion materials (Figure 9). DM-driven hosts typically exhibit a helical ground state, in which neighboring spins rotate by less than 180° due to the influence of the DM vector. In contrast, some RKKY-driven hosts display AFM or incommensurate non-collinear AFM ground states, indicating that JRKKYJ_{RKKY} can be sufficiently strong to reverse a spin from its neighbor.

RKKY-driven skyrmion evolution is particularly susceptible to the topology of the Fermi surface. A vital feature that facilitates skyrmion formation in magnetic metals is Fermi surface nesting—large, parallel sections of the Fermi surface that lead to anomalies in the electronic susceptibility at certain wavevectors. However, this nesting is a low-temperature phenomenon; as temperature increases, the Fermi–Dirac distribution becomes broader, smearing out these features and reducing the effectiveness of the spin-mediated coupling. Thus, elevated temperatures can suppress the underlying mechanism that stabilizes RKKY-induced skyrmions. In fact, RKKY-driven skyrmions typically emerge at high fields and low temperatures (Figure 2). But once the skyrmions are stabilized, their presence extends from TcT_c down to base temperature (Figure 9b).

Taken together, the discussion on the underlying DM and RKKY interactions illuminates, in part, some insight into why DM-driven skyrmions tend to evolve over a wide temperature window at low fields, while RKKY-induced skyrmions typically form at low temperatures and high fields (Figure 2, Figure 9). This framework aligns well with the observations discussed in § 6.1 through 4.3. For instance, the different field and temperature conditions seen in dd- and ff-electron systems can now be better understood by considering both DOE and magnetic exchange interactions. This understanding could further guide design strategies for skyrmion hosts at ambient conditions by providing some hints of the electronic origins that foster a phase transition to skyrmions.

Implications for Skyrmion Discovery and Design Strategies

The perspective highlights the importance of tailoring the electronic structure, through careful magnetic element selection, strategic electron fillings of valence orbitals, and desired bonding character, as a guiding strategy for choosing skyrmion materials of interest. This framework offers a predictive pathway for evaluating and engineering systems that stabilize skyrmions under accessible field and temperature conditions.

One important factor not addressed in depth here is the size of skyrmions, which critically impacts their stability, dynamics, and suitability for device uses. It has been experimentally observed that DM-driven skyrmions often have larger sizes (a few to hundreds of nm) compared to those stabilized by RKKY-type interactions (a few nm). Moreover, another salient experimental observation is that skyrmion sizes are temperature-dependent. RKKY-driven skyrmions, typically evolving at higher magnetic fields (μ0Hc>1\mu_0 H_c > 1 T) and lower temperatures, display small sizes. DM-induced skyrmions emerging at higher temperatures are typically larger in size than those forming at low temperatures, suggesting that thermal excitations are at play. Although many excellent studies have modeled skyrmion sizes using micromagnetic simulations or continuum models, the microscopic origin of skyrmion sizes still remains to be further investigated.102-107

In this perspective, DOE and DOEs_s analysis is based on spin-polarized DFT that studies the ground-state electronic structure. While this approach provides valuable insights into electronic instabilities that can give rise to a skyrmion phase transition, it cannot resolve the spatial size or profile of skyrmions. As mentioned earlier, noncollinear DFT calculations using supercells that explicitly incorporate the skyrmion spin configuration would allow for direct comparison of total energies, electronic structure, and spin texture features, offering more quantitative guidance to link the electronic structure of the excited state and skyrmion sizes.

Nevertheless, this approach provides an effective framework for judiciously tuning the electronic structure of the material to promote susceptibility to skyrmion evolution. Let us take Mn2_2RhSn as an example. This pristine compound does not host skyrmions; however, its electron-doped systems Mn2_2Rh0.95_{0.95}Ir0.05_{0.05}Sn and Mn1.4_{1.4}Pt0.9_{0.9}Pd0.1_{0.1}Sn stabilize skyrmions at high temperatures. This occurs because electron doping positions the Fermi level into a region where electronic instabilities are amplified.108, 109 Another example that also benefits from the approach is the B20 compounds (e.g., FeSi, MnGe). Hole-doping these systems with Co or Cr, such as Fe1−xCoxSi\mathrm{Fe}_{1-x}\mathrm{Co}_x\mathrm{Si} and Cr0.82Mn0.18Ge\mathrm{Cr}_{0.82}\mathrm{Mn}_{0.18}\mathrm{Ge}, results in the evolution of skyrmions at a wide temperature range and low fields. The electronic tunability approach resonates with other excellent studies in the field.49, 110-114

