betaFeedback

WorksPhysics and Astronomy

Chiral-Angle-Controlled Spin Splitting and Spin Transport in Nanotubes Rolled from d-wave Altermagnets

Abstract

Altermagnets combine compensated collinear magnetic order with momentum-dependent spin splitting in the electronic band structure. Here, we show that rolling a two-dimensional (2D) dd-wave altermagnet into a nanotube converts this momentum dependence into chiral-angle-controlled one-dimensional (1D) spin splitting through dimensional projection. A minimal tight-binding model reveals a characteristic nodal–antinodal dependence on the chiral angle θθ, with the central circumferential subband exhibiting a cos⁡(2θ)\cos(2θ) scaling and the projected spin splitting vanishing for the nodal orientation and reversing sign between orthogonal antinodal orientations. First-principles calculations for V2_2O and representative symmetric and Janus systems demonstrate that this nodal–antinodal selection rule persists despite curvature-induced structural asymmetry and magnetic moment imbalance. We further show that the projected electronic structure produces chiral-angle-controlled spin-polarized transport: antinodal nanotubes exhibit spin-polarized transmission, whereas the nodal nanotube remains conducting with identical spin-channel transmission. These results demonstrate how dimensional projection can translate the momentum-space spin splitting of a 2D altermagnet into geometrically controlled electronic and transport properties in nanotubes.

Altermagnets have recently emerged as a distinct class of collinear magnetic materials that exhibit momentum-dependent spin splitting despite compensated magnetic order [1–6]. Unlike conventional ferromagnets, where exchange-induced spin splitting is approximately uniform in momentum space, altermagnetic spin splitting changes sign between symmetry-related regions of the Brillouin zone and can form highly anisotropic dd-, gg-, or higher-order wave patterns [7, 8]. The coexistence of sizable spin splitting with compensated magnetic order has stimulated considerable interest in altermagnets as platforms for spin-dependent electronic phenomena with strongly reduced macroscopic stray fields [9–11]. Momentum-dependent spin splitting has now been observed experimentally in several candidate materials, including MnTe, CrSb, and KV2_2Se2_2O, through angle-resolved photoemission spectroscopy and complementary probes [12–14]. These developments have been accompanied by rapid progress in symmetry classification, materials discovery, and first-principles studies of bulk and low-dimensional altermagnets, including an expanding family of two-dimensional (2D) systems [15–24].

A defining feature of many altermagnetic states is the presence of symmetry-protected momentum-space directions along which the spin splitting vanishes. In square-lattice 2D dd-wave altermagnets, this structure takes a particularly transparent form: a dx2−y2d_{x^2-y^2}-type spin splitting produces antinodal directions with large and opposite-sign splitting, separated by nodal directions where the bands remain spin degenerate. This nodal–antinodal structure raises a more general question concerning dimensional reduction. When a higher-dimensional electronic structure is confined to one dimension, transverse momentum becomes quantized while dispersion remains along the one-dimensional propagation direction. Different geometrical orientations can therefore sample distinct sectors of the original momentum-dependent electronic structure. This suggests that dimensional projection could select between nodal and antinodal sectors of a dd-wave altermagnet and thereby convert momentum-space spin anisotropy into a geometrically controlled one-dimensional spin splitting.

Nanotubes provide a natural setting in which to realize this principle. Rolling a 2D layer into a cylinder imposes periodic boundary conditions around the circumference, quantizing the transverse momentum and producing a set of one-dimensional (1D) subbands dispersing along the nanotube axis. In carbon nanotubes, the wrapping geometry determines the allowed momentum-space cuts through the graphene band structure and thereby whether the resulting nanotube is metallic or semiconducting [26] [25]. Related dimensional-confinement effects arise more broadly in nanotubes formed from layered materials, for which substantial experimental progress has been achieved in transition-metal dichalcogenides and Janus systems [27, 28, 29, 30, 31, 32]. Recent theoretical work has also explored altermagnetic nanotubes in the context of Lifshitz-transition-induced open Fermi contours and chirality-dependent persistent spin currents [33]. Together with the growing set of predicted 2D dd-wave altermagnets [34, 35, 36] [16], these developments establish altermagnetic nanotubes as a promising setting for geometrically controlled spin-dependent phenomena. Here, we focus instead on how the sign-changing momentum-dependent spin splitting of the parent dd-wave state is transferred to the one-dimensional electronic structure through dimensional projection. The nanotube geometry can thereby select different sectors of the nodal–antinodal structure, transforming momentum-space magnetic anisotropy into a real-space geometrical control parameter.

Whether this projected spin splitting also controls transport is a separate question. A spin-split band structure does not by itself imply a spin-polarized current, because transport depends on the propagating states available near the Fermi level. Establishing whether the nodal–antinodal structure survives at the level of the conducting channels is therefore essential for connecting dimensional projection to spintronic functionality. Nanotubes are particularly appealing in this respect because their geometry is structurally fixed and can, in principle, control the relative contributions of the two spin channels without changing the underlying magnetic order. Whether such geometrical control can be realized in realistic nanotubes rolled from 2D dd-wave altermagnets remains unexplored.

