betaFeedback

WorksPhysics and Astronomy

Direction-selective triplet pairing and spin-edge locking in altermagnetic metals

Abstract

We investigate self-consistent unconventional superconductivity in a two-dimensional dd-wave altermagnetic metal. We find that momentum-dependent altermagnetic spin splitting suppresses opposite-spin singlet pairing and stabilizes highly anisotropic equal-spin triplet order. In the spin-conserving limit, this directional triplet pairing gives rise to nearly dispersionless Majorana boundary states associated with effective one-dimensional topological channels. Rashba spin-orbit coupling mixes spin sectors, activates additional pairing components, and drives the system into a mixed-parity superconducting state with dispersive Majorana boundary states. The spin-resolved boundary spectra further reveal a characteristic locking between boundary orientation and spin polarization, reflecting the underlying altermagnetic symmetry. These results identify altermagnetic spin splitting as an intrinsic mechanism for selecting unconventional pairing and generating spin-resolved Majorana boundary states without external magnetic fields.

Introduction

The emergence of unconventional quantum phases in correlated and topological materials often arises from the interplay of multiple electronic degrees of freedom, including charge, spin, orbital, and superconducting orders [1, 2, 3, 4, 5]. Among these, the interplay between magnetism and superconductivity has attracted extensive attention [6, 7]. Magnetic order generally suppresses conventional spin-singlet superconductivity through pair-breaking effects, but it can also reconstruct the spin and momentum structure of quasiparticles and thereby promote unconventional pairing [1]. Understanding how magnetism suppresses, selects, or stabilizes superconducting order is therefore central to the search for novel superconducting and topological phases [8, 9, 10, 11].

Spin splitting plays a central role in realizing unconventional and topological superconductivity, as it lifts spin degeneracy [9, 10, 12]. When combined with spin-orbit coupling, spin splitting can generate effective pp-wave pairing and thereby support Majorana excitations [9, 13]. In conventional platforms, spin splitting is typically produced by external Zeeman fields or ferromagnetic order [11, 14, 15]. However, these routes often introduce orbital and paramagnetic depairing, as well as stray-field effects, which are detrimental to superconductivity [1, 8]. This motivates the search for magnetic systems in which intrinsic spin splitting can coexist with superconductivity.

Altermagnets have recently emerged as a distinct class of collinear compensated magnets beyond the conventional ferromagnet-antiferromagnet dichotomy [16, 17, 18, 19, 20, 21, 22, 23, 24]. Unlike conventional antiferromagnets, altermagnets can host momentum-dependent spin splitting despite vanishing net magnetization and, in many cases, even the absence of relativistic spin-orbit coupling [16, 25]. This spin splitting originates from crystalline symmetries, where opposite-spin sublattices are related by time reversal combined with rotation or mirror operations rather than by inversion or translation [25]. Consequently, altermagnets exhibit anisotropic spin-polarized Fermi surfaces that are qualitatively distinct from those generated by Zeeman fields or ferromagnets. This makes them a natural platform for exploring unconventional superconductivity in the absence of macroscopic magnetization [21].

Superconductivity in altermagnetic systems has therefore attracted increasing attention [21, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63]. Previous studies have explored both intrinsic pairing instabilities in altermagnetic metals and proximity-induced superconducting phenomena in altermagnet-superconductor hybrid structures. These works have revealed a variety of unconventional responses, including equal-spin triplet pairing, mixed singlet-triplet pairing, finite-momentum superconductivity, anomalous Josephson and Andreev effects, and possible routes to topological superconductivity [21, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 64]. More recent self-consistent calculations have further indicated that dd-wave altermagnetism, especially when combined with Rashba spin-orbit coupling (RSOC), can support mixed-parity superconducting states [37, 38, 39, 64].

Despite these advances, the microscopic mechanism by which altermagnetic spin splitting selects superconducting order remains unclear. When singlet and triplet pairing channels compete, a central question is whether the superconducting state is imposed by construction or emerges self-consistently. A further key issue is how the anisotropy dictated by altermagnetic symmetry propagates from the normal-state electronic structure to the pairing order, and ultimately to Majorana boundary states and their spin-resolved spectral signatures.

In this work, we develop a minimal self-consistent The remainder of this paper is organized as follows. In § 2, we introduce the minimal model for a two-dimensional dd-wave altermagnetic metal with RSOC and outline the self-consistent mean-field framework for determining the superconducting order parameters. In § 3, we present the pairing phase diagram, analyze the altermagnetism-driven anisotropy in the singlet-triplet competition, and discuss the resulting boundary spectra, including Majorana modes and their spin-resolved structure. Finally, Sec. IV summarizes the main results and discusses their broader implications.

Model and Method

We consider an effective two-band model for a two-dimensional dd-wave altermagnetic metal with RSOC. This model can be viewed as the low-energy projection of a four-band altermagnetic system in the strong spin-sublattice locking regime, where the bands crossing the Fermi level are described by an effective spin degree of freedom. The normal-state Hamiltonian is HN=∑kck†h0(k)ckH_N=\sum_k c_k^\dagger h_0(k)c_k, where ck†=(ck↑†,ck↓†)c_k^\dagger=(c_{k\uparrow}^\dagger,c_{k\downarrow}^\dagger), and the matrix h0(k)h_0(k) is given by

h0(k)=ϵ(k)σ0+2J0(cos⁡kx−cos⁡ky)σz+2λ(sin⁡kyσx−sin⁡kxσy).(1)h_0(\mathbf{k})=\epsilon(\mathbf{k})\sigma_0+2J_0(\cos k_x-\cos k_y)\sigma_z+2\lambda(\sin k_y\sigma_x-\sin k_x\sigma_y). \tag*{(1)}

Here ϵ(k)=−2t0(cos⁡kx+cos⁡ky)−μ\epsilon(\mathbf{k})=-2t_0(\cos k_x+\cos k_y)-\mu, where t0t_0 is the nearest-neighbor hopping amplitude and μ\mu is the chemical potential. The second term in Eq. (1) describes the dx2−y2d_{x^2-y^2}-type altermagnetic exchange field. Its form factor cos⁡kx−cos⁡ky\cos k_x-\cos k_y produces momentum-dependent spin splitting with opposite signs in different regions of the Brillouin zone. This term breaks T\mathcal{T} and C4zC_{4z} separately, but preserves the combined C4zTC_{4z}\mathcal{T} symmetry, leading to spin-split yet compensated bands. The last term is the RSOC induced by inversion-symmetry breaking, which mixes the two spin sectors.

We consider superconductivity arising from magnetic exchange interactions on nearest-neighbor bonds. The corresponding interaction Hamiltonian can be written as

HI=−V∑⟨ij⟩∑σ,σ′ciσ†cjσ′†ciσ′cjσ,(2)H_I=-V\sum_{\langle ij\rangle}\sum_{\sigma,\sigma'}c_{i\sigma}^{\dagger}c_{j\sigma'}^{\dagger}c_{i\sigma'}c_{j\sigma}, \tag*{(2)}

where VV denotes the pairing strength. The cases σ≠σ′\sigma\ne\sigma' and σ=σ′\sigma=\sigma' correspond to spin-singlet and equal-spin triplet pairings, respectively. At the mean-field level, we decouple this interaction to obtain the following bond order parameters: Δi,js=V2⟨ci↑cj↓−ci↓cj↑⟩\Delta_{i,j}^{s}=\frac{V}{2}\langle c_{i\uparrow}c_{j\downarrow}-c_{i\downarrow}c_{j\uparrow}\rangle and Δi,jt,σ=V⟨ciσcjσ⟩\Delta_{i,j}^{t,\sigma}=V\langle c_{i\sigma}c_{j\sigma}\rangle. The symmetry properties under bond reversal are:

Δi,i+μ^s=+Δi+μ^,is,(3)\Delta_{i,i+\hat{\mu}}^{s}=+\Delta_{i+\hat{\mu},i}^{s}, \tag*{(3)}
Δi,i+μ^t,σ=−Δi+μ^,it,σ.(4)\Delta_{i,i+\hat{\mu}}^{t,\sigma}=-\Delta_{i+\hat{\mu},i}^{t,\sigma}. \tag*{(4)}

where μ=x,y\mu=x,y denotes the bond direction. The plus sign for singlet pairing reflects its even orbital parity, while the minus sign for triplet pairing reflects its odd parity.