Summary and Outlook

In this perspective, we have articulated a unified framework of the electronic origins underpinning skyrmion formation across a diverse set of materials. The material examples in this study are established skyrmion hosts; thus, they already possess a pathway that enables a magnetic phase transition to skyrmions. By examining EspinE_{\mathrm{spin}} and EreqE_{\mathrm{req}} derived from the DOE and DOEs\mathrm{DOE}_s and experimental data, a semi-quantitative model for linking the electronic instability and the temperature and field conditions of skyrmion evolution is developed. We demonstrate that this approach can be applied across RKKY- and DM-driven systems, spanning metallic and insulating hosts with dd- and ff-electron magnetism. Normalizing EspinE_{\mathrm{spin}} and EreqE_{\mathrm{req}} by a unit-cell volume enables a meaningful comparison for the diverse set of skyrmion materials. We then discuss broader design considerations, highlighting how orbital character, electron fillings, magnetic exchange, and structural factors influence a skyrmion phase transition. In addition, the approach can serve as a predictive tool: (i) doping can turn a non-skyrmion host into a skyrmion host by positioning the Fermi level into a regime in the DOE where electronic instabilities are enhanced (modifying EspinE_{\mathrm{spin}} and EreqE_{\mathrm{req}}), and (ii) for known skyrmion hosts, the DOEs\mathrm{DOE}_s curve can also be used as a guide to tune skyrmion formation conditions by modifying the electronic instability through chemical adjustment.

The normalized phase diagrams highlight distinct characteristics of skyrmion regions stabilized by DM or RKKY interactions. DM-driven skyrmions are confined to a narrow region between conventional magnetic states, whereas RKKY-induced skyrmions typically span a broad range, extending from the critical temperature down to the base temperature. Moreover, DM-driven skyrmions can evolve over a wide temperature window at low fields, while RKKY-driven skyrmions generally emerge only at high fields and low temperatures.

Taking all into consideration, we predict that noncentrosymmetric semi-metals with slightly above half-filled dd-orbitals would be excellent candidates for hosting skyrmions at ambient conditions (the treasure in Figure 2). The framework elucidates the electronic origins governing skyrmion evolution and establishes a useful link between these insights and experimentally measurable parameters—specifically, the temperature and field conditions of a skyrmion phase transition. This perspective shares a new way of approaching quantum materials research that could assist researchers in advancing the design and development of skyrmions at ambient conditions, while appreciating the current remarkable research efforts in the field that have built the foundation necessary for many more forthcoming, tantalizing breakthroughs now within reach.

ASSOCIATED CONTENT

Supporting Information.

AUTHOR INFORMATION

Corresponding Author

Email: [email protected]

Author Contributions

The manuscript was written through the contributions of all authors. All authors have given approval to the final version of the manuscript.

Funding Sources

This work was supported by the Arnold and Mabel Beckman Foundation grant 2023 BYI and the Camille and Henry Dreyfus Foundation grant 2025 Camille Dreyfus Teacher-Scholar award TC-25-071. The synthesis of some of the skyrmions was supported by the U.S. National Science Foundation grant NSF-DMR-CAREER-2338014.

Notes

Any additional relevant notes should be placed here.

ACKNOWLEDGMENT

We thank Athira Babu and Abigail Corales for their assistance in the synthesis of two skyrmion host materials.