Here, we establish dimensional projection as a mechanism for converting the momentum-dependent spin splitting of a 2D dd-wave altermagnet into geometrically controlled 1D electronic and transport properties. A minimal tight-binding model reveals a characteristic nodal–antinodal dependence on the chiral angle θ\theta, with the central circumferential subband exhibiting a cos⁡(2θ)\cos(2\theta) scaling of the projected spin splitting. We then verify the resulting nodal–antinodal selection rule using first-principles calculations for V2_2O and structurally distinct symmetric and Janus nanotubes, showing that it persists despite curvature-induced structural asymmetry and magnetic moment imbalance. Finally, we demonstrate that this geometrical dependence is reflected in spin-polarized transport, with the spin-channel imbalance progressively suppressed toward the nodal orientation while finite charge transmission is retained.

Results

Dimensional projection of dd-wave spin splitting

We first establish the general mechanism by which the momentum-dependent spin splitting of a 2D dd-wave altermagnet is converted into angle-dependent one-dimensional electronic states upon rolling. To separate the underlying geometrical and symmetry principles from material-specific effects, we consider a minimal two-sublattice tight-binding model on a square lattice describing a generic metallic dd-wave altermagnet. The model provides an analytical framework for determining how the orientation of a nanotube axis relative to the nodal–antinodal structure of the parent system controls the projected 1D spin splitting. Model parameters and further numerical details are provided in the Supplementary Information. The Bloch Hamiltonian is

Hσ(k)=ε(k)τ0+tAB(k)τx+[Jσ+λgd(k)]τz,(1)H_{\sigma}(\mathbf{k}) = \varepsilon(\mathbf{k})\tau_{0} + t_{AB}(\mathbf{k})\tau_{x} + [J\sigma+ \lambda g_{d}(\mathbf{k})]\tau_{z}, \tag*{(1)}

where σ=±1\sigma= \pm1 denotes the spin index, τi\tau_i are Pauli matrices acting in the magnetic sublattice space, ε(k)\varepsilon(\mathbf{k}) describes the spin-independent dispersion, tAB(k)t_{AB}(\mathbf{k}) is the intersublattice hopping, JJ is the staggered exchange field, and λ\lambda controls the strength of the dd-wave term. The characteristic momentum dependence of the dd-wave state is encoded in the form factor

gd(k)=cos⁡kx−cos⁡ky,(2)g_{d}(\mathbf{k}) = \cos k_x - \cos k_y, \tag*{(2)}

which changes sign under a π/2\pi/2 rotation and vanishes along kx=±kyk_x = \pm k_y. The resulting spin splitting therefore has two orthogonal antinodal directions with opposite signs, separated by nodal lines along which the bands remain spin degenerate. This dx2−y2d_{x^2-y^2} structure is illustrated by the parent 2D band structure and momentum-resolved spin splitting in Figure 1(a) and Figure 1(b), respectively.

Figure 1. Dimensional projection of dd-wave spin splitting into one-dimensional nanotube states. (a) Band structure of the parent 2D dd-wave altermagnet along X–Γ–Y. (b) Momentum-resolved spin splitting ΔE(k)\Delta E(k) of the parent system, showing antinodal regions with opposite signs of the spin splitting separated by symmetry-protected nodal lines. (c)–(e) Spin-resolved 1D band structures for nanotubes with chiral angles θ=0∘\theta= 0^\circ, 45∘45^\circ, and 90∘90^\circ, respectively. These three limiting geometries are achiral. Antinodal orientations produce finite spin splitting with opposite signs for θ=0∘\theta= 0^\circ and 90∘90^\circ, whereas the nodal orientation at θ=45∘\theta= 45^\circ yields spin-degenerate bands. (f) Spin-resolved energies of the k⊥=0k_\perp= 0 subband at the nanotube zone-boundary point ZZ as a function of chiral angle θ\theta. The splitting evolves continuously between the two antinodal orientations, vanishes at the nodal orientation, and reverses sign, following the cos⁡(2θ)\cos(2\theta) scaling of the projected dd-wave form factor for this circumferential mode.

Rolling the 2D system into a nanotube imposes periodic boundary conditions around the circumference and restricts the parent electronic structure to a discrete set of one-dimensional modes. We define the chiral angle θ\theta by the orientation of the nanotube axis relative to an antinodal direction of the parent layer and introduce momentum components (k∥,k⊥)(k_{\parallel}, k_{\perp}) parallel and perpendicular to the nanotube axis, respectively. The circumferential component k⊥k_{\perp} is quantized, whereas k∥k_{\parallel} remains the continuous crystal momentum along the nanotube axis, following the standard nanotube zone-folding construction [26] [25]. Changing θ\theta therefore changes which sectors of the parent nodal–antinodal structure contribute to the 1D dispersion.

This geometrical dependence follows directly from the transformation of the dd-wave form factor. Expanding Eq. (2) around the Γ\Gamma point and expressing the momentum in the rotated nanotube coordinates gives

gd∝cos⁡(2θ)(k∥2−k⊥2)−2sin⁡(2θ)k∥k⊥.(3)g_{d} \propto\cos(2\theta)(k_{\parallel}^{2} - k_{\perp}^{2}) - 2\sin(2\theta)k_{\parallel}k_{\perp}. \tag*{(3)}

For the k⊥=0k_{\perp} = 0 circumferential mode, the projected spin splitting therefore scales as

Δtube(θ)∝cos⁡(2θ)k∥2.(4)\Delta_{\mathrm{tube}}(\theta) \propto\cos(2\theta) k_{\parallel}^{2}. \tag*{(4)}

Equation (4) provides a simple analytical description of the nodal–antinodal angular dependence for this projected subband: the spin splitting is maximal in magnitude for the antinodal orientations θ=0∘\theta= 0^\circ and 90∘90^\circ, vanishes for the nodal orientation θ=45∘\theta= 45^\circ, and reverses sign between the two orthogonal antinodal directions. For the remaining quantized circumferential modes with k⊥≠0k_{\perp} \ne0, additional terms contribute, so that the spin splitting does not in general follow a universal cos⁡(2θ)\cos(2\theta) dependence.