At the mean-field level, the magnitude of the singlet and triplet order parameters along the μ\mu-direction can be expressed as,

Δμs=V2N∑kcos⁡kμ F↑↓(k),(5)\Delta_{\mu}^{s}=\frac{V}{2N}\sum_k\cos k_\mu\,\mathcal{F}_{\uparrow\downarrow}(\mathbf{k}), \tag*{(5)}
Δμt,σ=V2N∑kisin⁡kμ Fσσ(k),(6)\Delta_{\mu}^{t,\sigma}=\frac{V}{2N}\sum_k i\sin k_\mu\,\mathcal{F}_{\sigma\sigma}(\mathbf{k}), \tag*{(6)}

where NN is the number of lattice sites and Fσσ′(k)=⟨ckσc−kσ′⟩\mathcal{F}_{\sigma\sigma'}(\mathbf{k})=\langle c_{k\sigma}c_{-k\sigma'}\rangle is the anomalous expectation value. We introduce the Nambu spinor Ψk=(ck↑,ck↓,c−k↑†,c−k↓†)T\Psi_k=(c_{k\uparrow},c_{k\downarrow},c_{-k\uparrow}^{\dagger},c_{-k\downarrow}^{\dagger})^T. The mean-field Bogoliubov–de Gennes (BdG) Hamiltonian is HBdG=12∑kΨk†HBdG(k)ΨkH_{\mathrm{BdG}}=\frac{1}{2}\sum_k\Psi_k^\dagger\mathcal{H}_{\mathrm{BdG}}(\mathbf{k})\Psi_k, with

HBdG(k)=(h0(k)Δ(k)Δ†(k)−h0T(−k)),(7)\mathcal{H}_{\mathrm{BdG}}(\mathbf{k})= \begin{pmatrix} h_0(\mathbf{k}) & \Delta(\mathbf{k})\\ \Delta^\dagger(\mathbf{k}) & -h_0^T(-\mathbf{k}) \end{pmatrix}, \tag*{(7)}

where Δ(k)\Delta(\mathbf{k}) is the pairing matrix:

Δ(k)=(Δt,↑(k)Δs(k)−Δs(k)Δt,↓(k)),(8)\Delta(\mathbf{k})= \begin{pmatrix} \Delta^{t,\uparrow}(\mathbf{k}) & \Delta^s(\mathbf{k})\\ -\Delta^s(\mathbf{k}) & \Delta^{t,\downarrow}(\mathbf{k}) \end{pmatrix}, \tag*{(8)}

and the pairing functions in the singlet and triplet channels are expressed as

Δs(k)=2Δxscos⁡kx+2Δyscos⁡ky,(9)\Delta^s(\mathbf{k})=2\Delta_x^s\cos k_x+2\Delta_y^s\cos k_y, \tag*{(9)}
Δt,σ(k)=2i(Δxt,σsin⁡kx+Δyt,σsin⁡ky).(10)\Delta^{t,\sigma}(\mathbf{k})=2i\left(\Delta_x^{t,\sigma}\sin k_x+\Delta_y^{t,\sigma}\sin k_y\right). \tag*{(10)}

Diagonalizing HBdG(k)\mathcal{H}_{\mathrm{BdG}}(\mathbf{k}) gives the quasiparticle eigenvalues EknE_{k n} and eigenvectors Φkn=(u1kn,u2kn,u3kn,u4kn)T\Phi_{k n}=(u_{1kn},u_{2kn},u_{3kn},u_{4kn})^T, where the four components correspond to (ck↑,ck↓,c−k↑†,c−k↓†)(c_{k\uparrow},c_{k\downarrow},c_{-k\uparrow}^{\dagger},c_{-k\downarrow}^{\dagger}). Using Eqs. (4a) and (4b), the order parameter magnitudes along the μ\mu direction can be expressed as

Δμs=V2N∑k,nu1knu4kn∗tanh⁡(Ekn2T)cos⁡kμ,\Delta_{\mu}^{s}=\frac{V}{2N}\sum_{\mathbf{k},n}u_{1kn}u_{4kn}^{*}\tanh\left(\frac{E_{kn}}{2T}\right)\cos k_{\mu},
Δμt,↑=V2N∑k,ni u1knu3kn∗tanh⁡(Ekn2T)sin⁡kμ.(11)\Delta_{\mu}^{t,\uparrow}=\frac{V}{2N}\sum_{\mathbf{k},n}i\,u_{1kn}u_{3kn}^{*}\tanh\left(\frac{E_{kn}}{2T}\right)\sin k_{\mu}. \tag*{(11)}
Δμt,↓=V2N∑k,ni u2knu4kn∗tanh⁡(Ekn2T)sin⁡kμ.\Delta_{\mu}^{t,\downarrow}=\frac{V}{2N}\sum_{\mathbf{k},n}i\,u_{2kn}u_{4kn}^{*}\tanh\left(\frac{E_{kn}}{2T}\right)\sin k_{\mu}.

To characterize the boundary topology, we diagonalize the self-consistent BdG Hamiltonian in ribbon geometries. The spin-resolved local spectral function is evaluated from the retarded Green’s function GR(kμ,ω)=[ω+iΓ−HBdG(kμ)]−1G^{R}(k_{\mu},\omega)=[\omega+i\Gamma-\mathcal{H}_{\mathrm{BdG}}(k_{\mu})]^{-1} as

Aσ(m,kμ,ω)=−1πIm⁡Gασ(m),ασ(m)R(kμ,ω),(12)A_{\sigma}(m,k_{\mu},\omega)=-\frac{1}{\pi}\operatorname{Im}G_{\alpha_{\sigma}(m),\alpha_{\sigma}(m)}^{R}(k_{\mu},\omega), \tag*{(12)}

where kμk_{\mu} is the conserved momentum along the ribbon, Γ\Gamma is a small broadening factor, and mm labels the chain normal to the boundary. In the electron sector of the Nambu basis, α↑(m)=2m−1\alpha_{\uparrow}(m)=2m-1 and α↓(m)=2m\alpha_{\downarrow}(m)=2m. The boundary spectral functions are obtained by evaluating Aσ(m,kμ,ω)A_{\sigma}(m,k_{\mu},\omega) on the outermost chains.

In the following calculations, all energies are measured in units of t0t_{0} (with t0=1t_{0}=1). Unless otherwise specified, we set μ=0\mu=0, T=10−6T=10^{-6}, a broadening factor Γ=0.006\Gamma=0.006 for the spectral functions, and the interaction strength V=1.5V=1.5.

Results and Discussion

We begin by examining the superconducting order parameters obtained from the self-consistent calculation. The self-consistent solutions reveal several robust features [65]. For the spin-singlet channel, we always obtain Δxs=−Δys\Delta_{x}^{s}=-\Delta_{y}^{s}, indicating a dominant dx2−y2d_{x^{2}-y^{2}}-wave pairing (hereafter referred to as dd-wave). For the equal-spin triplet channel, the pairing is highly anisotropic. Unlike previous studies that often impose predefined pairing symmetries (e.g., chiral px+ipyp_{x}+ip_{y}) [30], our self-consistent calculation treats the xx and yy bond orders independently Eq. (11), revealing a direction-selective triplet pairing. In the absence of RSOC (λ=0\lambda=0), the triplet order becomes quasi-one-dimensional within each spin sector: for spin up, only the yy-directed bond component is finite (Δyt,↑≠0\Delta_{y}^{t,\uparrow}\ne0, Δxt,↑≈0\Delta_{x}^{t,\uparrow}\approx0); for spin down, the opposite holds (Δxt,↓≠0\Delta_{x}^{t,\downarrow}\ne0, Δyt,↓≈0\Delta_{y}^{t,\downarrow}\approx0). The preserved C4zTC_{4z}T symmetry further relates the two spin sectors, enforcing Δyt,↑=Δxt,↓\Delta_{y}^{t,\uparrow}=\Delta_{x}^{t,\downarrow}. Thus, the triplet pairing (henceforth pp-wave) in opposite spin sectors is linked by a π/2\pi/2 rotation. Consequently, when presenting the self-consistent pairing amplitudes, we may focus on the spin-up triplet components without loss of generality.