Supporting Information-1

Supplementary Materials for

Experimental Section

Synthesis

Polycrystalline VOSe2O5\mathrm{VO}_{\mathrm{Se}_2}\mathrm{O}_5 was prepared via a conventional solid state reaction. VO2\mathrm{VO}_2 and SeO2\mathrm{SeO}_2 were weighed in stoichiometric amounts, thoroughly ground using an agate mortar and pestle, and pressed into pellets. The pellets were sealed under vacuum in quartz tube and annealed at 400 °C for 96 h, with 3 intermediate regrinding steps to enhance homogeneity. The resulting pale-green powder was confirmed single-phase by powder X-ray diffraction

For Cu2OSeO3\mathrm{Cu}_2\mathrm{OSeO}_3, stoichiometric CuO and a slight molar excess of SeO2\mathrm{SeO}_2 were similarly mixed, ground, and pelletized. The pellets were vacuum-sealed in quartz tube and heated at 600 °C for 72 h, including 3 intermittent grinding stage to ensure complete reaction. Phase-pure olive-green powders were verified by powder X-ray diffraction.

Magnetization

DC magnetization measurements on VOSe2O5\mathrm{VOSe}_2\mathrm{O}_5 and Cu2OSeO3\mathrm{Cu}_2\mathrm{OSeO}_3 powder samples were conducted using the Vibrating Sample Magnetometer option of the Quantum Design Magnetic Properties Measurement System (MPMS-3). The magnetic entropy change (ΔSmag\Delta S_{\mathrm{mag}}) was subsequently extracted from temperature-dependent magnetization data by evaluating dM/dTdM/dT across field sweeps and integrating using the Maxwell relation.

Density Functional Theory Calculation

Pseudo-potential DFT calculations were conducted using Quantum Espresso (QE) software package[1] in conjunction with the Generalized Gradient Approximation (GGA+UU)[2] for the exchange-correlation potential, specifically adopting the PBE parametrization[3]. For the projector-augmented wave (PAW) potentials, all elements were sourced from the PSlibrary v.1.0.0 dataset,[4] the kk-mesh employed and kinetic energy cutoffs for both charge density and wavefunctions were optimized for each Skyrmion hosts, and the computation details are summarized in Table S1. To elucidate electronic structure, the spin-resolved density of energy were computed using the LOBSTER software.[5, 6]

Density of energy for GdRu₂Si₂

Figure S1. Density of energy for GdRu₂Si₂: (a) spin-resolved density of energy (DOE); (b) spin asymmetry (DOEs_s) and its integration.

Density of energy for GdRu₂Ge₂

Figure S2. Density of energy for GdRu₂Ge₂: (a) spin-resolved density of energy (DOE); (b) spin asymmetry (DOEs_s) and its integration.

Density of energy plots for EuAl₄

Figure S3. Density of energy for EuAl₄: (a) spin-resolved density of energy (DOE); (b) spin asymmetry (DOEs_s) and its integration.

Density of energy plots for EuGa₂Al₂

Figure S4. Density of energy for EuGa₂Al₂: (a) spin-resolved density of energy (DOE); (b) spin asymmetry (DOEs_s) and its integration.

Density of energy plots for EuGa4

Figure S5. Density of energy for EuGa4: (a) spin-resolved density of energy (DOE); (b) spin asymmetry (DOEs_s) and its integration.

Density of energy plots for MnSi

Figure S6. Density of energy for MnSi (a) spin-resolved density of energy (DOE); (b) spin asymmetry (DOEs_s) and its integration.

Density of energy for MnGe: spin-resolved density of energy and spin asymmetry with integration

Figure S7. Density of energy for MnGe: (a) spin-resolved density of energy (DOE); (b) spin asymmetry (DOEs_s) and its integration.

Density of energy for FeGe: spin-resolved density of energy and spin asymmetry with integration

Figure S8. Density of energy for FeGe: (a) spin-resolved density of energy (DOE); (b) spin asymmetry (DOEs_s) and its integration.

Spin-resolved density of energy and spin asymmetry for VOSe₂O₅

Figure S9. Density of energy for VOSe2_2O5_5: (a) spin-resolved density of energy (DOE); (b) spin asymmetry (DOEs_s) and its integration.

Spin-resolved density of energy and spin asymmetry for GaMo₄S₈

Figure S10. Density of energy for GaMo4_4S8_8: (a) spin-resolved density of energy (DOE); (b) spin asymmetry (DOEs_s) and its integration.