The corresponding nanotube band structures are shown in Figure 1(c)–Figure 1(e). The three representative angles θ=0∘\theta= 0^\circ, 45∘45^\circ, and 90∘90^\circ correspond to achiral nanotube geometries, while intermediate values of θ\theta generally describe chiral nanotubes. For θ=0∘\theta= 0^\circ, projection along an antinodal direction produces pronounced spin splitting of the 1D bands [Figure 1(c)]. At θ=45∘\theta= 45^\circ, the nanotube axis is aligned with a nodal direction and the bands become spin degenerate [Figure 1(d)]. Rotating further to the orthogonal antinodal orientation at θ=90∘\theta= 90^\circ restores finite spin splitting with the opposite sign [Figure 1(e)]. These three achiral orientations thus sample the positive, zero, and negative sectors of the parent dd-wave spin-splitting pattern.

Dimensional reduction also folds states from different momenta of the parent 2D Brillouin zone into the reduced 1D Brillouin zone of the nanotube. Consequently, spin-split states originating at finite momenta of the parent system can appear at the nanotube zone center, even though the primitive-cell Γ\Gamma point of the parent 2D altermagnet is spin degenerate by symmetry. This distinction between dimensional folding and curvature-induced modifications will become particularly useful below when the nanotube bands are compared with those of corresponding rectangular 2D supercells.

The continuous evolution between the limiting achiral orientations is shown in Figure 1(f), where the spin-resolved energies at the nanotube zone-boundary point ZZ are plotted as a function of the chiral angle θ\theta. For the k⊥=0k_{\perp} = 0 subband, the spin splitting evolves continuously from one antinodal orientation to the other, vanishes at θ=45∘\theta= 45^\circ, and reverses sign across the nodal direction, consistent with Eq. (4). The chiral angle therefore converts the momentum-space anisotropy of the parent dd-wave state into a geometrically tunable property of the 1D electronic structure.

First-principles realization in V2_2O nanotubes

We next examine how the dimensional-projection mechanism identified in the tight-binding model is realized in an atomistic system. As a minimal first-principles platform, we consider checkerboard V2_2O, whose square-lattice magnetic structure exhibits the dx2−y2d_{x^2-y^2} spin splitting required for the nodal–antinodal projection mechanism. Figure 2(a) shows the parent 2D V2_2O layer together with the high-symmetry crystallographic directions [10], [11], and [01]. The [10] and [01] directions are antinodal, whereas the diagonal [11] direction is nodal.

Atomistic realization of nanotubes rolled from a 2D $d$-wave altermagnet

Figure 2. Atomistic realization of nanotubes rolled from a 2D dd-wave altermagnet. (a) Checkerboard 2D V2OV_2O with opposite-spin V sublattices and high-symmetry crystallographic directions [10], [11], and [01]. (b) Representative achiral V2OV_2O nanotubes with chiral angles θ=0∘\theta= 0^\circ, 45∘45^\circ, and 90∘90^\circ, defined by the orientation of the nanotube axis relative to an antinodal direction of the parent layer. The θ=0∘\theta= 0^\circ and 90∘90^\circ geometries correspond to antinodal orientations of the parent dd-wave spin splitting, whereas θ=45∘\theta= 45^\circ corresponds to the nodal orientation.

Using the chiral-angle convention introduced above, the representative nanotubes in Figure 2(b) correspond to θ=0∘\theta= 0^\circ, 45∘45^\circ, and 90∘90^\circ. These three limiting geometries are achiral and provide atomistic realizations of the two orthogonal antinodal orientations and the intervening nodal orientation identified in the tight-binding model.

The corresponding first-principles electronic structure of the parent V2OV_2O monolayer is shown in Figure 3(a). Along the X–Γ\Gamma–Y path, the bands exhibit pronounced momentum-dependent spin splitting away from the Brillouin-zone center, while the Γ\Gamma point remains spin degenerate due to the symmetry of the parent checkerboard lattice. The splitting reverses sign between the orthogonal X–Γ\Gamma and Γ\Gamma–Y directions, displaying the characteristic dx2−y2d_{x^2-y^2} structure anticipated by the tight-binding model.

Spin-resolved band structures of 2D V$_2$O, a rectangular V$_2$O supercell, and V$_2$O nanotubes

Figure 3. First-principles dimensional projection of the V2_2O electronic structure. (a) Spin-resolved band structure of primitive-cell 2D V2_2O along X–Γ–Y. (b) Band structure of a 1 × 18 rectangular V2_2O supercell along Γ–X, showing the folded electronic structure prior to rolling. (c)–(e) Spin-resolved band structures of the achiral V2_2O nanotubes with chiral angles θ=0∘\theta= 0^\circ, 45∘45^\circ, and 90∘90^\circ, respectively, calculated along the nanotube axis (Γ–Z). The two antinodal nanotubes exhibit finite spin splitting with opposite signs, whereas the nodal nanotube remains spin degenerate.