Figure 1 summarizes the self-consistent order parameters as functions of J0J_{0} and λ\lambda. Figure 1(a)-Figure 1(c) show the magnitudes of the dd-wave singlet amplitude Δx\Delta_{x} and the spin-up triplet components Δ↑,px\Delta_{\uparrow,p_{x}} and Δ↑,py\Delta_{\uparrow,p_{y}} (where Δ↑,pμ≡Δμt,↑\Delta_{\uparrow,p_{\mu}}\equiv\Delta_{\mu}^{t,\uparrow}) over the J0J_{0}-λ\lambda plane. Figure 1(d) displays the evolution with J0J_{0} at λ=0\lambda=0, clearly demonstrating the emergence of Δ↑,py\Delta_{\uparrow,p_{y}} and Δ↓,px\Delta_{\downarrow,p_{x}} (the latter not shown) while Δ↑,px\Delta_{\uparrow,p_{x}} and Δ↓,py\Delta_{\downarrow,p_{y}} remain negligible. Figure 1(e) shows the dependence on λ\lambda at fixed J0J_{0}, and Figure 1(f) illustrates the momentum-space winding of the triplet pairing at a representative point.

Image not extracted

Figure 1. Self-consistent superconducting order parameters and their phase structure in the dd-wave altermagnetic system with RSOC. (a)–(c) Magnitudes of the spin-singlet dd-wave order parameter Δd\Delta_d and the equal-spin triplet components Δ↑,px\Delta_{\uparrow,\mathit{p}x} and Δ↑,py\Delta_{\uparrow,\mathit{p}y} as functions of the altermagnetic splitting J0J_0 and the RSOC strength λ\lambda. The insets show the Fermi surfaces for (a) λ=0\lambda=0, J0=0.01J_0=0.01 and (b) λ=0.2\lambda=0.2, J0=0.2J_0=0.2. (d) Order parameters as functions of J0J_0 at λ=0\lambda=0. (e) Order parameters as functions of λ\lambda at fixed J0=0.2J_0=0.2. (f) Phase of the equal-spin triplet pairing along the Fermi surface for the spin-up and spin-down sectors at the star-marked point (λ=0.2,J0=0.12)(\lambda=0.2,J_0=0.12).

For small J0J_{0}, the altermagnetic spin splitting is modest, as seen in the Fermi surface inset of Figure 1(a); the Fermi-surface mismatch between opposite spins is limited, which favors singlet dd-wave pairing. Indeed, Figure 1(a) shows that the dd-wave component dominates at weak J0J_{0}, while triplet components remain small. With increasing J0J_{0}, the momentum-dependent spin splitting reconstructs the Fermi surface and enhances the spin-dependent mismatch, leading to a progressive suppression of singlet pairing and a relative enhancement of triplet pairing.

RSOC qualitatively modifies this pairing structure. As indicated by the Fermi surface in the inset of Figure 1(b), the electronic states acquire mixed spin character. This relaxes the spin-selective pairing pattern at λ=0\lambda=0 and activates otherwise suppressed pairing components, allowing singlet and triplet orders to coexist in a mixed-parity superconducting state. Most notably, in each spin sector the originally quasi-one-dimensional triplet pairing acquires a finite pxp_{x} component, evolving into a px+ipyp_{x}+ip_{y}-like state, as evidenced by the finite Δ↑,px\Delta_{\uparrow,p_{x}} seen in Figure 1(b) at finite λ\lambda. Nevertheless, the pyp_{y} component remains dominant over the pxp_{x} component, as can be seen by comparing Δ↑,py\Delta_{\uparrow,p_{y}} [Fig. 1(c)] and Δ↑,px\Delta_{\uparrow,p_{x}} [Fig. 1(b)], indicating that the induced p+ipp+ip pairing retains a strong anisotropy inherited from the altermagnetic spin splitting.

To further examine the internal structure of the triplet order parameter, we focus on the λ=0\lambda=0 limit, where szs_{z} is conserved and the two spin sectors are decoupled. In this limit, the altermagnetic exchange field lifts the equivalence between the two in-plane directons within each spin sector, allowing the triplet order to select a preferred bond direction. As shown in Figure 1(d), increasing J0J_{0} suppresses the singlet amplitude, while only two equal-spin triplet components become sizable: Δ↑,py\Delta_{\uparrow,p_{y}} (and, by C4zTC_{4z}T symmetry, Δ↓,px\Delta_{\downarrow,p_{x}}). In contrast, Δ↑,px\Delta_{\uparrow,p_{x}} and Δ↓,py\Delta_{\downarrow,p_{y}} remain strongly suppressed [65]. This demonstrates that altermagnetic spin splitting not only favors triplet pairing but also imposes a spin-dependent directional selection within the triplet sector.

The origin of this directional selectivity can be understood as follows. From the perspective of the self-consistent pairing kernel Eq. (4b), the two triplet components in the spin-up sector can be written schematically as

Δμp,↑∝∑kF↑↑(k)sin⁡kμ,μ=x,y.(13)\Delta_{\mu}^{p,\uparrow}\propto\sum_{\mathbf{k}}F_{\uparrow\uparrow}(\mathbf{k})\sin k_{\mu},\qquad\mu=x,y. \tag*{(13)}

The relative magnitudes of Δ↑,px\Delta_{\uparrow,\mathit{p}x} and Δ↑,py\Delta_{\uparrow,\mathit{p}y} are governed by the projections of F↑↑(k)F_{\uparrow\uparrow}(\boldsymbol{k}) onto the odd-parity form factors sin⁡kx\sin k_x and sin⁡ky\sin k_y. Because the altermagnetic exchange field has a dx2−y2d_{x^2-y^2} character, it distinguishes the two in-plane directions within a fixed spin sector and yields an anisotropic pairing kernel, making the two projections unequal. This selects one triplet bond component over the other.

The same phenomenon can be further understood within a phenomenological Ginzburg–Landau theory [65]. For a conventional C4zC_{4z}-symmetric two-component pp-wave order parameter η=(ηx,ηy)\boldsymbol{\eta}=(\eta_x,\eta_y), the quadratic free energy is isotropic,

F(2)=α(∣ηx∣2+∣ηy∣2),F^{(2)}=\alpha\left(|\eta_x|^2+|\eta_y|^2\right),

so that pxp_x and pyp_y are degenerate at quadratic order. In the altermagnetic state, the unitary C4zC_{4z} symmetry is broken within each spin sector, allowing an anisotropic quadratic free energy. For the spin-up sector, we may write

F↑(2)=αx↑∣ηx↑∣2+αy↑∣ηy↑∣2,F_{\uparrow}^{(2)}=\alpha_{x\uparrow}|\eta_{x\uparrow}|^2+\alpha_{y\uparrow}|\eta_{y\uparrow}|^2,

with αx↑≠αy↑\alpha_{x\uparrow}\ne\alpha_{y\uparrow} in general. If αy↑<αx↑\alpha_{y\uparrow}<\alpha_{x\uparrow}, the leading instability occurs in the pyp_y channel, giving (ηx↑,ηy↑)∝(0,1)(\eta_{x\uparrow},\eta_{y\uparrow})\propto(0,1) (see Supplemental Material for details [65]). Thus, this symmetry-constrained quadratic anisotropy explains why Δ↑,py\Delta_{\uparrow,p_y} (and Δ↓,px\Delta_{\downarrow,p_x}) dominate in the self-consistent solution, while Δ↑,px\Delta_{\uparrow,p_x} (and Δ↓,py\Delta_{\downarrow,p_y}) remain strongly suppressed.

We next turn to the finite-Rashba-coupling regime. Figure 1(e) shows the evolution of the superconducting order parameters with λ\lambda at fixed J0J_0. As λ\lambda increases, the triplet components suppressed in the λ=0\lambda=0 limit acquire finite amplitudes, accompanied by an enhancement of the dd-wave component. This behavior reflects Rashba-induced spin mixing, which relaxes the spin-selective pairing constraint and allows singlet and triplet components to admix. Consequently, the superconducting state evolves from a direction-selective triplet state into a mixed-parity phase with coexisting pairing components. The coexistence of the two triplet bond components also gives rise to a nontrivial phase texture in momentum space [65]. When both pxp_x and pyp_y components are finite, the triplet pairing function in Eq. (7b) acquires a momentum-dependent complex phase. We evaluate ϕσ(k)=arg⁡[Δp,σ(k)]\phi_\sigma(\boldsymbol{k})=\arg[\Delta^{p,\sigma}(\boldsymbol{k})] along a circular loop k(θ)=R(cos⁡θ,sin⁡θ)\boldsymbol{k}(\theta)=R(\cos\theta,\sin\theta) centered at Γ\Gamma and define the corresponding winding as

Wσ(R)=12π∫02πdθ ∂θϕσ[k(θ)].(14)W_\sigma(R) = \frac{1}{2\pi}\int_{0}^{2\pi} d\theta\,\partial_\theta\phi_\sigma[k(\theta)]. \tag*{(14)}

As shown in Figure 1(f), the two spin sectors exhibit opposite windings, with W↑=−1W_{\uparrow} = -1 and W↓=+1W_{\downarrow} = +1. This indicates that Rashba-induced spin mixing generates opposite momentum-space phase windings of the equal-spin triplet pairing in the two spin sectors.

Boundary spectrum and topological properties

The momentum-space phase winding of the triplet pairing motivates an analysis of the boundary spectrum. We therefore diagonalize the BdG Hamiltonian in ribbon geometries. We first consider the λ=0\lambda= 0 limit. As discussed above, the self-consistent solution is dominated by two equal-spin triplet components, Δ↓↓p\Delta^{p}_{\downarrow\downarrow} and Δ↓↑p\Delta^{p}_{\downarrow\uparrow}, whereas Δ↑↑p\Delta^{p}_{\uparrow\uparrow} and Δ↑↓p\Delta^{p}_{\uparrow\downarrow} remain negligible. Thus, each spin sector realizes pp-wave pairing predominantly along a single crystalline direction, rather than a two-dimensional chiral px±ipyp_x \pm i p_y state. This anisotropic pairing structure effectively decomposes the system into momentum-resolved one-dimensional spinless pp-wave superconducting channels [66] and is directly reflected in the cylinder spectrum shown in Figure 2(a). For a fixed momentum along the periodic direction, the two-dimensional BdG Hamiltonian reduces to an effective one-dimensional problem along the open direction. Over a finite momentum interval, this effective one-dimensional subsystem supports Majorana end modes at the two open boundaries. These momentum-resolved end states appear as nearly dispersionless zero-energy boundary states in the cylinder spectrum.

Image not extracted

Figure 2. Boundary spectra and zero-mode wave functions of the self-consistent superconducting phases with J0=0.12J_0 = 0.12. (a) Cylinder spectrum at λ=0\lambda= 0. (b) Energy levels and spatial distributions of the two zero-energy states at a representative momentum kx=0k_x = 0 selected from (a). (c) Cylinder spectrum at λ=0.3\lambda= 0.3. (d) Cylinder spectrum at λ=0.6\lambda= 0.6.

To verify the boundary localization, we select a representative momentum kxk_x from the zero-energy manifold and diagonalize the corresponding finite one-dimensional BdG Hamiltonian. As shown in Figure 2(b), the two zero-energy eigenstates are localized at opposite open boundaries, consistent with Majorana end modes. This behavior has a simple real-space interpretation, as illustrated in Figure 3(a). In the λ=0\lambda= 0 limit, the anisotropic pairing effectively decomposes the system into decoupled Kitaev-chain-like channels stacked along the xx direction. Each channel supports Majorana zero modes at its ends, and their momentum-resolved collection gives rise to the nearly flat zero-energy boundary states in Figure 2(a).

Image not extracted

Figure 3. Schematic illustration of the evolution from decoupled Kitaev chains to a two-dimensional Majorana edge mode. (a) Decoupled Kitaev chains stacked along the xx direction, each hosting Majorana zero modes at its two ends. (b) Interchain coupling hybridizes the boundary Majorana modes. (c) In the coupled two-dimensional limit, the hybridized Majorana modes form a dispersive edge channel along the boundary.

With finite RSOC, the boundary spectrum changes qualitatively, as shown in Figure 2(c). The nearly flat zero-energy modes become dispersive away from the high-symmetry momenta, while gapless crossings remain at kx=0k_x = 0 and kx=πk_x = \pi. This evolution reflects Rashba-induced spin mixing, which couples the previously decoupled spin sectors and allows both pxp_x and pyp_y pairing components to coexist. In the real-space picture, finite RSOC introduces coupling between the Kitaev-chain-like channels, as illustrated in Figure 3(b). The resulting hybridization of boundary Majorana modes converts the flat zero-energy states into dispersive edge modes of the two-dimensional superconducting state, as shown in Figure 3(c). The residual gapless crossings at kx=0k_x = 0 and kx=πk_x = \pi can be traced to the vanishing of the Rashba term proportional to sin⁡kx\sin k_x at these momenta. The suppressed spin mixing limits the hybridization of boundary states and allows the crossings to persist. For stronger RSOC, the hybridization between boundary and bulk states is further enhanced, as shown in Figure 2(d). The edge modes gradually lose their spectral separation from the bulk continuum. To clarify the topological character of these boundary features, we evaluate the bulk invariant of the two-dimensional BdG Hamiltonian.

The superconducting state belongs to class D because time-reversal symmetry is broken by the altermagnetic order [67]. In two dimensions, a nonzero Chern number in this class would correspond to a chiral topological superconducting phase with a net chiral Majorana edge channel. Our calculation, however, yields a vanishing total Chern number. This is consistent with the boundary spectrum, where the edge branches near kx=0k_x = 0 and kx=πk_x = \pi disperse with opposite chiralities and therefore do not produce a net chiral edge response [Fig. 4].

The residual gapless crossings nevertheless admit a natural momentum-resolved interpretation. At kx=0k_x = 0 and kx=πk_x = \pi, the Rashba term proportional to sin⁡kx\sin k_x vanishes, thereby suppressing the hybridization between the relevant boundary channels. At these fixed momenta, the BdG Hamiltonian can be viewed as an effective one-dimensional pp-wave superconductor along the open direction, which supports Majorana end states. The gapless crossings in the ribbon spectrum can therefore be interpreted as the momentum-resolved spectral signatures of these one-dimensional Majorana end states [30, 35].

Boundary orientation and spin polarization

Motivated by the symmetry properties of altermagnets, where spin degeneracy is lifted in a momentum-dependent manner while the net magnetization remains zero, we characterize the spin structure of the boundary states through spin-resolved spectral functions. The spin polarization depends sensitively on both momentum and boundary orientation, making a spin-resolved analysis essential. We therefore evaluate the boundary spectral functions using the retarded Green’s function defined in Eq. (12). Figure 4(a)-Figure 4(d) show representative spectral weights at the four boundaries of the ribbon geometry. For clarity, only the dominant spin component at each boundary is displayed; the complete set of spin-resolved spectra for all boundary-spin combinations is provided in the Supplemental Material [65].

Image not extracted

Figure 4. Spin-resolved boundary spectral functions calculated from the Green’s function with J0=0.12J_0 = 0.12 and λ=0.6\lambda= 0.6. (a) Spin-up spectral weight at the lower edge (y=1y = 1). (b) Spin-up spectral weight at the upper edge (y=Ny = N). (c) Spin-down spectral weight at the left edge (x=1x = 1). (d) Spin-down spectral weight at the right edge (x=Nx = N). The boundary states exhibit clear dispersion, with opposite slopes near k=0k = 0 and k=πk = \pi.

For the boundaries at y=1y = 1 and y=Ny = N, the low-energy spectral weight is dominated by the spin-up component, whereas for the boundaries at x=1x = 1 and x=Nx = N, the dominant contribution arises from the spin-down component. Thus, changing the boundary orientation from the yy edge to the xx edge switches the dominant spin sector of the boundary states. This establishes a clear correlation between boundary orientation and spin polarization, reflecting the momentum-dependent spin splitting of the altermagnetic normal state rather than conventional Rashba spin-momentum locking.

This behavior follows directly from the C4zTC_{4z}T symmetry of the altermagnetic state. Under this symmetry, a π/2\pi/2 rotation in real space is accompanied by spin reversal, mapping the boundary states on the yy edges to those on the xx edges with opposite dominant spin polarization. Consequently, the boundary orientation determines the dominant spin character of the boundary states. Such a spin-edge locking provides a characteristic signature of altermagnetic superconductivity and may be probed by spin-resolved tunneling spectroscopy [68] or spin-sensitive quasiparticle-interference measurements [69].

Summary

In summary, we have investigated the influence of dd-wave altermagnetic spin splitting on superconducting pairing and boundary properties in a two-dimensional metal. Self-consistent calculations demonstrate that altermagnetism suppresses singlet pairing and stabilizes anisotropic equal-spin triplet order. Within a Ginzburg-Landau framework, this behavior is attributed to a symmetry-allowed quadratic anisotropy between the pxp_x and pyp_y channels, which selects direction-dependent triplet components, while the C4zTC_{4z}T symmetry constrains their relation between opposite spin sectors. RSOC further mixes the spin sectors and activates additional pairing channels, resulting in a mixed-parity superconducting state with nontrivial triplet phase winding. This modification is reflected in the boundary spectrum, which evolves from nearly flat Majorana modes associated with effectively one-dimensional channels to dispersive boundary excitations.

The spin-resolved boundary spectra reveal a characteristic spin-edge locking, in which the dominant spin polarization of the boundary states is determined by the crystallographic orientation of the edge. This feature follows from the underlying C4zTC_{4z}T symmetry and provides a distinct spectroscopic signature. These results establish altermagnetic metals as a platform for symmetry-controlled pairing and spin-selective Majorana boundary phenomena without external magnetic fields.

Supplementary material for “Direction-selective triplet pairing and spin-edge locking in altermagnetic metals”

In this Supplementary Material, we provide additional theoretical analysis and supporting results for the main text. Section S-1 presents a Ginzburg–Landau analysis of the two-component pp-wave order parameter and shows how altermagnetic anisotropy leads to direction-selective triplet pairing. Section S-2 gives a complete account of the self-consistent superconducting state, including the pairing amplitudes in both spin sectors, their symmetry relations, the momentum-space phase winding of the triplet pairing, and the spin-resolved boundary spectral functions underlying the spin–edge correspondence.

GINZBURG–LANDAU ANALYSIS OF ALTERMAGNETIC ANISOTROPY

We analyze the effect of altermagnetic anisotropy on the equal-spin triplet order parameter within a Ginzburg–Landau (GL) theory [8]. For a two-component pp-wave order parameter, the superconducting gap can be written as

Δ(k)=ηxφx(k)+ηyφy(k),(15)\Delta(\mathbf{k}) = \eta_x\varphi_x(\mathbf{k}) + \eta_y\varphi_y(\mathbf{k}), \tag*{(15)}

where ηx\eta_x and ηy\eta_y are complex order-parameter components and the basis functions may be chosen as φx(k)∼sin⁡kx\varphi_x(\mathbf{k}) \sim\sin k_x and φy(k)∼sin⁡ky\varphi_y(\mathbf{k}) \sim\sin k_y. The GL free energy is expanded as

F=F(2)+F(4)+⋯ ,(16)F = F^{(2)} + F^{(4)} + \cdots, \tag*{(16)}

with

F(2)=∑ijαijηi∗ηj,F(4)=∑ijklβijklηi∗ηj∗ηkηl.(17)F^{(2)} = \sum_{ij} \alpha_{ij}\eta_i^*\eta_j,\qquad F^{(4)} = \sum_{ijkl}\beta_{ijkl}\eta_i^*\eta_j^*\eta_k\eta_l. \tag*{(17)}

Here, the quadratic coefficients determine the leading superconducting instability, whereas the quartic terms determine how different components compete or coexist once the order parameter becomes finite.

We first recall the conventional C4C_4-symmetric case for comparison. In this case, the two components pxp_x and pyp_y form a two-dimensional representation. The quadratic term is constrained to be isotropic,

Fiso(2)=α(∣ηx∣2+∣ηy∣2),(18)F_{\mathrm{iso}}^{(2)} = \alpha\left(|\eta_x|^2 + |\eta_y|^2\right), \tag*{(18)}

so that pxp_x and pyp_y remain degenerate at quadratic order. The relative amplitude and phase are therefore selected by quartic invariants. A convenient form of the quartic free energy is

Fiso(4)=β1(∣ηx∣2+∣ηy∣2)2+β2∣ηx2+ηy2∣2+β3∣ηx∣2∣ηy∣2.(19)F_{\mathrm{iso}}^{(4)} = \beta_1\left(|\eta_x|^2 + |\eta_y|^2\right)^2 + \beta_2\left|\eta_x^2 + \eta_y^2\right|^2 + \beta_3|\eta_x|^2|\eta_y|^2. \tag*{(19)}

The β2\beta_2 term controls the relative phase between the two components. For β2>0\beta_2 > 0, the free energy favors minimizing ∣ηx2+ηy2∣2|\eta_x^2 + \eta_y^2|^2, which is achieved by the chiral combinations

(ηx,ηy)∝(1,±i).(20)(\eta_x,\eta_y) \propto(1,\pm i). \tag*{(20)}

For β2<0\beta_2 < 0, the free energy instead favors real combinations. Within the real sector, the β3\beta_3 term further distinguishes whether a single-component state, such as (1,0)(1,0) or (0,1)(0,1), or a two-component real state, such as (1,±1)(1,\pm1), is preferred. Thus, in a C4C_4-symmetric system, pxp_x and pyp_y are degenerate at quadratic order, and the detailed pairing structure is selected by quartic terms.

In the present model, the unitary C4zC_{4z} symmetry and time-reversal symmetry T\mathcal{T} are separately broken, while the combined symmetry C4zTC_{4z}\mathcal{T} is preserved. Consequently, within a fixed spin sector, pxp_x and pyp_y are no longer related by a unitary fourfold rotation and need not have the same quadratic coefficient. The quadratic GL free energy for one spin sector can therefore contain an anisotropic term,

FAM(2)=αx∣ηx∣2+αy∣ηy∣2,(21)F_{\mathrm{AM}}^{(2)} = \alpha_x|\eta_x|^2 + \alpha_y|\eta_y|^2, \tag*{(21)}

with αx≠αy\alpha_x \ne\alpha_y in general. Equivalently, one may write

αx=α+δα,αy=α−δα,(22)\alpha_x = \alpha+ \delta\alpha,\qquad\alpha_y = \alpha- \delta\alpha, \tag*{(22)}

so that δα\delta\alpha measures the altermagnetic anisotropy between the two pp-wave components. This anisotropy changes the mechanism of order-parameter selection. If δα<0\delta\alpha< 0, then αx<αy\alpha_x < \alpha_y, and the pxp_x component has the stronger superconducting instability. The leading state is then

(ηx,ηy)∝(1,0).(23)(\eta_x,\eta_y) \propto(1,0). \tag*{(23)}

Conversely, if δα>0\delta\alpha> 0, the pyp_y component is favored and

(ηx,ηy)∝(0,1).(24)(\eta_x,\eta_y) \propto(0,1). \tag*{(24)}

Therefore, unlike the isotropic C4C_4-symmetric case, the altermagnetic state can select a single pp-wave component already at the quadratic level. The quartic terms still affect possible subleading admixtures and the stability of two-component states, but the leading directional selection is controlled by the anisotropic quadratic coefficients.

For the full spinful altermagnetic system, the remaining C4zTC_{4z}T symmetry relates the two spin sectors. This symmetry maps a spin-up component with pyp_y character to a spin-down component with pxp_x character, and similarly maps a spin-up component with pxp_x character to a spin-down component with pyp_y character. Accordingly, the quadratic coefficients obey

αy↑=αx↓,αx↑=αy↓.(25)\alpha_{y\uparrow}=\alpha_{x\downarrow},\qquad\alpha_{x\uparrow}=\alpha_{y\downarrow}. \tag*{(25)}

Thus, if the spin-up sector favors ηy↑\eta_{y\uparrow}, the spin-down sector favors ηx↓\eta_{x\downarrow}. The resulting leading pairing structure is

ηy↑≠0,ηx↓≠0,ηx↑≃0,ηy↓≃0.(26)\eta_{y\uparrow}\ne0,\qquad\eta_{x\downarrow}\ne0,\qquad\eta_{x\uparrow}\simeq0,\qquad\eta_{y\downarrow}\simeq0. \tag*{(26)}

This is in direct agreement with the self-consistent solution, where Δy↑p\Delta^p_{y\uparrow} and Δx↓p\Delta^p_{x\downarrow} are the dominant equal-spin triplet components. The essential point is that the directional selection in the altermagnetic state is not primarily determined by the sign of a quartic coefficient such as β2\beta_2, but by the quadratic anisotropy induced by the dx2−y2d_{x^2-y^2}-type altermagnetic spin splitting.

Complete self-consistent pairing structure

Pairing amplitudes and C4zTC_{4z}T symmetry

In the main text, we show representative self-consistent order parameters to illustrate the evolution of the superconducting state. For completeness, here we present the full set of pairing components in both spin sectors. This analysis provides a direct check of how the self-consistent pairing structure is constrained by the combined C4zTC_{4z}T symmetry of the altermagnetic state.

We first consider the case without Rashba spin-orbit coupling, λ=0\lambda=0. In this limit, szs_z is conserved and the two spin sectors are decoupled. We extract the equal-spin triplet components Δxσp\Delta^p_{x\sigma} and Δyσp\Delta^p_{y\sigma} for σ=↑,↓\sigma=\uparrow,\downarrow, together with the opposite-spin dd-wave singlet component Δd\Delta_d. The full results are shown in Figure S1 and Figure S1. For the spin-up sector, the dominant triplet component is Δ↑,py\Delta_{\uparrow,p_y}, whereas Δ↑,px\Delta_{\uparrow,p_x} remains strongly suppressed. For the spin-down sector, the dominant component is Δ↓,px\Delta_{\downarrow,p_x}, while Δ↓,py\Delta_{\downarrow,p_y} is negligible. Thus, the anisotropic triplet pairing is selected along opposite crystalline directions in the two spin sectors, as dictated by the combined C4zTC_{4z}T symmetry. The operation rotates the momentum structure by π/2\pi/2 and flips the spin, thereby relating

Image not extracted

Figure S1. Complete self-consistent superconducting order parameters in the two spin sectors. (a),(b) Order parameters as functions of J0J_0 at λ=0\lambda= 0 for the spin-up and spin-down sectors, respectively. (c),(d) Order parameters as functions of λ\lambda at fixed J0=0.2J_0 = 0.2 for the spin-up and spin-down sectors, respectively. The equal-spin triplet components Δxσp\Delta^{p}_{x\sigma} and Δyσp\Delta^{p}_{y\sigma} and the opposite-spin dd-wave singlet component Δd\Delta_d are shown.

Δ↑,px⟷Δ↓,py,Δ↑,py⟷Δ↓,px.(27)\Delta_{\uparrow,p_x}\longleftrightarrow\Delta_{\downarrow,p_y},\qquad\Delta_{\uparrow,p_y}\longleftrightarrow\Delta_{\downarrow,p_x}. \tag*{(27)}

We next examine the finite-Rashba-coupling case. For λ≠0\lambda\ne0, spin is no longer a good quantum number because the Rashba term mixes the two spin sectors. Consequently, pairing components that are strongly suppressed at λ=0\lambda=0, such as Δ↑,px\Delta_{\uparrow,p_x} and Δ↓,py\Delta_{\downarrow,p_y}, acquire finite amplitudes. The corresponding complete results are shown in Figure S1 and Figure S1. Although Rashba spin-orbit coupling activates additional components, the pairing structure remains anisotropic: the components connected to the dominant λ=0\lambda=0 channels remain larger than the previously suppressed ones. This demonstrates that the altermagnetic anisotropy persists even after spin mixing is introduced.

The complete self-consistent order parameters therefore support the symmetry-based interpretation used in the main text. In the spin-decoupled limit, C4zTC_{4z}T relates the dominant components Δ↑,py\Delta_{\uparrow,p_y} and Δ↓,px\Delta_{\downarrow,p_x}. With finite Rashba spin-orbit coupling, additional components appear, but the hierarchy of the triplet order parameters continues to reflect the underlying altermagnetic anisotropy.

Momentum-space phase structure of the triplet pairing

We further analyze the momentum-space phase structure of the equal-spin triplet pairing. For a given spin sector, the triplet pairing function is written as

Δσp(k)=Δxp,σsin⁡kx+Δyp,σsin⁡ky,(28)\Delta^{p}_{\sigma}(\mathbf{k}) = \Delta^{p,\sigma}_{x}\sin k_x + \Delta^{p,\sigma}_{y}\sin k_y, \tag*{(28)}

and its phase is defined by

ϕσ(k)=arg⁡[Δσp(k)].(29)\phi_{\sigma}(\mathbf{k}) = \arg[\Delta^{p}_{\sigma}(\mathbf{k})]. \tag*{(29)}

Figures S2(a) and S2(c) show the phase distribution ϕσ(k)\phi_{\sigma}(\mathbf{k}) in the Brillouin zone for the spin-up and spin-down triplet pairing functions, respectively. The color scale denotes the phase angle in the range [−π,π][-\pi, \pi]. The phase exhibits strong momentum dependence, indicating a nontrivial internal phase structure of the equal-spin triplet pairing.

To quantify this phase texture, we evaluate the phase winding along a closed loop in momentum space. The loop is chosen as a circle centered at Γ\Gamma,

k(θ)=k0(cos⁡θ,sin⁡θ),0≤θ<2π,(30)\mathbf{k}(\theta) = k_0(\cos\theta,\sin\theta), \qquad0 \leq\theta< 2\pi, \tag*{(30)}

where k0k_0 is a fixed radius chosen to avoid nodes of Δσp(k)\Delta^{p}_{\sigma}(\mathbf{k}). Along this path, the phase becomes

ϕσ(θ)=arg⁡[Δσp(k(θ))].(31)\phi_{\sigma}(\theta) = \arg[\Delta^{p}_{\sigma}(\mathbf{k}(\theta))]. \tag*{(31)}

The corresponding winding number is

Wσ=12π∫02πdθ ∂θϕσ(θ).(32)W_{\sigma} = \frac{1}{2\pi}\int_{0}^{2\pi} d\theta\,\partial_{\theta}\phi_{\sigma}(\theta). \tag*{(32)}

The resulting ϕσ(θ)\phi_{\sigma}(\theta) is shown in Figure S2 and Figure S2. As θ\theta evolves from 00 to 2π2\pi, the phase winds by approximately ±2π\pm2\pi, giving a nonzero winding number.

Image not extracted

Figure S2. Momentum-space phase structure of the equal-spin triplet pairing. (a),(c) Phase distribution ϕσ(k)\phi_{\sigma}(\mathbf{k}) in the Brillouin zone for the spin-up and spin-down triplet pairing functions, respectively. (b),(d) Phase ϕσ(θ)\phi_{\sigma}(\theta) evaluated along a circular loop k(θ)=R(cos⁡θ,sin⁡θ)\mathbf{k}(\theta) = R(\cos\theta, \sin\theta) centered at Γ\Gamma for the spin-up and spin-down sectors, respectively. Common parameters are J0=0.12J_0 = 0.12 and λ=0.6\lambda= 0.6.

Spin-resolved boundary spectral functions

In the main text, we present representative spin-resolved boundary spectra to illustrate the spin-edge correspondence of the superconducting state. For completeness, here we show the full set of boundary spectral functions, including both spin components on all four edges of the ribbon geometry. The spin-resolved boundary spectral function is evaluated from the retarded Green’s function of the Bogoliubov-de Gennes Hamiltonian,

GR(k∥,ω)=[ω+iΓ−HBdG(k∥)]−1,(33)G^{R}(k_{\parallel},\omega) = \left[\omega+ i\Gamma- H_{\mathrm{BdG}}(k_{\parallel})\right]^{-1}, \tag*{(33)}

where k∥k_{\parallel} denotes the conserved momentum along the periodic direction and Γ\Gamma is a small broadening parameter. The local spectral function for spin σ\sigma on the mm-th chain is given by

Aσ(m,k∥,ω)=−1πIm⁡Gασ(m),ασ(m)R(k∥,ω),(34)A_{\sigma}(m,k_{\parallel},\omega) = -\frac{1}{\pi}\operatorname{Im}G^{R}_{\alpha_{\sigma}(m),\alpha_{\sigma}(m)}(k_{\parallel},\omega), \tag*{(34)}

where α↑(m)=2m−1\alpha_{\uparrow}(m) = 2m - 1 and α↓(m)=2m\alpha_{\downarrow}(m) = 2m label the electron sector of the Nambu basis.

For each boundary, we evaluate the spectral function at the outermost chains, i.e., m=1m = 1 or m=Nm = N, and resolve the two spin components separately. The complete results are shown in Figure S3, where the spectral weights are plotted for the four boundaries at y=1y = 1, y=Ny = N, x=1x = 1, and x=Nx = N for both spin-up and spin-down sectors. The full data confirm a pronounced correlation between boundary orientation and spin polarization. For the boundaries normal to the yy direction (y=1y = 1 and y=Ny = N), the low-energy spectral weight is dominated by the spin-up component, while the spin-down contribution remains strongly suppressed. In contrast, for the boundaries normal to the xx direction (x=1x = 1 and x=Nx = N), the dominant spectral weight arises from the spin-down component. Thus, changing the boundary orientation from the yy edge to the xx edge switches the dominant spin sector of the boundary states.

Image not extracted

Figure S3. Complete spin-resolved boundary spectral functions. (a1),(a2) Spin-up and spin-down spectral weights at the lower edge (y=1y = 1), respectively. (a3),(a4) Spin-up and spin-down spectral weights at the upper edge (y=Ny = N), respectively. (b1),(b2) Spin-up and spin-down spectral weights at the left edge (x=1x = 1), respectively. (b3),(b4) Spin-up and spin-down spectral weights at the right edge (x=Nx = N), respectively. Common parameters are J0=0.12J_0 = 0.12 and λ=0.6\lambda= 0.6.

The subdominant components remain finite due to Rashba spin–orbit coupling, but their spectral weights are much smaller than those of the dominant components. The boundary states therefore retain a well-defined spin character on each edge. This spin–edge correspondence is summarized schematically in Figure S4. The boundaries normal to the yy direction are dominated by spin-up spectral weight, whereas those normal to the xx direction are dominated by spin-down spectral weight.

Image not extracted

Figure S4. Schematic illustration of the spin–edge correspondence in the altermagnetic superconducting state. The boundaries normal to the yy direction (y=1,Ny = 1, N) host spin-up dominated modes, while those normal to the xx direction (x=1,Nx = 1, N) host spin-down dominated modes. This edge-dependent spin polarization reflects the underlying C4zTC_{4z}T symmetry of the altermagnetic state.

This edge-dependent spin polarization follows from the C4zTC_{4z}T symmetry of the altermagnetic state. The C4zC_{4z} operation exchanges the xx and yy directions, while TT reverses the spin, leading to the correspondence

(y-edge,↑)⟷(x-edge,↓).(35)(y\text{-edge}, \uparrow) \longleftrightarrow(x\text{-edge}, \downarrow). \tag*{(35)}

Thus, boundary states on orthogonal edges are symmetry-related and carry opposite dominant spin components. The complete spin-resolved boundary spectra, together with the schematic in Figure S4, provide a comprehensive confirmation of the spin–edge correspondence discussed in the main text.

References

  1. [1]F. S. Bergeret, A. F. Volkov, and K. B. Efetov, Odd triplet superconductivity and related phenomena in superconductor-ferromagnet structures, Rev. Mod. Phys. 77, 1321 (2005).DOI
  2. [2]E. Berg, E. Fradkin, S. A. Kivelson, and J. M. Tranquada, Striped superconductors: How spin, charge and superconducting orders intertwine in the cuprates, New J. Phys. 11, 115004 (2009).DOI
  3. [3]D. J. Scalapino, A common thread: The pairing interaction for unconventional superconductors, Rev. Mod. Phys. 84, 1383 (2012).DOI
  4. [4]E. Fradkin, S. A. Kivelson, and J. M. Tranquada, Colloquium: Theory of intertwined orders in high temperature superconductors, Rev. Mod. Phys. 87, 457 (2015).DOI
  5. [5]D. J. Scalapino, E. Loh, and J. E. Hirsch, d-wave pairing near a spin-density-wave instability, Phys. Rev. B 34, 8190 (1986).pubmed.ncbi.nlm.nih.gov/9939525
  6. [6]D. A. Dikin, M. Mehta, C. W. Bark, C. M. Folkman, C. B. Eom, and V. Chandrasekhar, Coexistence of superconductivity and ferromagnetism in two dimensions, Phys. Rev. Lett. 107, 056802 (2011).DOI
  7. [7]U. Li, C. Richter, J. Mannhart, and R. C. Ashoori, Coexistence of magnetic order and two-dimensional superconductivity at LaAlO3/SrTiO3 interfaces, Nat. Phys. 7, 762 (2011).DOI
  8. [8]M. Sigrist and K. Ueda, Phenomenological theory of unconventional superconductivity, Rev. Mod. Phys. 63, 239 (1991).DOI
  9. [9]J. Alicea, New directions in the pursuit of majorana fermions in solid state systems, Rep. Prog. Phys. 75, 076501 (2012).DOI
  10. [10]X.-L. Qi and S.-C. Zhang, Topological insulators and superconductors, Rev. Mod. Phys. 83, 1057 (2011).DOI
  11. [11]K. Flensberg, F. von Oppen, and A. Stern, Engineered platforms for topological superconductivity and majorana zero modes, Nat. Rev. Mater. 6, 944 (2021).arxiv.org/abs/2103.05548
  12. [12]M. Sato and Y. Ando, Topological superconductors: A review, Rep. Prog. Phys. 80, 076501 (2017).
  13. [13]C. Beenakker, Search for majorana fermions in superconductors, Annu. Rev. Condens. Matter Phys. 4, 113 (2013).
  14. [14]R. Cai, I. Žutić, and W. Han, Superconductor/ferromagnet heterostructures: A platform for superconducting spintronics and quantum computation, Adv. Quantum Technol. 6, 2200080 (2023).arxiv.org/abs/2211.08625
  15. [15]S. Das Sarma, K. Laubscher, H. Pan, J. D. Sau, and T. D. Stanescu, Rashba spin-orbit coupling and artificially engineered topological superconductors, Phys. Rev. B 113, 139604 (2026).DOI
  16. [16]L. Šmejkal, J. Sinova, and T. Jungwirth, Emerging research landscape of altermagnetism, Phys. Rev. X 12, 040501 (2022).arxiv.org/abs/2204.10844
  17. [17]I. Mazin and The PRX Editors, Editorial: Altermagnetism—a new punch line of fundamental magnetism, Phys. Rev. X 12, 040002 (2022).DOI
  18. [18]Z. Xiao, J. Zhao, Y. Li, R. Shindou, and Z.-D. Song, Spin space groups: Full classification and applications, Phys. Rev. X 14, 031037 (2024).DOI
  19. [19]L. Bai, W. Feng, S. Liu, L. Šmejkal, Y. Mokrousov, and Y. Yao, Altermagnetism: Exploring new frontiers in magnetism and spintronics, Adv. Funct. Mater. 34, 2409327 (2024).
  20. [20]S. S. Fender, O. Gonzalez, and D. K. Bediako, Altermagnetism: A chemical perspective, J. Am. Chem. Soc. 147, 2257 (2025).DOI
  21. [21]Z. Liu, H. Hu, and X.-J. Liu, Altermagnetism and superconductivity: A short historical review, arXiv:2510.09170.DOI
  22. [22]T. Jungwirth, J. Sinova, R. M. Fernandes, Q. Liu, H. Watanabe, S. Murakami, S. Nakatsuji, and L. Šmejkal, Symmetry, microscopy and spectroscopy signatures of altermagnetism, Nature(London) 649, 837 (2026).
  23. [23]X. Chen, W. Chen, and Q. Liu, The rise of unconventional magnetism, arXiv:2603.27505.arxiv.org/abs/2603.27505
  24. [24]X.-J. Luo, J.-X. Hu, M. Hu, and K. T. Law, Spin group symmetry criteria for unconventional magnetism, arXiv:2603.07643.
  25. [25]L. Šmejkal, J. Sinova, and T. Jungwirth, Beyond conventional ferromagnetism and antiferromagnetism: A phase with nonrelativistic spin and crystal rotation symmetry, Phys. Rev. X 12, 031042 (2022).DOI
  26. [26]Y. Fukaya, B. Lu, K. Yada, Y. Tanaka, and J. Cayao, Superconducting phenomena in systems with unconventional magnets, J. Phys.: Condens. Matter 37, 313003 (2025).DOI
  27. [27]J. A. Ouassou, A. Brataas, and J. Linder, Dc josephson effect in altermagnets, Phys. Rev. Lett. 131, 076003 (2023).DOI
  28. [28]M. Papaj, Andreev reflection at the altermagnet-superconductor interface, Phys. Rev. B 108, L060508 (2023).DOI
  29. [29]C. Sun, A. Brataas, and J. Linder, Andreev reflection in altermagnets, Phys. Rev. B 108, 054511 (2023).
  30. [30]D. Zhu, Z.-Y. Zhuang, Z. Wu, and Y. Yan, Topological superconductivity in two-dimensional altermagnetic metals, Phys. Rev. B 108, 184505 (2023).DOI
  31. [31]S. A. A. Ghorashi, T. L. Hughes, and J. Cano, Altermagnetic routes to majorana modes in zero net magnetization, Phys. Rev. Lett. 133, 106601 (2024).DOI
  32. [32]M. Ezawa, Topological insulators and superconductors based on p-wave magnets: Electrical control and detection of a domain wall, Phys. Rev. B 110, 165429 (2024).DOI
  33. [33]K. Parshukhov and A. P. Schnyder, Exotic superconducting states in altermagnets, arXiv:2507.10700.
  34. [34]N. Heinsdorf and M. Franz, Proximitizing altermagnets with conventional superconductors, Phys. Rev. B 113, L020501 (2026).
  35. [35]O. Alam, A. Pal, P. Dutta, and A. Saha, Proximity-induced superconductivity and emerging topological phases in altermagnet-based heterostructures, Phys. Rev. B 113, 155429 (2026).arxiv.org/abs/2510.26894
  36. [36]X. Xiao and A. Bansil, Nodal topological superconductivity driven by crystalline antiunitary symmetry in altermagnets, arXiv:2605.13656.arxiv.org/abs/2605.13656
  37. [37]X. Zou, R. M. Fernandes, and E. Fradkin, Superconducting states and intertwined orders in metallic altermagnets, arXiv:2603.04503.DOI
  38. [38]V. S. de Carvalho and H. Freire, Unconventional superconductivity in altermagnets with spin-orbit coupling, Phys. Rev. B 110, L220503 (2024).DOI
  39. [39]D. Chakraborty and A. M. Black-Schaffer, Constraints on superconducting pairing in altermagnets, Phys. Rev. B 112, 014516 (2025).arxiv.org/abs/2408.03999
  40. [40]C. W. J. Beenakker and T. Vakhtel, Phase-shifted andreev levels in an altermagnet josephson junction, Phys. Rev. B 108, 075425 (2023).
  41. [41]Y.-X. Li and C.-C. Liu, Majorana corner modes and tunable patterns in an altermagnet heterostructure, Phys. Rev. B 108, 205410 (2023).DOI
  42. [42]D. Zhu, D. Liu, Z.-Y. Zhuang, Z. Wu, and Z. Yan, Field-sensitive dislocation bound states in two-dimensional d-wave altermagnets, Phys. Rev. B 110, 165141 (2024).DOI
  43. [43]S.-B. Zhang, L.-H. Hu, and T. Neupert, Finite-momentum cooper pairing in proximitized altermagnets, Nat. Commun. 15, 1801 (2024).
  44. [44]B. Lu, K. Maeda, H. Ito, K. Yada, and Y. Tanaka, φ josephson junction induced by altermagnetism, Phys. Rev. Lett. 133, 226002 (2024).DOI
  45. [45]Q. Cheng and Q.-F. Sun, Orientation-dependent josephson effect in spin-singlet superconductor/altermagnet/spin-triplet superconductor junctions, Phys. Rev. B 109, 024517 (2024).DOI
  46. [46]Y.-X. Li, Realizing tunable higher-order topological superconductors with altermagnets, Phys. Rev. B 109, 224502 (2024).DOI
  47. [47]J. Gondolf, A. Kreisel, M. Roig, Y. Yu, D. F. Agterberg, and B. M. Andersen, Local signatures of altermagnetism, Phys. Rev. B 111, 174436 (2025).DOI
  48. [48]H.-Y. Sun, L. Liu, Y.-T. Zhang, and Z. Qiao, Altermagnetism-induced iron-based third-order topological superconductivity, Phys. Rev. B 112, 195402 (2025).DOI
  49. [49]M. Alipourzadeh and Y. Hajati, Andreev bound states and supercurrent in an unconventional superconductor-altermagnet josephson junction, Phys. Rev. B 111, 214515 (2025).DOI
  50. [50]J.-X. Hu, O. Matsyshyn, and J. C. W. Song, Nonlinear superconducting magnetoelectric effect, Phys. Rev. Lett. 134, 026001 (2025).DOI
  51. [51]Z.-T. Sun, X. Feng, Y.-M. Xie, B. T. Zhou, J.-X. Hu, and K. T. Law, Pseudo-ising superconductivity induced by p-wave magnetism, Phys. Rev. B 112, 214504 (2025).DOI
  52. [52]D. Mondal, A. Pal, A. Saha, and T. Nag, Distinguishing between topological majorana and trivial zero modes via transport and shot noise study in an altermagnet heterostructure, Phys. Rev. B 111, L121401 (2025).arxiv.org/abs/2409.08009
  53. [53]A. Hadjipaschalis, S. A. A. Ghorashi, and J. Cano, Majoranas with a twist: Tunable majorana zero modes in altermagnetic heterostructures, Phys. Rev. B 112, 214430 (2025).DOI
  54. [54]A. Pal, D. Mondal, T. Nag, and A. Saha, Josephson current signature of floquet majorana and topological accidental zero modes in altermagnet heterostructures, Phys. Rev. B 112, L201408 (2025).DOI
  55. [55]K. Maeda, Y. Fukaya, K. Yada, B. Lu, Y. Tanaka, and J. Cayao, Classification of pair symmetries in superconductors with unconventional magnetism, Phys. Rev. B 111, 144508 (2025).DOI
  56. [56]C. Li, J.-X. Hu, F.-C. Zhang, S.-B. Zhang, and L.-H. Hu, Spin-polarized josephson supercurrent in nodeless altermagnets, Phys. Rev. Lett. 136, 116701 (2026).arxiv.org/abs/2509.13838
  57. [57]R. Dsouza, A. Kreisel, B. M. Andersen, D. F. Agterberg, and M. H. Christensen, Odd-parity magnetism in fe-based superconductors with coplanar magnetic order, Phys. Rev. B 113, 144509 (2026).DOI
  58. [58]P. Sharma and N. Mohanta, Double-peak majorana bound states in altermagnet–superconductor heterostructures, arXiv:2603.25844.arxiv.org/abs/2603.25844
  59. [59]X.-J. Yi, Y. Mao, C.-M. Miao, and Q.-F. Sun, Majorana modes in a helical altermagnet without net magnetism and spin-orbit coupling, Phys. Rev. B 113, L060408 (2026).DOI
  60. [60]K. Monkman, J. Weng, N. Heinsdorf, A. Nocera, Y. Barlas, and M. Franz, Persistent spin currents in superconducting altermagnets, Phys. Rev. X 16, 011057 (2026).
  61. [61]X. Wan, Z. Qin, F. Zhan, J. Zeng, D.-H. Xu, and R. Wang, Chiral superconductivity in periodically driven altermagnet/superconductor heterostructures, arXiv:2604.01596.arxiv.org/abs/2604.01596
  62. [62]Z. Liu, H. Hu, and X.-J. Liu, Fulde-ferrell-larkin-ovchinnikov states and topological bogoliubov fermi surfaces in altermagnets: An analytical study, Phys. Rev. B 113, 024518 (2026).DOI
  63. [63]A. A. Mazanik, R. de las Heras, and F. S. Bergeret, Superconductivity as a probe of altermagnetism: Critical temperature, field, and current, arXiv:2603.06188.arxiv.org/abs/2603.06188
  64. [64]C. L. H. Rasmussen, J. Gondolf, M. Barkman, M. Roig, D. F. Agterberg, A. Kreisel, and B. M. Andersen, Inherent momentum-dependent gap structure of altermagnetic superconductors, arXiv:2509.03247.DOI
  65. [65]See Supplemental Material [at URL will be inserted by publisher] for details on (I) the Ginzburg–Landau analysis of altermagnetic anisotropy, (II) the complete self-consistent pairing amplitudes and their symmetry relations, (III) the momentum-space phase winding of the triplet pairing, and (IV) the full spin-resolved boundary spectral functions, which includes Ref. [8].
  66. [66]A. Y. Kitaev, Unpaired majorana fermions in quantum wires, Phys. Usp. 44, 131 (2001).DOI
  67. [67]A. P. Schnyder, S. Ryu, A. Furusaki, and A. W. W. Ludwig, Classification of topological insulators and superconductors in three spatial dimensions, Phys. Rev. B 78, 195125 (2008).DOI
  68. [68]R. Wiesendanger, Spin mapping at the nanoscale and atomic scale, Rev. Mod. Phys. 81, 1495 (2009).DOI
  69. [69]J. E. Hoffman, Spectroscopic scanning tunneling microscopy insights into Fe-based superconductors, Rep. Prog. Phys. 74, 124513 (2011).DOI

Paper details

vlnc:W_2jca4mgm
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 6

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