Magnetic entropy of Cu2OSeO3

Figure S13. Magnetic entropy of Cu2OSeO3\mathrm{Cu}_{2}\mathrm{OSeO}_{3}.

Density of energy for Fe(IO3)3

Figure S14. Density of energy for Fe(IO3)3\mathrm{Fe}(\mathrm{IO}_{3})_{3}: (a) spin-resolved density of energy (DOE); (b) spin asymmetry (DOEs_{\mathrm{s}}) and its integration.

Spin-resolved density of energy and spin asymmetry for Mn₂RhSn

Figure S15. Density of energy for Mn2RhSn\mathrm{Mn}_{2}\mathrm{RhSn}: (a) spin-resolved density of energy (DOE); (b) spin asymmetry (DOEs_{\mathrm{s}}) and its integration.

Figure S16. Density of energy for EuPtSi\mathrm{EuPtSi}: (a) spin-resolved density of energy (DOE); (b) spin asymmetry (DOEs_{\mathrm{s}}) and its integration.

Spin-resolved density of energy and spin asymmetry for Gd$_3$Ru$_4$Al$_{12}$

Figure S17. Density of energy for Gd3_3Ru4_4Al12_{12}: (a) spin-resolved density of energy (DOE); (b) spin asymmetry (DOEs_s) and its integration.

Spin-resolved density of energy and spin asymmetry for SrFeO$_3$

Figure S18. Density of energy for SrFeO3_3: (a) spin-resolved density of energy (DOE); (b) spin asymmetry (DOEs_s) and its integration.

Semi-Quantitative Assessment and Experimental Correlation

Pure theoretical approach

If the skyrmion formation is actually closely related to the available destabilized excited states in the electronic structure, we would expect the energy provided by the field and temperature to be close to the energy required for the skyrmion formation. In order to cover more skyrmion candidates, we used a pure theoretical model of the external energy (EexE_{\mathrm{ex}}) that can be extended to most of the hosts:

EexM1=9NkBT4θD3+H2Vμ0μ(5)E_{\mathrm{ex}}^{\mathrm{M1}} = \frac{9Nk_{\mathrm{B}}T^{4}}{\theta_{\mathrm{D}}^{3}} + \frac{H^{2}V}{\mu_{0}\mu} \tag*{(5)}

The first term describes the thermal energy input arising from phonon excitations. Here, NN is the number of atoms per A˚3\mathring{\mathrm{A}}^{3}. One may intend to use the number of magnetic atoms, but the EreqE_{\mathrm{req}} is extracted from spin-asymmetry DOEs; this characteristic value still represents the material instability change. kBk_{\mathrm{B}} is Boltzmann’s constant, TT is the absolute temperature, and θD\theta_{\mathrm{D}} is the Debye temperature, which sets the energy scale of lattice vibrations. The θD\theta_{\mathrm{D}} were fitted and fixed at 135 K for all Skyrmion hosts, which is far from accurate but sufficient for a rough validation. Within the low-temperature limit (T≪θDT \ll\theta_{\mathrm{D}}), this T4T^{4}-dependence stems from the Debye model and quantifies the contribution of thermal fluctuations to destabilizing the ground-state electronic and magnetic order. The second term accounts for the magnetic energy supplied by an external field HH, where VV is the unit-cell volume, μ0\mu_{0} is the vacuum permeability, and μ\mu is the magnetic permeability of the material. This term corresponds to the magnetic work that can overcome exchange interactions and drive the system into an excited magnetic state, such as a skyrmion phase. A fitted μfitted=0.26\mu_{\mathrm{fitted}} = 0.26 H/m was used for all hosts, again, far from accurate, just to approach the magnitude. Together, these two contributions describe the total energy available to promote the formation of skyrmions under finite temperature and magnetic field.