To separate the effects of dimensional folding from those introduced by curvature, we first consider a 1×181 \times18 rectangular supercell of the 2D monolayer containing the same number of atoms as the corresponding θ=0∘\theta= 0^\circ nanotube unit cell. Its band structure along Γ\Gamma–X, corresponding to the direction that becomes the nanotube axis, is shown in Figure 3(b). The enlarged real-space periodicity folds finite-momentum states of the primitive 2D system into the reduced supercell Brillouin zone. Consequently, spin-split states can appear at the supercell zone center even though the primitive-cell Γ\Gamma point is spin degenerate. The rectangular supercell therefore provides a curvature-free reference for identifying the electronic changes associated with subsequent rolling.

The nanotube band structures for θ=0∘\theta= 0^\circ, 45∘45^\circ, and 90∘90^\circ are shown in Figure 3(c)–Figure 3(e), respectively. The two antinodal nanotubes retain pronounced spin splitting along the 1D Γ\Gamma–Z direction, with opposite signs for the two orthogonal orientations. In contrast, the θ=45∘\theta= 45^\circ nanotube remains spin degenerate, consistent with alignment of the nanotube axis along a nodal direction of the parent dd-wave state. These first-principles band structures therefore reproduce the nodal–antinodal dependence predicted by the tight-binding model.

The correspondence between the rectangular-supercell and nanotube band structures further shows that the characteristic spin splitting is already encoded in the di-

mensionally folded electronic structure rather than being generated by curvature. Rolling modifies this electronic structure through curvature and structural relaxation, but preserves the underlying nodal–antinodal dependence. The relation between the supercell and nanotube bands is further illustrated in Fig. S1 of the Supplementary Information.

The V2OV_2O nanotubes also remain nearly magnetically compensated after rolling. For a representative radius of R=11.04R = 11.04 Å, the calculated magnetic moments on the two V sublattices are +2.63 μB+2.63\ \mu_B and −2.61 μB-2.61\ \mu_B, corresponding to a residual moment imbalance ΔM=∣M1+M2∣≃0.02 μB\Delta M = \lvert M_1 + M_2 \rvert\simeq0.02\ \mu_B, while the O moments remain negligible. The persistence of the chiral-angle-dependent spin splitting in the presence of this small imbalance shows that the projected electronic structure is robust against a slight departure from exact magnetic compensation.

Chiral-angle-controlled spin transport

Having established that the chiral angle controls the projected spin splitting of the nanotube bands, we next examine how this geometrical dependence is reflected in electronic transport. We calculate the spin-resolved ballistic transmission along the nanotube axis for V2OV_2O nanotubes with chiral angles θ=0∘\theta= 0^\circ, ≃18∘\simeq18^\circ, ≃27∘\simeq27^\circ, and 45∘45^\circ. The θ=0∘\theta= 0^\circ and 45∘45^\circ nanotubes represent the achiral antinodal and nodal limits, respectively, whereas the intermediate θ≃18∘\theta\simeq18^\circ and ≃27∘\simeq27^\circ nanotubes are chiral. These structures therefore allow us to track how the spin polarization of the conducting channels evolves from the antinodal to the nodal orientation as the projected dd-wave spin splitting is progressively suppressed.

Figure 4(a)–Figure 4(d) show the zero-bias total and spin-resolved ballistic transmission spectra. For the antinodal θ=0∘\theta= 0^\circ nanotube, the two spin channels exhibit a pronounced imbalance around the Fermi level, with T↑(EF)=52T_{\uparrow}(E_F) = 52 and T↓(EF)=23T_{\downarrow}(E_F) = 23, giving a total transmission of T(EF)=75T(E_F) = 75. For θ≃18∘\theta\simeq18^\circ and ≃27∘\simeq27^\circ, the corresponding spin-resolved values are (46, 28) and (42, 28), with total transmissions of 74 and 70, respectively. At the nodal orientation, θ=45∘\theta= 45^\circ, the two spin channels become equivalent, T↑(EF)=T↓(EF)=35T_{\uparrow}(E_F) = T_{\downarrow}(E_F) = 35, while the total transmission remains finite at T(EF)=70T(E_F) = 70. Thus, the spin-channel imbalance is progressively suppressed on approaching the nodal direction, whereas the overall number of conducting channels remains comparable.

Chiral-angle-controlled spin transport in V₂O nanotubes

Figure 4. Chiral-angle-controlled spin transport in V₂O nanotubes. (a)–(d) Zero-bias spin-resolved ballistic transmission spectra for nanotubes with chiral angles θ=0∘\theta= 0^\circ, ≃18∘\simeq18^\circ, ≃27∘\simeq27^\circ, and 45∘45^\circ, respectively. The θ=0∘\theta= 0^\circ and 45∘45^\circ nanotubes are achiral, whereas the intermediate-angle nanotubes are chiral. The total transmission and the spin-up and spin-down contributions are shown as functions of energy, with the equilibrium Fermi level set to zero. The transmission polarization decreases on approaching the nodal direction and vanishes for the θ=45∘\theta= 45^\circ nanotube. (e)–(h) Energy-resolved spectral currents evaluated from the corresponding zero-bias transmission spectra with electrochemical potentials EF,L=−0.2 eVE_{F,L} = -0.2\ \mathrm{eV} and EF,R=+0.2 eVE_{F,R} = +0.2\ \mathrm{eV} relative to the equilibrium Fermi level, defining a 0.4 eV0.4\ \mathrm{eV} energy window. The charge and spin spectral currents are defined as Ich=I↑+I↓I_{\mathrm{ch}} = I_{\uparrow} + I_{\downarrow} and Isp=I↑−I↓I_{\mathrm{sp}} = I_{\uparrow} - I_{\downarrow}, respectively. The charge spectral current remains of comparable magnitude for the different chiral angles, whereas the spin spectral current decreases toward the nodal orientation and vanishes at θ=45∘\theta= 45^\circ.