The calculated external energy was then compared to the extracted EreqE_{\mathrm{req}}. (Figure S19) Given that EreqE_{\mathrm{req}} lies in the order of 10−210^{-2} to 10−110^{-1} eV/Å3^{3}, this model successfully captures the required energy scale for several systems. However, it clearly overestimates the magnetic-field contribution, as hosts requiring HSmax>1H_{\mathrm{Smax}} > 1 T (e.g. EuGa4_{4}, SrFeO3_{3}, and MnGe)[7, 8, 9] exhibit energy scales far exceeding EreqE_{\mathrm{req}}. (Figure 7b) In addition, the thermal contribution is underestimated for hosts with TSmax>11T_{\mathrm{Smax}} > 11 K, such as MnSi, FeGe, and GaMo4_{4}S8_{8}. These discrepancies may be improved with more accurate values of θD\theta_{\mathrm{D}} and μ\mu, but the agreement between calculated EexE_{\mathrm{ex}} and EreqE_{\mathrm{req}} for low-field and low-temperature systems supports the idea that further destabilization of states above EFE_{\mathrm{F}} is closely linked to skyrmion formation.

Experimental data supported approach

Since the theoretical approach indicates the need for accurate data, we then propose the second model which is data-based for the external energy supplied to the system:

EexM2=TΔSmagmcellMVcell+12MHV(6)E_{\rm ex}^{M2} = \frac{T \Delta S_{\rm mag} m_{\rm cell}}{M V_{\rm cell}} + \frac{1}{2} M H V \tag*{(6)}

The first term represents the thermal energy contribution derived from the magneto-entropy ΔSmag\Delta S_{\rm mag}, scaled by temperature TT, magnetic moment per unit cell mcellm_{\rm cell}, molar mass MM, and unit-cell volume VcellV_{\rm cell}. This quantifies the thermal energy needed to reach the magnetically excited state observed in magneto-entropy analysis from magnetization of heat capacity measurements. The second term corresponds to the magnetic energy supplied by the external field HH, where MM is the magnetization per unit volume. Unlike the previous, more theoretical model, this formulation is grounded in direct experimental parameters (e.g. ΔSmag\Delta S_{\mathrm{mag}}, MM, VV, HH) and thus provides a practical means to estimate the energy scale driving skyrmion formation.

The calculated external energy was also compared with EreqE_{\mathrm{req}} (Figure S20), and this second model reproduces the required energy scale (10−210^{-2} to 10−110^{-1} eV/A˚3\AA^{3}) much more accurately.

Although the limited availability of data restricts the number of compounds considered, the model successfully captures the behavior of EuAl4, MnSi, and FeGe. This further supports the idea that the destabilization of states above EFE_{\mathrm{F}} plays a central role in skyrmion formation.

Proposed approach

Despite the improvements in the second model, we still observe a ∼\sim30% or greater difference between Ereq/A˚3E_{\mathrm{req}}/\AA^{3} and Eex/A˚3E_{\mathrm{ex}}/\AA^{3}. This discrepancy may be partly due to the assumptions made in the ground-state calculation. Specifically, we relied on collinear spin-polarized DFT and performed a spin-resolved DOE analysis under this constraint. However, several of the compounds considered do not exhibit a classical collinear magnetic ground state. For instance, EuAl4 adopts a collinear AFM order, but its magnetic structure is incommensurate with the lattice,[10] and other examples such as MnSi, FeGe, EuPtSi, and Cu2OSeO3 are known to host helical magnetic ground states.[11, 12, 13, 14, 15] These departures from conventional FM or AFM ordering may themselves originate from the electronic instability near EFE_{\mathrm{F}}. Furthermore, in the excited states under finite temperature or field, the electronic structure may be substantially altered relative to the ground-state configuration. Hence, extracting EreqE_{\mathrm{req}} solely from the ground-state band structure above EFE_{\mathrm{F}} is likely to be an oversimplification.

A more rigorous approach would involve noncollinear DFT calculations with supercells that reflect both the true magnetic structure of the ground-state and the skyrmion (excited-state) configuration. The normalized total energy difference between these two states would then provide a more accurate estimate of EreqE_{\mathrm{req}}. However, such calculations are computationally expensive, especially if we aim to apply this methodology to large materials databases in search of new skyrmion candidates. Nevertheless, the present model already captures the key trends and provides a reasonable starting point for rapid screening.