A convenient measure of the spin polarization of the transmission is

PT(E)=T↑(E)−T↓(E)T↑(E)+T↓(E).(5)P_T(E) = \frac{T_{\uparrow}(E) - T_{\downarrow}(E)}{T_{\uparrow}(E) + T_{\downarrow}(E)}. \tag*{(5)}

At the Fermi level, PTP_T decreases from approximately 38.7% for θ=0∘\theta= 0^\circ to 24.3% and 20.0% for θ≃18∘\theta\simeq18^\circ and ≃27∘\simeq27^\circ, respectively, and vanishes for θ=45∘\theta= 45^\circ. The transport response therefore preserves the nodal–antinodal trend identified in the electronic structure: antinodal projections support a finite transmission polarization, whereas the nodal projection restores equal contributions from the two spin channels.

The transmission polarization should not be expected to follow the simple cos⁡(2θ)\cos(2\theta) dependence obtained for the k⊥=0k_{\perp} = 0 subband. This scaling follows from the projected dd-wave form factor for this particular circumferential mode, whereas the transmission at a given energy generally receives contributions from multiple propagating subbands and depends on their dispersions and crossings with that energy. Moreover, commensurate nanotubes with different chiral angles generally differ slightly in radius and unit-cell geometry, so the small variation of the total transmission from 70 to 75 cannot be attributed uniquely to the chiral angle. The robust transport signature is instead the progressive reduction of the spin-channel imbalance, from T↑−T↓=29T_{\uparrow} - T_{\downarrow} = 29 at θ=0∘\theta= 0^\circ to zero at θ=45∘\theta= 45^\circ, while finite charge transmission is retained.

To examine the corresponding current response, Figure 4(e)–Figure 4(h) show the energy-resolved spectral currents evaluated from the zero-bias transmission spectra. The left and right electrochemical potentials were set to EF,L=−0.2E_{F,L} = -0.2 eV and EF,R=+0.2E_{F,R} = +0.2 eV relative to the equilibrium Fermi level, respectively, corresponding to a separation of 0.4 eV. The spin-resolved spectral current is

Iσ(E)=eℏTσ(E)[fR(E)−fL(E)],(6)I_{\sigma}(E) = \frac{e}{\hbar}T_{\sigma}(E)\left[f_R(E) - f_L(E)\right], \tag*{(6)}

where fLf_L and fRf_R are the Fermi–Dirac distributions associated with the left and right electrochemical potentials. Because Tσ(E)T_{\sigma}(E) is taken from the zero-bias calculation and is not recalculated self-consistently for the imposed electrochemical-potential difference, these spectral currents represent a fixed zero-bias transmission approximation rather than a self-consistent finite-bias transport calculation. We define the charge and spin spectral currents as

Ich(E)=I↑(E)+I↓(E),Isp(E)=I↑(E)−I↓(E).(7)I_{\mathrm{ch}}(E) = I_{\uparrow}(E) + I_{\downarrow}(E), \qquad I_{\mathrm{sp}}(E) = I_{\uparrow}(E) - I_{\downarrow}(E). \tag*{(7)}

Consistent with the transmission spectra, the charge spectral current remains of comparable magnitude for the different chiral angles, whereas the spin spectral current decreases toward the nodal orientation and vanishes at θ=45∘\theta= 45^\circ, where the two spin-resolved transmissions coincide.

The spectral-current results therefore connect the chiral-angle-controlled electronic structure directly to spin-polarized transport. Antinodal nanotubes support a finite spin spectral current, whereas the nodal nanotube remains conducting but exhibits no spin spectral current within the considered electrochemical-potential window. Thus, the nodal–antinodal structure inherited from the parent 2D dd-wave altermagnet governs not only the projected 1D spin splitting but also the spin polarization of the propagating channels.

Robustness across material realizations

The V₂O nanotubes considered above provide a minimal realization of the dimensional-projection mechanism, but an important question is whether the same behavior persists in chemically and structurally more complex systems. Rolling generally makes the inner and outer sides of a nanotube inequivalent and can modify local bonding, hybridization, and magnetic moments. Such effects are expected to become particularly important in low-symmetry and Janus materials, where structural asymmetry is already present in the parent monolayer. We therefore test the robustness of the chiral-angle-controlled spin splitting in representative symmetric and Janus nanotubes with different degrees of curvature-induced magnetic imbalance.

The broader materials landscape provides several possible parent systems for such realizations. Layered V–O–chalcogen compounds, including KV₂Se₂O and RbV₂Te₂O, provide experimentally realized systems with closely related magnetic and structural motifs [37, 38] [14], while recent computational searches have identified a wider range of predicted 2D altermagnetic candidates [39, 40]. Representative parent monolayers are listed in Table S1 of the Supplementary Information. These materials should be regarded as a candidate space rather than as validated nanotube systems, since the stability and electronic structure of each rolled geometry must be established explicitly.