References

  1. [1]Giannozzi, P.; Baroni, S.; Bonini, N.; Calandra, M.; Car, R.; Cavazzoni, C.; Ceresoli, D.; Chiarotti, G. L.; Cococcioni, M.; Dabo, I. QUANTUM ESPRESSO: a modular and open-source software project for quantum simulations of materials. Journal of physics: Condensed matter 2009, 21 (39), 395502.DOI
  2. [2]Wang, L.; Maxisch, T.; Ceder, G. Oxidation energies of transition metal oxides within the GGA+ U framework. Physical Review B 2006, 73 (19), 195107.
  3. [3]Perdew, J. P.; Ruzsinszky, A.; Csonka, G. I.; Vydrov, O. A.; Scuseria, G. E.; Constantin, L. A.; Zhou, X.; Burke, K. Restoring the density-gradient expansion for exchange in solids and surfaces. Physical review letters 2008, 100 (13), 136406.pubmed.ncbi.nlm.nih.gov/18517979
  4. [4]Dal Corso, A. Pseudopotentials periodic table: From H to Pu. Computational Materials Science 2014, 95, 337-350.DOI
  5. [5]Maintz, S.; Deringer, V. L.; Tchougréeff, A. L.; Dronskowski, R. LOBSTER: A tool to extract chemical bonding from plane - wave based DFT. 2016.
  6. [6]Nelson, R.; Ertural, C.; George, J.; Deringer, V. L.; Hautier, G.; Dronskowski, R. LOBSTER: Local orbital projections, atomic charges, and chemical - bonding analysis from projector - augmented - wave - based density - functional theory. Journal of Computational Chemistry 2020, 41 (21), 1931-1940.DOI
  7. [7]Chakraverty, S.; Matsuda, T.; Wadati, H.; Okamoto, J.; Yamasaki, Y.; Nakao, H.; Murakami, Y.; Ishiwata, S.; Kawasaki, M.; Taguchi, Y. Multiple Helimagnetic Phases and Topological Hall Effect in Epitaxial Thin Films of Pristine and Co-Doped SrFeO₃. Physical Review B—Condensed Matter and Materials Physics 2013, 88 (22), 220405.
  8. [8]Ishiwata, S.; Tokunaga, M.; Kaneko, Y.; Okuyama, D.; Tokunaga, Y.; Wakimoto, S.; Kakurai, K.; Arima, T.; Taguchi, Y.; Tokura, Y. Versatile Helimagnetic Phases Under Magnetic Fields in Cubic Perovskite SrFeO₃. Physical Review B—Condensed Matter and Materials Physics 2011, 84 (5), 054427.DOI
  9. [9]Ishiwata, S.; Nakajima, T.; Kim, J.-H.; Inosov, D.; Kanazawa, N.; White, J.; Gavilano, J.; Georgii, R.; Seemann, K.; Brandl, G. Emergent Topological Spin Structures in the Centrosymmetric Cubic Perovskite SrFeO₃. Physical Review B 2020, 101 (13), 134406.
  10. [10]Takagi, R.; Matsuyama, N.; UkleeV, V.; Yu, L.; White, J. S.; Francoual, S.; Mardegan, J. R.; Hayami, S.; Saito, H.; Kaneko, K. Square and Rhombic Lattices of Magnetic Skyrmions in a Centrosymmetric Binary Compound. Nature communications 2022, 13 (1), 1472.DOI
  11. [11]Franke, K. J.; Dean, P. R.; Hatnean, M. C.; Birch, M. T.; Khalyavin, D.; Manuel, P.; Lancaster, T.; Balakrishnan, G.; Hatton, P. D. Investigating the Magnetic Ground State of the Skyrmion Host Material Cu₂OSeO₃ Using Long-Wavelength Neutron Diffraction. AIP Advances 2019, 9 (12).arxiv.org/abs/1807.11333
  12. [12]Karhu, E.; Kahwaji, S.; Robertson, M.; Fritzsche, H.; Kirby, B. J.; Majkrzak, C. F.; Monchesky, T. Helical Magnetic Order in MnSi Thin Films. Physical Review B—Condensed Matter and Materials Physics 2011, 84 (6), 060404.DOI
  13. [13]Kakihana, M.; Aoki, D.; Nakamura, A.; Honda, F.; Nakashima, M.; Amako, Y.; Takeuchi, T.; Harima, H.; Hedo, M.; Nakama, T. Unique Magnetic Phases in the Skyrmion Lattice and Fermi Surface Properties in Cubic Chiral Antiferromagnet EuPtSi. Journal of the Physical Society of Japan 2019, 88 (9), 094705.DOI
  14. [14]Felcher, G.; Jorgensen, J. Magnetic Structures of Monoclinic FeGe. Journal of Physics C: Solid State Physics 1983, 16 (32), 6281.DOI
  15. [15]Sakakibara, T.; Nakamura, S.; Kittaka, S.; Kakihana, M.; Hedo, M.; Nakama, T.; Ōnuki, Y. Magnetic Phase Transitions of the 4f Skyrmion Compound EuPtSi Studied by Magnetization Measurements. Journal of the Physical Society of Japan 2021, 90 (6), 064701.DOI
  16. [16]White, J. S.; Levatić, I.; Omrani, A.; Egetenmeyer, N.; Prša, K.; Živković, I.; Gavилano, J. L.; Kohlbrecher, J.; Bartkowiak, M.; Berger, H. Electric Field Control of the Skyrmion Lattice in Cu₂OSeO₃. Journal of Physics: Condensed Matter 2012, 24 (43), 432201.DOI
  17. [17]Qian, F.; Bannenberg, L. J.; Wilhelm, H.; Chaboussant, G.; Debeer-Schmitt, L. M.; Schmidt, M. P.; Aqeel, A.; Palstra, T. T.; Brück, E.; Lefering, A. J. New Magnetic Phase of the Chiral Skyrmion Material Cu₂OSeO₃. Science Advances 2018, 4 (9), eaat7323.DOI
  18. [18]Janson, O.; Rousochatzakis, I.; Tsirlin, A. A.; Belesi, M.; Leonov, A. A.; Rößler, U. K.; Van Den Brink, J.; Rosner, H. The Quantum Nature of Skyrmions and Half-Skyrmions in Cu₂OSeO₃. Nature communications 2014, 5 (1), 5376.DOI
  19. [19]Bocarsly, J. D.; Need, R. F.; Seshadri, R.; Wilson, S. D. Magnetoentropic Signatures of Skyrmionic Phase Behavior in FeGe. Physical Review B 2018, 97 (10), 100404.DOI
  20. [20]Schueller, E. C.; Kitchaev, D. A.; Zuo, J. L.; Bocarsly, J. D.; Cooley, J. A.; Van der Ven, A.; Wilson, S. D.; Seshadri, R. Structural Evolution and Skyrmionic Phase Diagram of the Lacunar Spinel GaMo₄Se₈. Physical Review Materials 2020, 4 (6), 064402.
  21. [21]Nikolaev, S.; Solovyev, I. Skyrmionic Order and Magnetically Induced Polarization Change in Lacunar Spinel Compounds GaV₄S₈ and GaMo₄S₈: Comparative Theoretical Study. Physical Review B 2020, 102 (1), 014414.DOI
  22. [22]Pocha, R.; Johrendt, D.; Pöttgen, R. Electronic and Structural Instabilities in GaV₄S₈ and GaMo₄S₈. Chemistry of materials 2000, 12 (10), 2882-2887.DOI
  23. [23]Tanigaki, T.; Shibata, K.; Kanazawa, N.; Yu, X.; Onose, Y.; Park, H. S.; Shindo, D.; Tokura, Y. Real-Space Observation of Short-Period Cubic Lattice of Skyrmions in MnGe. Nano letters 2015, 15 (8), 5438-5442.DOI
  24. [24]Ge, M.; Zhang, L.; Menzel, D.; Han, H.; Jin, C.; Zhang, C.; Pi, L.; Zhang, Y. Scaling Investigation of the Magnetic Entropy Change in Helimagnet MnSi. Journal of Alloys and Compounds 2015, 649, 46-49.DOI