To test the mechanism beyond V₂O, we performed first-principles calculations for V₂OSe₂, V₂OSeTe, and Fe₂SSe. V₂OSe₂ provides a symmetric reference, V₂OSeTe introduces intrinsic Janus asymmetry, and Fe₂SSe represents a more strongly perturbed Janus case in which rolling produces a substantially larger magnetic moment imbalance. For each material, three representative nanotube radii were considered to assess the effects of curvature on the magnetic and electronic structure.

The calculated magnetic moments are summarized in tbl:S2 of the Supplementary Information. For V2OSe2\mathrm{V}_{2}\mathrm{OSe}_{2}, the residual moment imbalance ΔM\Delta M decreases from 0.15 μB0.15\ \mu_B at R≃5R \simeq5 Å to 0.04 μB0.04\ \mu_B at R≃10R \simeq10 Å and 0.02 μB0.02\ \mu_B at R≃15R \simeq15 Å. The Janus V2OSeTe\mathrm{V}_{2}\mathrm{OSeTe} nanotubes exhibit larger imbalances of 0.35 μB0.35\ \mu_B, 0.15 μB0.15\ \mu_B, and 0.10 μB0.10\ \mu_B over the same approximate radius range. An even stronger imbalance is found for Fe2SSe\mathrm{Fe}_{2}\mathrm{SSe}, where ΔM\Delta M reaches 0.69 μB0.69\ \mu_B for the smallest nanotube and decreases to 0.42 μB0.42\ \mu_B and 0.26 μB0.26\ \mu_B at radii of approximately 1010 and 1515 Å, respectively. The systematic decrease of ΔM\Delta M with increasing radius is consistent with the weakening of curvature-induced inequivalence between the inner and outer sides of the nanotube.

Despite these substantial differences in magnetic imbalance, the characteristic nodal–antinodal dependence of the electronic structure remains preserved. fig:S2 of the Supplementary Information compares the parent monolayer, rectangular-supercell, and nanotube band structures for representative V2OSeTe\mathrm{V}_{2}\mathrm{OSeTe} and Fe2SSe\mathrm{Fe}_{2}\mathrm{SSe} systems. For V2OSeTe\mathrm{V}_{2}\mathrm{OSeTe}, the θ=0∘\theta= 0^\circ nanotube retains pronounced spin splitting and closely follows the folded electronic structure of the corresponding rectangular supercell, whereas the θ=45∘\theta= 45^\circ nanotube remains spin degenerate along the nanotube axis. The same qualitative distinction persists in Fe2SSe\mathrm{Fe}_{2}\mathrm{SSe} despite the much larger curvature-induced magnetic imbalance.

The Fe2SSe\mathrm{Fe}_{2}\mathrm{SSe} bands also show that stronger curvature-induced structural and magnetic modifications can lead to quantitative deviations from the simple folded-supercell picture. These deviations, however, do not eliminate the geometrical selection rule in the calculated cases: antinodal orientations retain finite spin splitting, whereas the nodal orientation remains spin degenerate. The dimensional-projection mechanism is therefore not contingent on exact magnetic compensation or on an electronically rigid rolling process. Rather, curvature-induced changes act as perturbations to the nodal–antinodal projection imposed by the momentum dependence of the parent dd-wave spin splitting.

Overall, the V2O\mathrm{V}_{2}\mathrm{O}, V2OSe2\mathrm{V}_{2}\mathrm{OSe}_{2}, V2OSeTe\mathrm{V}_{2}\mathrm{OSeTe}, and Fe2SSe\mathrm{Fe}_{2}\mathrm{SSe} calculations show that the projected nodal–antinodal spin-splitting pattern persists across structurally distinct nanotube realizations and over a broad range of curvature-induced magnetic imbalance. These results establish dimensional projection as a robust geometrical mechanism for transferring the momentum-dependent spin splitting of 2D dd-wave altermagnets into controllable one-dimensional electronic states.

Discussion

An important consequence of dimensional reduction concerns the relation between the symmetry of the parent 2D altermagnet and that of the resulting nanotube. In the checkerboard 2D system, the combined C4TC_4T symmetry relates the two magnetic sublattices and constrains the characteristic sign-changing dd-wave spin-splitting pattern between orthogonal momentum-space directions. Rolling the layer into a nanotube generally breaks this global C4TC_4T symmetry, and the resulting 1D structure should therefore not, in general, be regarded as an altermagnet in the same symmetry-classification sense as its 2D parent. Nevertheless, the local magnetic environments and exchange interactions responsible for the parent dd-wave spin splitting remain largely preserved upon rolling. Dimensional projection then selects particular sectors of this inherited electronic structure, allowing the nanotube bands to retain an altermagnet-like spin splitting whose magnitude and sign depend on the chiral angle. The nanotube therefore preserves a characteristic electronic fingerprint of the parent altermagnet even though the corresponding global symmetry relation of the 2D parent is no longer present in the rolled geometry.

This distinction also clarifies the role of curvature and magnetic compensation. Rolling makes the inner and outer sides of the nanotube inequivalent, modifying local bonding and magnetic moments and producing a finite magnetic moment imbalance, particularly in the Janus systems. These effects perturb, rather than generate, the chiral-angle dependence, which is already present in the curvature-free dimensional-projection picture. The persistence of the nodal–antinodal distinction across the calculated nanotubes therefore shows that neither retention of the parent’s global C4TC_4T symmetry nor exact magnetic compensation of the final nanotube is required for the projected spin-splitting pattern to survive.

Beyond the electronic structure, the transport calculations establish that the inherited nodal–antinodal character also governs spin-polarized conduction. A spin-split band structure does not by itself determine the spin polarization of a current, because transport is governed by the propagating states available at a particular energy. Our results show that this nodal–antinodal information remains encoded in the propagating channels, providing chiral-angle control over both the projected 1D spin splitting and the spin polarization of the conducting states. This geometrical control suggests potential spintronic applications, as nanotubes with different chiral angles can exhibit finite or vanishing spin polarization and could provide active elements in spin filters, spin valves, or magnetic tunnel junctions. Particularly attractive are compensated or nearly compensated nanotubes, where pronounced spin-polarized transport can coexist with a vanishing or very small net magnetic moment.

These prospective functionalities should, however, be viewed within the scope of the present calculations. The behavior demonstrated here is confirmed for several structurally distinct nanotube realizations, whereas the broader set of 2D altermagnets identified in the Supplementary Information should be regarded as a candidate space rather than as validated nanotube systems. The stability, magnetic configuration, and electronic structure of each rolled material must therefore be established individually. Likewise, the present transport calculations describe zero-bias ballistic transmission and spectral currents evaluated within a fixed zero-bias transmission approximation, rather than a self-consistent finite-bias transport calculation. Contacts, disorder, defects, finite-temperature broadening, and scattering may modify the resulting spin polarization in realistic devices, and explicit device-level calculations will be required to assess the proposed spintronic functionalities.

More broadly, our results illustrate how dimensional projection can transfer characteristic electronic signatures of a higher-dimensional magnetic state to a reduced-dimensional system even when the global symmetry classification of the parent is not retained. In the present case, the momentum-dependent dd-wave spin splitting of a 2D altermagnet becomes a geometrically tunable property of a 1D nanotube, with the chiral angle controlling both the projected spin splitting and the spin polarization of the conducting channels. Similar principles may apply to nanoscrolls, curved ribbons, and other quasi-one-dimensional structures formed from anisotropic magnetic layers, providing a route for translating symmetry-constrained momentum-space electronic structure into geometrically controlled low-dimensional electronic and transport properties.

Methods

Tight-binding calculations

The tight-binding calculations were performed using the minimal two-sublattice model introduced in the main text. Nanotube states were obtained by rotating the momentum coordinates according to the chiral angle θ\theta and imposing periodic boundary conditions along the circumferential direction, resulting in quantized transverse momenta. The model parameters and further numerical details are provided in the Supplementary Information.

First-principles calculations

First-principles calculations were performed within density-functional theory using the generalized gradient approximation in the Perdew–Burke–Ernzerhof (PBE) parametrization for the exchange–correlation functional. All calculations were carried out using the QuantumATK package with norm-conserving PseudoDojo pseudopotentials and localized atomic-orbital basis sets [41, 42, 43]. Collinear magnetic calculations were performed without spin–orbit coupling. Atomic structures were relaxed until the residual forces were below 0.01 eV/Å. A vacuum separation larger than 20 Å was used between periodic images along the non-periodic directions. Brillouin-zone integrations for the 2D monolayers employed 20×20×120 \times20 \times1 Monkhorst–Pack meshes, while the nanotubes were sampled using at least 20 kk points along the one-dimensional Brillouin zone. Nanotubes were constructed from commensurate supercells of the corresponding parent monolayers with their axes aligned along the desired crystallographic directions. Further computational and structural details are provided in the Supplementary Information.

Spin-dependent transport

Spin-resolved ballistic transmission spectra were calculated using QuantumATK for the periodic bulk configurations of the nanotubes, with the nanotube axis chosen as the transport direction. No explicit device or scattering-region geometry was introduced. The resulting transmission Tσ(E)T_{\sigma}(E) describes the number of propagating channels of spin σ\sigma available at a given energy along the nanotube axis. All transmission spectra were calculated at zero bias.

Energy-resolved spectral currents were subsequently obtained from the zero-bias transmission spectra by setting the left and right electrochemical potentials to −0.2-0.2 and +0.2+0.2 eV relative to the equilibrium Fermi level, respectively, corresponding to an electrochemical-potential difference of 0.4 eV. Since the transmission functions were not recalculated self-consistently at finite bias, the spectral currents represent a fixed zero-bias transmission approximation rather than a self-consistent finite-bias transport calculation.

Data availability

The data supporting the findings of this study, including the numerical data underlying the electronic-structure and transport results, are available from the corresponding author upon reasonable request.

Code availability

The first-principles and transport calculations were performed using the commercially available QuantumATK software. The tight-binding model and all parameters required to reproduce the calculations are provided in the Supplementary Information. The custom codes used to implement the tight-binding calculations are available from the corresponding author upon reasonable request.

Acknowledgements

This work was supported by the Deutsche Forschungsgemeinschaft (DFG) through CRC/TRR 227, Project No. B12 (Project ID 328545488), and Grant No. LO 1659/10-1, and by the European Innovation Council (EIC) Pathfinder Open project, Grant No. 101129641. T.G.S. acknowledges computing time provided through the Gauss Centre for Supercomputing e.V. on the supercomputer JUWELS at Forschungszentrum Jülich under project superorb.

Author contributions

E.Ş. conceived the study and performed the first-principles and transport calculations. T.G.S. and B.G. performed the tight-binding calculations and contributed to the theoretical analysis. I.M. and S.L. supervised the project. All authors contributed substantially to the interpretation of the results. E.Ş. wrote the manuscript with input from all authors.

Competing interests

The authors declare no competing interests.


References

  1. [26]M. S. Dresselhaus, G. Dresselhaus, and P. Avouris, Carbon Nanotubes: Synthesis, Structure, Properties, and Applications (Springer, Berlin, 2001).
  2. [27]Y. Nakanishi et al., Structural diversity of single-walled transition metal dichalcogenide nanotubes grown via template reaction, Advanced Materials 35, 2306631 (2023).DOI
  3. [28]M. Liu et al., Photoluminescence from single-walled MoS₂ nanotubes coaxially grown on boron nitride nanotubes, ACS Nano 15, 8418 (2021).DOI
  4. [29]W. Dai et al., Metallic NbS₂ one-dimensional van der waals heterostructures, ACS Nano 19, 32800 (2025).arxiv.org/abs/2507.03274
  5. [30]C. Yang et al., Janus MoSSe nanotubes on 1d swcnt-bnnt van der waals heterostructures, Small 21, 2570208 (2025).DOI
  6. [31]A. E. Mikkelsen, F. T. Bølle, K. S. Thygesen, T. Vegge, and I. E. Castelli, Band structure of MoS₂ janus nanotubes, Physical Review Materials 5, 014002 (2021).
  7. [32]S. Zhao, C. Yang, Z. Zhu, X. Yao, and W. Li, Curvature-controlled band alignment transition in 1d van der waals heterostructures, npj Computational Materials 9, 92 (2023).DOI
  8. [33]W.-H. Kang, Y.-T. Hsiao, J. Genzor, Y. Zhumagulov, M.-H. Liu, and D. Kochan, Efficient Magnetic Spin-Filtering and Persistent Spin-Currents in Lifshitz-Transitioned Alternagnets: A Route to Open-Orbit Spintronics, arXiv preprint arXiv:2606.04587 (2026), arXiv:2606.04587 [cond-mat.mes-hall].arxiv.org/abs/2606.04587
  9. [34]K. Zou, Y. Yang, B. Xin, et al., Monolayer M₂X₂O as potential 2d altermagnets and half-metals: a first-principles study, Journal of Physics: Condensed Matter 37 (2024).
  10. [35]J.-Y. Li et al., Strain-induced valley polarization, topological states, and piezomagnetism in two-dimensional altermagnetic V₂Te₂O, V₂STeO, V₂SSeO, and V₂S₂O, arXiv (2024), 2411.19237.DOI
  11. [36]S. Biswas, A. Achintyaa, and P. Dutta, Altermagnetic phases and phase transitions in the Lieb-5 Hubbard model, Physical Review B 10.1103/xv8w-8sx3 (2026), accepted.
  12. [37]A. Ablimit, Y.-J. Sun, H. Jiang, C. Wu, L. Wei, X. Chen, and G. Cao, Weak metal–metal transition in the vanadium oxytelluride Rb₁−δV₂Te₂O, Physical Review Materials 2, 044801 (2018).
  13. [38]X. Cheng, Y. Gao, J. Peng, and J. Liu, Realistic tight-binding model for V₂Se₂O-family alternamagnets, arXiv (2026), arXiv:2602.09465 [cond-mat.mtrl-sci].arxiv.org/abs/2602.09465
  14. [39]R. Xu, Y. Gao, and J. Liu, Chemical design of monolayer alternamagnets, National Science Review 13, nwaf528 (2026).DOI
  15. [40]A. González-García, W. López-Pérez, P. Pacheco, L. Ramírez-Montes, and R. González-Hernández, Coexistence of d-wave altermagnetism and topological states in janus FeSe X (X = S, Te) monolayers, Physical Review Materials 10, 044004 (2026).
  16. [41]S. Smidtrup, T. Markussen, P. Vancraeyveld, J. Wellendorff, J. Schneider, T. Gunst, B. Verstichel, D. Stradi, P. A. Khomyakov, U. G. Vej-Hansen, M.-E. Lee, S. T. Chilk, F. Rasmussen, G. Penazzi, F. Corsetti, A. Ojanperä, K. Jensen, M. L. N. Palsgaard, U. Martinez, A. Blom, M. Brandbyge, and K. Stokbro, QuantumATK: an integrated platform of electronic and atomic-scale modelling tools, J. Phys.: Condens. Matter 32, 015901 (2019).
  17. [42]J. P. Perdew, K. Burke, and M. Ernzerhof, Generalized gradient approximation made simple, Phys. Rev. Lett. 77, 3865 (1996).pubmed.ncbi.nlm.nih.gov/10062328
  18. [43]M. J. Van Setten, M. Giantomassi, E. Bousquet, M. J. Verstraete, D. R. Hamann, X. Gonze, and G.-M. Rignanese, The PseudoDojo: Training and grading a 85 element optimized norm-conserving pseudopotential table, Comput. Phys. Commun. 226, 39 (2018).DOI

Paper details

vlnc:W_zpuhmbc3
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 10

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