  25. [25]Kurumaji, T.; Nakajima, T.; Feoktystov, A.; Babcock, E.; Salhi, Z.; Ukleev, V.; Arima, T.-h.; Kakurai, K.; Tokura, Y. Direct Observation of Cycloidal Spin Modulation and Field-Induced Transition in Néel-Type Skyrmion-Hosting VOSe₂O₅. Journal of the Physical Society of Japan 2021, 90 (2), 024705.DOI
  26. [26]Kurumaji, T.; Nakajima, T.; Ukleev, V.; Feoktystov, A.; Arima, T.-h.; Kakurai, K.; Tokura, Y. Néel-Type Skyrmion Lattice in the Tetragonal Polar Magnet VOSe₂O₅. Physical review letters 2017, 119 (23), 237201.DOI
  27. [27]Meier, W. R.; Torres, J. R.; Hermann, R. P.; Zhao, J.; Lavina, B.; Sales, B. C.; May, A. F. Thermodynamic Insights Into the Intricate Magnetic Phase Diagram of EuAl₄. Physical Review B 2022, 106 (9), 094421.DOI
  28. [28]Wang, S.; Luo, Q.; Luo, X.; Ma, S.; Chen, C.; Gao, F.; Lv, B.; Zhong, R.; Ren, W.; Zhong, Z. Tuning the Magnetism and Magnetocaloric Effects of EuAl₄ Single Crystals by Si and Cu Doping. Journal of Alloys and Compounds 2023, 968, 171977.DOI
  29. [29]Zhang, H.; Zhu, X.; Xu, Y.; Gawryluk, D. J.; Xie, W.; Ju, S.; Shi, M.; Shiroka, T.; Zhan, Q.; Pomjakushina, E. Giant Magnetoresistance and Topological Hall Effect in the EuGa₄ Antiferromagnet. Journal of Physics: Condensed Matter 2021, 34 (3), 034005.
  30. [30]Moya, J. M.; Lei, S.; Clements, E. M.; Kengle, C. S.; Sun, S.; Allen, K.; Li, Q.; Peng, Y.; Husain, A. A.; Mitrano, M. Incommensurate Magnetic Orders and Topological Hall Effect in the Square-Net Centrosymmetric EuGa₂Al₂ System. Physical Review Materials 2022, 6 (7), 074201.DOI
  31. [31]Mishra, A. K.; Ganesan, V. A-Phase, Field-Induced Tricritical Point, and Universal Magnetocaloric Scaling in EuPtSi. Physical Review B 2019, 100 (12), 125113.DOI
  32. [32]Ogunbunmi, M. O.; Nair, H. S.; Strydom, A. M. Magnetic Frustration-Driven Ground State Properties of Rare-Earth Magnetic Ions on a Breathing Kagome Lattice: A Review of the Gd₃Ru₄Al₁₂ Structure Type Magnets. Critical Reviews in Solid State and Materials Sciences 2023, 48 (4), 480-501.DOI
  33. [33]Yoshimochi, H.; Takagi, R.; Ju, J.; Khanh, N.; Saito, H.; Sagayama, H.; Nakao, H.; Itoh, S.; Tokura, Y.; Arima, T.-h. Multistep Topological Transitions Among Meron and Skyrmion Crystals in a Centrosymmetric Magnet. Nature Physics 2024, 20 (6), 1001-1008.
  34. [34]Gries, L.; Kleinbeck, T.; Mayoh, D.; Wood, G.; Balakrishnan, G.; Klingeler, R. Uniaxial Pressure Effects, Phase Diagram, and Tricritical Point in the Centrosymmetric Skyrmion Lattice Magnet GdRu₂Si₂. Physical Review B 2025, 111 (6), 064419.DOI
  35. [35]Yasui, Y.; Butler, C. J.; Khanh, N. D.; Hayami, S.; Nomoto, T.; Hanaguri, T.; Motome, Y.; Arita, R.; Arima, T.-h.; Tokura, Y. Imaging the Coupling Between Itinerant Electrons and Localised Moments in the Centrosymmetric Skyrmion Magnet GdRu₂Si₂. Nature communications 2020, 11 (1), 5925.DOI

Paper details

vlnc:W_gs394jrj
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 3

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents