betaFeedback

WorksPhysics and Astronomy

Generation of spin-triplet Cooper pairs via a canted antiferromagnet

Abstract

Spinful triplet Cooper pairs can be generated from their singlet counterparts available in a conventional superconductor (S) using two or more noncollinear magnetic moments, typically contributed by different magnets in a multilayered heterostructure. Here, we theoretically demonstrate that an S interfaced with a canted antiferromagnet (AF) harbors spinful triplet Cooper pairs capitalizing on the intrinsic noncollinearity between the two AF sublattice magnetizations. As the AF canting can be controlled by an applied field, our work proposes a simple bilayer structure that admits controllable generation of spin-triplet Cooper pairs. Employing the Bogoliubov-de Gennes framework, we delineate the spatial dependence of the spin-triplet correlations. We further evaluate the superconducting critical temperature as a function of the AF canting, which provides one experimental observable associated with the emergence of these triplet correlations.

Introduction

The dissipationless flow of charge in superconductors is partly responsible for their central role in the various emerging quantum technologies [1, 2]. The widely available conventional superconductors, such as Al and Nb, are made of spin-singlet Cooper pairs, which harbor no net spin [3]. A superconductor hosting spin-triplet Cooper pairs can support dissipationless spin currents [4, 5, 6, 7, 8], deemed valuable for switching magnetic memories [9, 10, 11, 12, 13], as well as exotic excitations, such as Majorana bound states [14, 15]. Such an unconventional superconductor can be engineered from its conventional counterpart employing heterostructures incorporating magnetic multilayers [5, 6, 7, 8, 16]. The basic requirement for achieving spinful triplets from spin-singlets is exposing the latter to two or more noncollinear spin-splitting fields. A wide variety of multilayered hybrids comprising conventional superconductors (S) and ferromagnets (F) has been employed to achieve the desired spinful triplets [17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33], coming a long way from the initial critical temperature studies [34, 35].

Since a Néel ordered antiferromagnet (AF) bears no net spin or magnetic moment, for some time it was considered inert at causing spin-splitting in an adjacent superconductor. Indeed, early experiments found a metallic AF to behave just like a normal metal when considering its effect on an adjacent superconductor [36]. More recent experiments, on the other hand, found the AF to substantially affect the adjacent S with intriguing dependencies [37, 38, 39, 40, 41, 42]. From the theory perspective, Josephson junctions [43, 44] and interfaces [45, 46] involving itinerant AFs were shown to exhibit non-trivial properties due to quasiparticle reflections. Moreover, a recent work demonstrates that an uncompensated interface of an insulating AF with an adjacent S induces a strong spin-splitting as well as spin-flip scattering thereby strongly influencing the S [47]. Subsequent work found even the fully compensated interface between the AF and S to be spin-active [48]. This has been understood as due to the AF inducing Néel triplets whose pairing amplitude has an alternating sign in space similar to the AF spin [49]. Altogether, the potential usefulness of AFs [50] in engineering novel superconducting effects and devices, such as a filter [51], is starting to be understood.

Pekar and Rashba [52] already recognized long ago that even though the net spin vanishes in an AF, at the lattice constant length scale, the AF harbors a spin or magnetization profile that rapidly varies in space changing its sign from one lattice site to the next, which should manifest itself in physical observables. As a result, the AF with its two sublattice magnetizations antiparallel to each other generates zero-spin Néel triplets [49]. Proceeding further along this line of thought, a homogeneous canted AF with its sublattice magnetizations deviating from an antiparallel alignment effectively harbors a noncollinear spin texture capable of generating spinful triplet Cooper pairs in an adjacent S. This exciting possibility is theoretically examined in the present work. Furthermore, a recent experiment [53] demonstrates generation and use of spinful triplet correlations employing the intrinsically noncollinear ground state of a kagome AF.

Here, employing the Bogoliubov-de Gennes (BdG) framework [54], we theoretically investigate a bilayer structure consisting of an insulating AF exchange coupled via a compensated interface to an adjacent S. We examine the critical temperature and spin-triplet correlations in the S as a function of the canting in the AF, which allows us to continuously tune the AF from its collinear antiparallel state to it effectively becoming an F. We find that Néel triplets are generated in the S both from the interband pairing channel considered recently [49], and from the conventional intraband pairing. The former channel is dominant at half-filling in which case states from two different electronic bands are energetically close to the chemical potential and can participate in forming these unconventional Néel triplet Cooper pairs [49]. This can be compared with pairing at finite energies in other multisublattice systems [55]. The conventional intraband pairing channel dominates away from half-filling when only states within the same electronic band are energetically close to each other and the chemical potential. We find that the intraband pairing channel results in Néel triplets formation due to an imprinting of the Néel character by the AF on the normal state electronic wavefunctions in the SS. Although this channel of Néel triplet generation is found to be much weaker than the interband pairing channel, it admits qualitatively new effects. We show that it is only in the intraband pairing channel that spinful Néel triplets are generated due to the intrinsic noncollinearity of a canted AF. The SS critical temperature variation as a function of the AF canting angle is found to be consistent with the intraband Néel triplets being weaker than their interband counterparts, and may offer a convenient experimental signature of this interplay.

The paper is organized as follows. Section II introduces the model and BdG framework employed in our analysis. The dependence of spin-triplet correlations on space and AF canting is discussed in Sec. III, while the variation of superconducting critical temperature is discussed in § 1.3. Until this point, we consider a one-dimensional model that allows a simple and semianalytic understanding of the essential physics. In § 1.4, we employ a two-dimensional (2-D) model for the SS to validate our prior results and further examine the spatial dependence of the s-wave and p-wave triplet correlations. We conclude with discussion and summary of the key points in Sec. VI. The appendices provide details of the BdG framework, analytic evaluation of the normal state electronic properties in the bilayer, a discussion of the decay length of the spin-triplet correlations inside the SS, critical temperature results for the 2-D system, triplet correlations for a different configuration of the AF sublattice magnetizations, and the parameters employed in our numerical routines.

AF/S bilayer model

We consider a bilayer structure comprising an insulating AF exchange coupled via a compensated interface to the SS layer, as depicted schematically in Figure 1(a). We anticipate that all three kinds of spin-triplet correlations will be generated in the SS when the AF sublattice magnetizations are canted [Figure 1(a)]. We employ the Bogoliubov-de Gennes method and numerically evaluate the superconducting properties self-consistently [54]. The canted-AF is taken to be an ideal insulator with a large band-gap. Consequently, the Hamiltonian is formulated only for the itinerant electrons in SS as they never enter the insulating AF. The AF’s influence on the S is accounted for by incorporating a spatially dependent spin-splitting caused by the AF spins [47, 56, 57]. Furthermore, with the aim of allowing a semianalytic understanding to the essential physics, we consider a one-dimensional model as depicted schematically in Figure 1(b). The resulting Hamiltonian is given by

Figure 1. (a) Schematic depiction of the system and key physics under investigation. Equal-spin and zero-spin triplet correlations are generated in a conventional s-wave spin-singlet superconductor when it is interfaced with a canted antiferromagnet (canted-AF). This results from the intrinsic noncollinearity between the two AF sublattice magnetizations. (b) Schematic depiction of our model investigated using the Bogoliubov-de Gennes method. The black circles represent lattice sites of the superconductor hosting itinerant electrons and Cooper pairs. Blue (red) circles represent A (B) sublattice sites of the electrically insulating canted-AF. The blue and red arrows denote the local AF magnetic moments. The canting angle θt\theta_t allows us to vary the magnet from being a collinear AF (θt=0\theta_t = 0) to a ferromagnet (θt=π/2\theta_t = \pi/2).

H=−μ∑j,σcj,σ†cj,σ−t∑⟨i,j⟩∑σci,σ†cj,σ−J2∑jM⃗j⋅S⃗j+∑j(∣Δj∣2U+Δj∗cj,↓†cj,↑†+Δjcj,↑cj,↓)(1)H = -\mu\sum_{j,\sigma} c_{j,\sigma}^{\dagger}c_{j,\sigma} - t \sum_{\langle i,j\rangle}\sum_{\sigma} c_{i,\sigma}^{\dagger}c_{j,\sigma} - \frac{J}{2}\sum_j \vec{M}_j \cdot\vec{S}_j + \sum_j \left(\frac{|\Delta_j|^2}{U} + \Delta_j^{*}c_{j,\downarrow}^{\dagger}c_{j,\uparrow}^{\dagger} + \Delta_j c_{j,\uparrow}c_{j,\downarrow}\right) \tag*{(1)}

where cj,σ†c_{j,\sigma}^{\dagger} (cj,σc_{j,\sigma}) is the creation (annihilation) operator of an electron with spin σ\sigma at site jj of the S layer, with j=1,2,…,Nj = 1, 2, \ldots, N as the site index. Here, the spin quantization axis is taken along the zz-axis. We further consider periodic boundary conditions by allowing electrons to hop between sites j=1j = 1 and j=Nj = N.

In Eq. (1), μ\mu is the chemical potential. Within our theoretical method, it is determined via the filling factor ff, which is the fraction of filled electronic states in the system. For f=0.5f = 0.5 (half-filled band), we obtain μ=0\mu= 0. This corresponds to the Fermi wave-vector kF=π/2ak_F = \pi/2a located at the AF Brillouin zone boundary, where aa is the lattice constant. The normal state electronic dispersion and properties for the case of an AF in its collinear anti-parallel state have been discussed in Appendix § 1.2.1. For f≠0.5f \ne0.5, μ\mu is non-zero and the Fermi level is away from the AF Brillouin zone boundary. In the following analysis, we will consider the two qualitatively distinct cases of f=0.5f = 0.5 and f≠0.5f \ne0.5, corresponding to μ=0\mu= 0 and μ≠0\mu\ne0.

The second term in Eq. (1) is the kinetic energy term, describing hopping between nearest neighboring sites ⟨i,j⟩\langle i,j\rangle with t>0t > 0 as the hopping parameter.

The third term in the Hamiltonian [Eq. (1)] accounts for the spin-splitting due to the localized magnetic moments in the canted-AF [6, 47, 58, 59]. JM⃗j/2J\vec{M}_j/2 is the local spin-splitting field which causes an energy shift of spin-up and spin-down electrons by ∓J/2\mp J/2, with respect to the local spin-quantization axis along M⃗j\vec{M}_j. The strength of this spin-splitting, and thus the value of JJ in our model, depends on the S thickness and several other parameters [47]. It can thus be tuned in a broad range via appropriate thin films fabrication [47] as per the experimental requirements. Here, M⃗j=[(−1)j+1cos⁡θtx^+sin⁡θty^]\vec{M}_j = [(-1)^{j+1}\cos\theta_t \hat{x} + \sin\theta_t \hat{y}] is the unit vector along the direction of local magnetic moment at the jthj^{\mathrm{th}} canted-AF site. The AF sublattice magnetizations are taken to be in the xx-yy plane to examine the spin-triplet correlation with the quantization axis (zz) perpendicular to this plane. A different magnetic configuration has also been investigated and discussed in the appendix.

S⃗j=[(cj,↑†cj,↑−cj,↓†cj,↓)x^+(cj,↑†cj,↓+cj,↓†cj,↑)x^+(−icj,↓†cj,↑+icj,↑†cj,↓)y^]\vec{S}_j = [(c^{\dagger}_{j,\uparrow}c_{j,\uparrow} - c^{\dagger}_{j,\downarrow}c_{j,\downarrow})\hat{x} + (c^{\dagger}_{j,\uparrow}c_{j,\downarrow} + c^{\dagger}_{j,\downarrow}c_{j,\uparrow})\hat{x} + (-ic^{\dagger}_{j,\downarrow}c_{j,\uparrow} + ic^{\dagger}_{j,\uparrow}c_{j,\downarrow})\hat{y}] is the spin operator of an electron at site jj of the superconductor.

The last term in Eq. (1) accounts for the conventional s-wave spin-singlet superconducting correlations. It is obtained by mean field approximation of the pairing interaction −U∑jnj,↑nj,↓-U\sum_j n_{j,\uparrow}n_{j,\downarrow}, where U>0U > 0 is the attractive pairing potential and nj,σ=cj,σ†cj,σn_{j,\sigma} = c^{\dagger}_{j,\sigma}c_{j,\sigma} is the number operator [54]. Δj=−U⟨cj,↓cj,↑⟩\Delta_j = -U\langle c_{j,\downarrow}c_{j,\uparrow}\rangle is the resulting superconducting order parameter.

The total Hamiltonian Eq. (1) is numerically diagonalized and the superconducting state is determined self-consistently as detailed in Appendix § 1.2.2. The exact parameters employed in our numerical routines have been specified in Appendix § 3.1.

Triplet correlations

In this section, we quantify and investigate the different spin-triplet correlations in the S.

Consider the anomalous Matsubara Green’s function Fjj,σσ′(τ)=−⟨Tτcj,σ(τ)cj,σ′(0)⟩F_{jj,\sigma\sigma'}(\tau)=-\langle T_\tau c_{j,\sigma}(\tau)c_{j,\sigma'}(0)\rangle, where τ=it~\tau=i\tilde{t} is the imaginary time with t~\tilde{t} as the time [60, 61]. Further, TτT_\tau is the ordering operator for imaginary time τ\tau. In the Fourier space, we obtain

Fjj,σσ′(iωl)=∫0βeiωlτFjj,σσ′(τ) dτ,(2)F_{jj,\sigma\sigma'}(i\omega_l)=\int_{0}^{\beta}e^{i\omega_l\tau}F_{jj,\sigma\sigma'}(\tau)\,\mathrm{d}\tau, \tag*{(2)}

where β=ℏ/kBT\beta=\hbar/k_BT, kBk_B is the Boltzmann constant, TT is the temperature, and ωl=(2l+1)π/β\omega_l=(2l+1)\pi/\beta are the fermionic Matsubara frequencies with integer ll. See Appendix § 1.2.2 for further calculation details. Since the spin-triplet correlations are odd in frequency [5], we take a sum over all positive Matsubara frequencies to define an appropriate dimensionless quantity that would allow us to quantify the correlations

Fj,σσ′=1β∑ωl>0Fjj,σσ′(iωl).(3)F_{j,\sigma\sigma'}=\frac{1}{\beta}\sum_{\omega_l>0}F_{jj,\sigma\sigma'}(i\omega_l). \tag*{(3)}

Employing this notation, we express the relevant superconducting correlations

Fjs=12(Fj,↓↑−Fj,↑↓),(4)F_j^s=\frac{1}{2}\left(F_{j,\downarrow\uparrow}-F_{j,\uparrow\downarrow}\right), \tag*{(4)}
Fjt,z=−12(Fj,↓↑+Fj,↑↓),(5)F_j^{t,z}=-\frac{1}{2}\left(F_{j,\downarrow\uparrow}+F_{j,\uparrow\downarrow}\right), \tag*{(5)}
Fjt,x=12(Fj,↑↑−Fj,↓↓),(6)F_j^{t,x}=\frac{1}{2}\left(F_{j,\uparrow\uparrow}-F_{j,\downarrow\downarrow}\right), \tag*{(6)}
Fjt,y=i2(Fj,↑↑+Fj,↓↓).(7)F_j^{t,y}=\frac{i}{2}\left(F_{j,\uparrow\uparrow}+F_{j,\downarrow\downarrow}\right). \tag*{(7)}

FsF^s is the spin-singlet correlation. Ft,zF^{t,z}, Ft,xF^{t,x}, and Ft,yF^{t,y} are the zero-spin triplet correlations when spin is measured along zz-, xx-, and yy-axis respectively. Together, the latter three [Eqs. (5)-(7)] allow us to express all three kinds of the spin-triplet correlations with zz quantization axis. We evaluate the quantities defined in Eqs. (4)-(7) to investigate the different superconducting correlations in our system.

Numerical results

In an isolated conventional superconductor, only FsF^s is non-zero while Ft,xF^{t,x}, Ft,yF^{t,y}, and Ft,zF^{t,z} are zero. Now, if we consider an F/S bilayer, then the propagating electronic wavefunctions in the normal state of the S film acquire a spin-dependent phase associated with their spin-dependent energies resulting from the spin-splitting induced by the F [62]. The result is a relative phase difference between the opposite spin electrons as they propagate. In the superconducting state of the S layer, this causes the zero-spin triplet correlation with spin-quantization axis along the F magnetization to become non-zero [5, 62]. For example, if the magnetization of the ferromagnet is along the zz-axis, then Ft,zF^{t,z} becomes non-zero while Ft,xF^{t,x} and Ft,yF^{t,y} remain zero. Similarly, Ft,xF^{t,x} and Ft,yF^{t,y} become non-zero when magnetization of the ferromagnet is along xx- and yy-axis, respectively.

Quasiclassical theory [60, 61] shows that the triplet vector F⃗jt=Fjt,xx^+Fjt,yy^+Fjt,zz^\vec{F}_j^t = F_j^{t,x}\hat{x}+F_j^{t,y}\hat{y}+F_j^{t,z}\hat{z} always has a component aligning with the local exchange field JM⃗j/2J\vec{M}_j/2 whether the magnetization of the ferromagnet in a F/S bilayer is homogeneous or inhomogeneous [4, 63]. We, therefore, study these correlations in order to decompose the contribution of antiferromagnetic and ferromagnetic components of the canted-AF.

For θt=0\theta_t = 0 (Figure 1), our considered AF becomes a collinear antiferromagnet with the axis of magnetic moments along the xx-direction. As we increase the value of θt\theta_t by a small amount, the canted-AF acquires a net magnetization along the yy-direction. So the canted-AF can be decomposed into an antiferromagnetic component (along the xx-axis) and a ferromagnetic component (along the yy-axis). For a collinear AF, we obtain Néel triplets. This means the component of F⃗jt\vec{F}_j^t parallel to the axis of the Néel vector modulates with the Néel order of the AF [49]. For θt=π/2\theta_t = \pi/2, we effectively obtain a F/S structure with the F magnetization along the yy-axis.

We now investigate the case of θt=π/4\theta_t = \pi/4 that produces maximum noncollinearity between the two AF sublattice magnetization. In Figure 2, triplet correlations for a canted-AF/S bilayer have been plotted as a function of space for filling factor f=0.6f = 0.6. We see that Ft,zF^{t,z}, the component of F⃗t\vec{F}^t perpendicular to the canted-AF sublattice magnetizations plane, is also being generated along with the in-plane components Ft,xF^{t,x} and Ft,yF^{t,y}. It oscillates from a constant positive to negative value with Néel order. This oscillation can be understood in terms of the noncollinearity that generates Ft,zF^{t,z}. From one lattice site to the next, the angle between the spin-splitting at the site and its direct neighbors is changing sign. However, Ft,zF^{t,z} only appears for non-zero μ\mu, and is zero at half-filling (f=0.5f = 0.5) where μ=0\mu= 0. Ft,xF^{t,x} too oscillates between a constant positive and negative value with Néel order, while Ft,yF^{t,y} is constant in space. Both Ft,xF^{t,x} and Ft,yF^{t,y} are imaginary, consistent with previous theoretical results for F/S bilayers [62, 63]. The AF sublattice magnetization configuration has been so chosen here to obtain and focus on the non-zero and spatially constant sum of the equal-spin triplets F↑↑F_{\uparrow\uparrow} and F↓↓F_{\downarrow\downarrow}. In Appendix F, we present the correlations for a configuration when the AF Néel order is aligned with the zz-axis [49]. To conclude this discussion, the intrinsic noncollinearity of the canted AF successfully generates all three components of the spin-triplet correlations. The generation of spinful triplet Cooper pairs has conventionally been accomplished via the noncollinearity between the magnetizations of different ferromagnetic layers [33].

Figure 2. Spatial variation of the normalized triplet correlations for 10 lattice sites considering maximal noncollinearity corresponding to the canting angle θt=π/4\theta_t = \pi/4 and filling factor f=0.6f = 0.6 (μ/Δ0≈37\mu/\Delta_0 \approx37, where 2Δ02\Delta_0 is the zero-temperature superconducting gap without the adjacent AF). We plot the real part of the zero-spin triplet correlation Fjt,zF_j^{t,z} [panel (a)], and the imaginary parts of the spin-triplet correlations Fjt,xF_j^{t,x} [panel (b)] and Fjt,yF_j^{t,y} [panel (c)]. The imaginary part of the former and the real parts of the latter two are zero. All the correlations are normalized by the spatially averaged magnitude of the singlet correlation ∣Fs∣\lvert F^s\rvert. The detailed parameters employed for the numerical evaluation are specified in Appendix § 3.1.

We now examine the dependence of these spin-triplets on the canting angle. To this end, we plot the average magnitudes of the three spin-triplet correlations vs. the canting angle θt\theta_t in Figure 3. As we change θt\theta_t from 00 to π/2\pi/2 the system changes from a collinear AF (along xx-axis) to a collinear F (along yy-axis). As discussed above, Ft,zF^{t,z} is found to vanish identically at μ=0\mu= 0 (f=0.5f = 0.5) for all canting angles. This will be explained further below. However, for non-zero μ\mu (away from the half-filling case), it increases from 0 to a finite value as we go from a collinear antiferromagnetic alignment to maximal noncollinearity between the sublattice magnetic moments, and decreases back to zero in the ferromagnetic alignment [Fig. 3(a)]. This component thus results directly due to the AF sublattice magnetization noncollinearity.

Image not extracted

Figure 3. Variation of triplet correlations with canting angle θt\theta_t for filling factors f=0.5f = 0.5 (μ=0\mu= 0) and f=0.6f = 0.6 (μ/Δ0≈37\mu/\Delta_0 \approx37), where Δ0\Delta_0 represents the zero-temperature value of the superconducting order parameter in absence of the canted-AF layer. (a) The average magnitude of the normalized spin-triplet correlation Ft,z‾\overline{F^{t,z}} is maximum when the noncollinearity between the two sublattices is maximum, i.e. at θt=π/4\theta_t = \pi/4. However, it is zero at half-filling (f=0.5f = 0.5). (b) Average magnitude of the normalized spin-triplet correlation Ft,x‾=[F↑↑−F↓↓]/2\overline{F^{t,x}} = [F_{\uparrow\uparrow} - F_{\downarrow\downarrow}]/2 decreases as the effective antiferromagnetism becomes weaker with increasing θt\theta_t. (c) The average magnitude of the normalized spin-triplet correlation Ft,y‾=[i(F↑↑+F↓↓)]/2\overline{F^{t,y}} = [i(F_{\uparrow\uparrow} + F_{\downarrow\downarrow})]/2 increases as the effective ferromagnetism becomes stronger with increasing θt\theta_t. The averages are taken over all the sites and are denoted via an overhead bar. The detailed parameters employed for the numerical evaluation are specified in Appendix § 3.1.

The spatially averaged value of ∣Ft,x∣|F^{t,x}| decreases as θt\theta_t goes from 00 to π/2\pi/2 [Fig. 3(b)]. This component, therefore, essentially follows the Néel vector magnitude and appears to stem directly from the antiferromagnetism [49]. It has been understood as being due to the interband pairing, which is feasible when μ\mu is smaller or comparable to the superconducting gap. However, we also find such Néel triplets to be present for μ≈37Δ0\mu\approx37\Delta_0 (f=0.6f = 0.6), although they are significantly weaker than for the case of μ=0\mu= 0 [Fig. 3(b)]. We attribute this observation to a modification of the normal state electronic wavefunctions by the AF, so that even the conventional intraband pairing causes a finite generation of the Néel triplets.

Finally, the average value of ∣Ft,y∣|F^{t,y}| increases with the canting angle and it appears to be caused primarily by the net magnetization.

Insights from simplified analytics

In order to understand the difference between μ=0\mu= 0 and μ≠0\mu\ne0 cases, we examine the electronic properties of a bilayer comprising a normal metal and an AF, as detailed in Appendix A. For μ=0\mu= 0, the Fermi wave-vector is kF=π/2ak_F = \pi/2a. This means that the electrons participating in the formation of Cooper pairs have ∣k∣∼π/2a|k| \sim\pi/2a. The eigenfunctions with ∣k∣∼π/2a|k| \sim\pi/2a are such that the probability of finding an electron is non-zero on one sublattice while it is zero on the other sublattice. Now, the electrons near the Fermi level of the superconductor which interact with sublattice A do not see sublattice B and vice-versa. Therefore the triplet correlations are generated independently by sublattices A and B. The resultant correlations we obtain are a sum of correlations generated by the two sublattices. So, the only non-zero components are Ft,xF^{t,x} and Ft,yF^{t,y}. This separation of the two sublattices at a special value of the electronic chemical potential is reminiscent of a similar result obtained for spin pumping via AFs [64, 65, 66, 67].

In contrast, for μ≠0\mu\ne0, the Fermi level is within one of the bands and the wavefunctions of the states near the Fermi energy are such that the electrons on a site of the A sublattice also have a non-zero probability at the sites of the B sublattice. There is no way for the electrons to arrange themselves to decouple the two sublattices [66, 67]. The electrons experience a spin-splitting field in one direction at site 1 (of sublattice A) and in another direction at site 2 (of sublattice B), then again the first orientation at site 3 (of sublattice A). So, the electrons see the noncollinearity between the magnetic moments on the adjacent lattice sites. Therefore, the correlation Ft,zF^{t,z} along the direction perpendicular to the plane of magnetic moments of the canted-AF becomes non-zero along with the in-plane components Ft,xF^{t,x} and Ft,yF^{t,y} when μ≠0\mu\ne0. At the same time, the electronic amplitudes at the two sublattices are different due to the adjacent AF, as detailed in Appendix A. This lends the normal electronic states a weak Néel character which manifests itself in the emergence of Néel triplets even for the conventional intraband pairing.

Critical temperature

The formation of spin-triplet Cooper pairs comes at the cost of destroying their spin-singlet counterparts that are originally produced in and stabilize the superconducting state [59, 68, 69]. Hence, the critical temperature is reduced with the formation of spin-triplets, which may offer a convenient experimental signature. Thus, we investigate the critical temperature of our AF/S bilayer now via numerical self-consistent solution of the BdG equation (1).

Critical temperature TcT_c vs. canting angle θt\theta_t is plotted in Figure 4 for (a) f=0.5f = 0.5 (μ=0\mu= 0) and (b) f=0.6f = 0.6 (μ≈37Δ0\mu\approx37\Delta_0). We find that for μ=0\mu= 0, TcT_c increases with θt\theta_t while it manifests the opposite dependence for μ≠0\mu\ne0. These intriguing and distinct dependencies can be understood based on our analysis of spin-triplets generation above.

Image not extracted

Figure 4. (a) Normalized critical temperature TcT_c vs. canting angle θt\theta_t for filling factor (a) f=0.5f = 0.5 and (b) f=0.6f = 0.6 considering different values of the spin-splitting JJ. Here, Tc0T_{c0} and 2Δ02\Delta_0 are respectively the critical temperature and the zero-temperature superconducting gap of the same superconductor without the AF layer. The detailed parameters employed for the numerical evaluation are specified in Appendix § 3.1.

Let us first consider the f=0.5f = 0.5, corresponding to μ=0\mu= 0, case presented in Figure 4(a). In this case, a strong generation of spin-zero Néel triplets [Fig. 3(b)] due to interband pairing leads to maximal TcT_c suppression at θt=0\theta_t = 0. Hence, the TcT_c increases with θt\theta_t since the TcT_c suppression is stronger for the collinear AF case (θt=0\theta_t = 0) than for the F case ($\theta_t = π/2\pi/2. Further, when the exchange field is large enough, we find a complete suppression of superconductivity at θt=0\theta_t = 0 corresponding to a vanishing TcT_c. The abrupt change in TcT_c with θt\theta_t here is attributed to the additional contribution to superconductivity suppression by the opening of a normal dispersion bandgap by the AF, as described in Appendix § 1.2.1. This normal state bandgap predominantly affects the superconducting pairing at half-filling when μ=0\mu= 0.

For the case of f=0.6f = 0.6, the Néel spin-triplets generation by the antiferromagnetic order is much weaker (Figure 3). On the other hand, the ordinary spin-triplets generation by a ferromagnet remain of the same order of magnitude as for f=0.5f = 0.5. Thus, TcT_c is largest for θt=0\theta_t = 0 and it decreases with θt\theta_t. The amplitude of spin-triplets generated due to the noncollinearity (Figure 3(a)) remains small and does not seem to affect the TcT_c dependence substantially. Contrary to the μ=0\mu= 0 case, the variation of TcT_c with θt\theta_t is smooth even for large values of the exchange field JJ.

Correlations in 2-D

In our discussion above, we have considered a one-dimensional (1-D) superconductor with the aim of examining essential physics employing analytic results discussed in Appendix § 1.2.1. We now validate these results using a two-dimensional (2-D) model for the S. This further allows us to examine how the spin-triplet correlations vary with space as we move away from the S/AF interface.

In Figure 1(b), the superconducting lattice is along the xx-axis. We add more such 1-D layers in the yy-direction to create our 2-D model for the S. Each site of this 2-D sheet is indexed as (jx,jy)(j_x,j_y) so that jxj_x takes a value between 1 to NxN_x, and jyj_y takes a value between 1 to NyN_y, where NxN_x and NyN_y are the number of sites along xx- and yy-direction, respectively. The spin-splitting effect is experienced only by the electrons at the sites next to the AF/S interface. Thus, it suffices to treat the AF via the same 1-D model as before. Overall, the Hamiltonian of Eq. (1) is modified by letting all site indices jj take the form (jx,jy)(j_x,j_y). Summation in the spin-splitting term is now over the sites with indices of the form (jx,1)(j_x,1). We continue to consider periodic boundary condition along the xx-axis, like in the 1-D case.

Carrying out the BdG diagonalization self-consistently and numerically, we evaluate the spatially-resolved spin-triplet correlations [Eqs. (4)-(7)] for this system and plot them in

Figure 5. We have the same observations in the first layer (along the AF/S interface) of the 2-D sheet as the 1-D case discussed in section III. Additional calculations not presented here confirm that Ft,zF^{t,z} appears only when μ\mu is non-zero. We observe that Ft,xF^{t,x} and Ft,zF^{t,z} show modulation between positive and negative values along the yy-axis apart from along the xx-axis. This is interesting because the system has nothing imposing Néel order along the yy-axis on the correlations. In addition, we find some layers in which the alternating pattern of positive and negative values is skipped. This is attributed to Friedel-like oscillations [49, 54]. We found that these skipping of pattern only appears for non-zero μ\mu while we get a perfect alternating pattern along the yy-axis for μ=0\mu= 0 case. Although these Néel triplet correlations flip sign at the length scale of a lattice constant, their magnitude decays at the coherence length scale (shown in Appendix C). The correlation component Ft,yF^{t,y} [Figure 5(c)] is constant in each layer along the interface (along the xx-axis) but decays as we move away from the interface along the yy-direction. So, the 2-D case is consistent with and corroborates our 1-D results, resulting in a similar trend in TcT_c suppression (Appendix § 3). We find that the spatial pattern (oscillating with Néel order or being constant in space) imposed on superconducting correlations along the interfacial direction also manifests itself perpendicular to the AF/S interface.

Image not extracted

Figure 5. Spatial variation of the spin-triplet correlations for a 2-D superconductor. The sites are indexed as (jx,jy)(j_x,j_y). The sites with index (jx,1)(j_x,1) form the layer adjacent to the AF. A small section of size 14×1414 \times14 of a superconducting sheet of size 102×14102 \times14 has been plotted here for clarity. The real part of the normalized zero-spin triplet correlation Fjt,zF_j^{t,z} [panel (a)], and the imaginary parts of the normalized spin-triplet correlations Fjt,xF_j^{t,x} [panel (b)] and Fjt,yF_j^{t,y} [panel (c)] have been plotted as colormaps. The imaginary part of the former and the real parts of the latter two are zero. We consider maximal noncollinearity (θt=π/4\theta_t = \pi/4) and filling fraction f=0.6f = 0.6 (μ=18.5∣Δ‾0∣\mu= 18.5|\overline{\Delta}_0|). Here, ∣Δ‾0∣|\overline{\Delta}_0| is the average magnitude of the superconducting order parameter in absence of the canted-AF layer and all the correlations are normalized with respect to the magnitude of the singlet correlation ∣Fjs∣|F_j^s|. The detailed parameters employed for the numerical evaluation are specified in Appendix § 3.1.

In the 2-D case, where the AF/S interface breaks translational symmetry along the yy-axis, even-frequency pyp_y-wave spin-triplet correlations are created, in addition to the odd-frequency ss-wave ones [13, 15]. These are quantified as [70]

Pjt,z=18[⟨cj,↓cj+y^,↑⟩+⟨cj,↑cj+y^,↓⟩−⟨cj,↓cj−y^,↑⟩−⟨cj,↑cj−y^,↓⟩],(8)P_j^{t,z} = \frac{1}{8}\left[\langle c_{j,\downarrow}c_{j+\hat{y},\uparrow}\rangle+\langle c_{j,\uparrow}c_{j+\hat{y},\downarrow}\rangle-\langle c_{j,\downarrow}c_{j-\hat{y},\uparrow}\rangle-\langle c_{j,\uparrow}c_{j-\hat{y},\downarrow}\rangle\right], \tag*{(8)}
Pjt,x=18[−⟨cj,↑cj+y^,↑⟩+⟨cj,↓cj+y^,↓⟩+⟨cj,↑cj−y^,↑⟩−⟨cj,↓cj−y^,↓⟩],(9)P_j^{t,x} = \frac{1}{8}\left[-\langle c_{j,\uparrow}c_{j+\hat{y},\uparrow}\rangle+\langle c_{j,\downarrow}c_{j+\hat{y},\downarrow}\rangle+\langle c_{j,\uparrow}c_{j-\hat{y},\uparrow}\rangle-\langle c_{j,\downarrow}c_{j-\hat{y},\downarrow}\rangle\right], \tag*{(9)}
Pjt,y=i8[−⟨cj,↑cj+y^,↑⟩−⟨cj,↓cj+y^,↓⟩+⟨cj,↑cj−y^,↑⟩+⟨cj,↓cj−y^,↓⟩],(10)P_j^{t,y} = \frac{i}{8}\left[-\langle c_{j,\uparrow}c_{j+\hat{y},\uparrow}\rangle-\langle c_{j,\downarrow}c_{j+\hat{y},\downarrow}\rangle+\langle c_{j,\uparrow}c_{j-\hat{y},\uparrow}\rangle+\langle c_{j,\downarrow}c_{j-\hat{y},\downarrow}\rangle\right], \tag*{(10)}

where the site index j±y^≡(jx,jy±1)j\pm\hat{y}\equiv(j_x,j_y\pm1) are the nearest neighbors of site jj along the yy-axis, and Pjt,αP_j^{t,\alpha} represent the zero-spin pyp_y-wave triplet correlation when the spin is measured along the α\alpha-axis (α=x,y,z\alpha= x,y,z).

While the ss-wave spin-triplets in Figure 5(a)-(Figure 5b) nearly follow a checkerboard pattern, Pjt,zP_j^{t,z} and Pjt,xP_j^{t,x} tend to form stripes along the yy-direction [Figure 6(a)-(Figure 6b)] [71]. Unlike the on-site ss-wave pairing, the pyp_y-wave pairing cannot take advantage of the complete decoupling of the two sublattices at half-filling as they are defined on the link between sites in the A and B sublattice. For Pjt,xP_j^{t,x}, this can be seen by realizing that ⟨cj,σcj+y^,σ⟩=−⟨c(j+y^),σc(j+y^)−y^,σ⟩\langle c_{j,\sigma}c_{j+\hat{y},\sigma}\rangle= -\langle c_{(j+\hat{y}),\sigma}c_{(j+\hat{y})-\hat{y},\sigma}\rangle. Thus, if the link in the positive yy-direction from site jj gives a positive contribution to Pjt,xP_j^{t,x}, then the link in the negative yy-direction from site j+y^j+\hat{y} also gives a positive contribution, as can be seen by insertion into Eq. (9). Thus, a link from the A sublattice to the B sublattice and a link from the B sublattice to the A sublattice give a contribution to Pjt,xP_j^{t,x} of the same sign. At half-filling, the sign of Pjt,xP_j^{t,x} is thus dictated by the sign of the spin-splitting induced at the lattice site closest to the interface. Away from half-filling, where the two sublattices are no longer perfectly decoupled, the perfect stripe pattern can be shifted along the xx-direction as can be seen for Pjt,zP_j^{t,z} in Figure 6(a). The ferromagnetic component Pjt,yP_j^{t,y} in Figure 6(c) remains positive inside the whole superconductor, similar to the ss-wave triplets in Figure 5(c). Although we here observe qualitatively different behavior in the nearest-neighbor (p-wave) pairing compared to the on-site (s-wave) one, we must note that the antiferromagnetic component and the component resulting from canting are both one order of magnitude larger for the latter. Thus, the AF’s influence on the critical temperature and other key superconducting properties is dominated by the odd-frequency ss-wave Néel triplets generated in the S.

Image not extracted

Figure 6. Spatial variation of the pyp_y-wave triplet correlations in a 2-D superconductor. The sites are indexed as (jx,jy)(j_x,j_y). The sites with index (jx,1)(j_x,1) form the layer adjacent to the AF. A small section of size 12×1212 \times12 of a superconducting sheet of size 202×14202 \times14 has been plotted here for clarity. The imaginary part of the normalized zero-spin pyp_y-wave triplet correlation Pjt,zP_j^{t,z} [panel (a)], and the real parts of the normalized pyp_y-wave spin-triplet correlations Pjt,xP_j^{t,x} [panel (b)] and Pjt,yP_j^{t,y} [panel (c)] have been plotted here. The real part of the former and the imaginary parts of the latter two are zero. We consider maximal noncollinearity (θt=π/4\theta_t=\pi/4) and filling factor f=0.6f=0.6. All the correlations are normalized with respect to the magnitude of the ss-wave singlet correlation ∣Fjs∣|F_j^s|. The detailed parameters employed for the numerical evaluation are specified in Appendix § 3.1.

Concluding remarks

We have theoretically demonstrated the generation of all, including the spinful, spin-triplet Cooper pairs in a conventional superconductor by an adjacent canted antiferromagnet. Our proposal leverages the intrinsic noncollinearity between the two sublattice magnetizations in a canted antiferromagnet for applications in superconducting hybrids. This canting can be induced intrinsically by Dzyaloshinskii-Moriya interaction, such as in hematite [72]. Additionally, it can be induced and controlled using an applied magnetic field [73]. The resulting spin-triplets have a predominantly Néel character, i.e., their amplitude oscillates in space on the lattice length scale similar to the Néel spin order. While we have considered a lattice matched interface for concreteness and simplicity, the essential physics remains the same even in the presence of disorder with a gradual suppression of the Néel triplets with increasing disorder [49]. Furthermore, a metallic AF would lead to similar phenomena as discussed here, with the additional complication of Cooper pairs leaking into the AF. The superconducting critical temperature is more strongly suppressed by the interband Néel spin-triplets than by a ferromagnetic spin-splitting field of similar magnitude, thereby offering an experimental signature of their generation. Altogether, our analysis highlights the noncollinear nature of homogeneous canted antiferromagnets by employing them for generating spin-triplet Cooper pairs in a simple superconducting bilayer. This manner of generating noncollinearity using a homogeneous antiferromagnet is expected to find use in other phenomena that have traditionally relied on magnetic multilayers or spin textures.

Normal metal interfaced with an antiferromagnet

In order to understand when interband and intraband pairing is favored, we consider the normal-state wave functions of an AF/normal metal bilayer. The Hamiltonian for the conducting electrons of a normal metal interfaced with an insulating AF is modeled as [74]

H=−μ∑j,σcj,σ†cj,σ−t∑⟨i,j⟩∑σci,σ†cj,σ−J2∑jM⃗j⋅S⃗j,(11)H=-\mu\sum_{j,\sigma}c_{j,\sigma}^{\dagger}c_{j,\sigma}-t\sum_{\langle i,j\rangle}\sum_{\sigma}c_{i,\sigma}^{\dagger}c_{j,\sigma}-\frac{J}{2}\sum_j\vec{M}_j\cdot\vec{S}_j, \tag*{(11)}

where M⃗j=(−1)j+1z^\vec{M}_j=(-1)^{j+1}\hat{z} gives the magnetic texture of a collinear AF. Other symbols have the same meaning as in section II of the main text.

In order to calculate the eigenenergies and eigenvectors, the Hamiltonian is written in terms of creation and annihilation operators for electrons at sublattices A and B. Creation operators for electrons at sublattices A and B are defined as aj,σ†=c2j−1,σ†a_{j,\sigma}^{\dagger}=c_{2j-1,\sigma}^{\dagger} and bj,σ†=c2j,σ†b_{j,\sigma}^{\dagger}=c_{2j,\sigma}^{\dagger}, where j=1,2,…,N/2j=1,2,\ldots,N/2. Then, the basis is changed from Wannier wavefunctions to Bloch wavefunctions using the relation

aj,σ†=∑k∈FBZ1N/2e−ik(2j−2)aak,σ†,bj,σ†=∑k∈FBZ1N/2e−ik(2j−1)abk,σ†,(12)a_{j,\sigma}^{\dagger}=\sum_{k\in FBZ}\frac{1}{\sqrt{N/2}}e^{-ik(2j-2)a}a_{k,\sigma}^{\dagger},\qquad b_{j,\sigma}^{\dagger}=\sum_{k\in FBZ}\frac{1}{\sqrt{N/2}}e^{-ik(2j-1)a}b_{k,\sigma}^{\dagger}, \tag*{(12)}

where aa is the lattice constant, and kk is a reciprocal lattice vector in the first Brillouin zone (FBZ). Now, the Hamiltonian can be written in the form

H=∑k(ak,↑† ak,↓† bk,↑† bk,↓†)H(k)(ak,↑ ak,↓ bk,↑ bk,↓)T,(13)H=\sum_k\left(a_{k,\uparrow}^{\dagger}\ a_{k,\downarrow}^{\dagger}\ b_{k,\uparrow}^{\dagger}\ b_{k,\downarrow}^{\dagger}\right)H(k)\left(a_{k,\uparrow}\ a_{k,\downarrow}\ b_{k,\uparrow}\ b_{k,\downarrow}\right)^{T}, \tag*{(13)}

giving the dispersion relation

E±(k)=−μ±4t2cos⁡2(ka)+(J/2)2,(14)E_{\pm}(k)=-\mu\pm\sqrt{4t^2\cos^2(ka)+(J/2)^2}, \tag*{(14)}

where the two energy bands E±(k)E_{\pm}(k) are two-fold degenerate. The eigenvectors for energy band E−(k)E_{-}(k) are

ψ1(k)=N2(2tcos⁡(ka)0−J/2+4t2cos⁡2(ka)+(J/2)20)=(uAk↑(1)uAk↓(1)uBk↑(1)uBk↓(1)),and(15)\psi_1(k)=\mathcal{N}_2 \begin{pmatrix} 2t\cos(ka)\\ 0\\ -J/2+\sqrt{4t^2\cos^2(ka)+(J/2)^2}\\ 0 \end{pmatrix} = \begin{pmatrix} u_{Ak\uparrow}^{(1)}\\ u_{Ak\downarrow}^{(1)}\\ u_{Bk\uparrow}^{(1)}\\ u_{Bk\downarrow}^{(1)} \end{pmatrix}, \quad\text{and} \tag*{(15)}
ψ2(k)=N1(02tcos⁡(ka)0J/2+4t2cos⁡2(ka)+(J/2)2)=(uAk↑(2)uAk↓(2)uBk↑(2)uBk↓(2)).(16)\psi_2(k)=\mathcal{N}_1 \begin{pmatrix} 0\\ 2t\cos(ka)\\ 0\\ J/2+\sqrt{4t^2\cos^2(ka)+(J/2)^2} \end{pmatrix} = \begin{pmatrix} u_{Ak\uparrow}^{(2)}\\ u_{Ak\downarrow}^{(2)}\\ u_{Bk\uparrow}^{(2)}\\ u_{Bk\downarrow}^{(2)} \end{pmatrix}. \tag*{(16)}

Whereas the eigenvectors for the energy band E+(k)E_{+}(k) are

ψ3(k)=N1(2tcos⁡(ka)0−J/2−4t2cos⁡2(ka)+(J/2)20)=(uAk↑(3)uAk↓(3)uBk↑(3)uBk↓(3)),and(17)\psi_3(k)=\mathcal{N}_1 \begin{pmatrix} 2t\cos(ka)\\ 0\\ -J/2-\sqrt{4t^2\cos^2(ka)+(J/2)^2}\\ 0 \end{pmatrix} = \begin{pmatrix} u_{Ak\uparrow}^{(3)}\\ u_{Ak\downarrow}^{(3)}\\ u_{Bk\uparrow}^{(3)}\\ u_{Bk\downarrow}^{(3)} \end{pmatrix}, \quad\text{and} \tag*{(17)}
ψ4(k)=N2(02tcos⁡(ka)0J/2−4t2cos⁡2(ka)+(J/2)2)=(uAk↑(4)uAk↓(4)uBk↑(4)uBk↓(4)).(18)\psi_4(k)=\mathcal{N}_2 \begin{pmatrix} 0\\ 2t\cos(ka)\\ 0\\ J/2-\sqrt{4t^2\cos^2(ka)+(J/2)^2} \end{pmatrix} = \begin{pmatrix} u_{Ak\uparrow}^{(4)}\\ u_{Ak\downarrow}^{(4)}\\ u_{Bk\uparrow}^{(4)}\\ u_{Bk\downarrow}^{(4)} \end{pmatrix}. \tag*{(18)}

Here, N1\mathcal{N}_1 and N2\mathcal{N}_2 are the normalization factors of the eigenvectors. ψ1\psi_1 and ψ3\psi_3 correspond to the wavefunctions of spin ↑\uparrow electrons, whereas ψ2\psi_2 and ψ4\psi_4 correspond to spin ↓\downarrow electrons. Figure 7 shows a plot of the non-zero components of ψ1\psi_1 for positive kk values . Let us compare the case of J=0J=0 and J≠0J\ne0 to study how the states of a metal are modified when it is brought in contact with an AF.

Figure 7. Non-zero components of eigenfunction ψ1\psi_1 (Eq. (15)) of a normal metal/AF bilayer for different strengths of magnetic exchange interaction as a function of positive kk values in the first Brillouin zone.

From Figure 7, we see that the probabilities of finding a spin ↑\uparrow electron at sites of sublattice A and B are equal for J=0J=0. However, for J≠0J\ne0, the probability on A sublattice is more than that on B sublattice. It is important to note that the probability of finding spin ↑\uparrow electrons at sublattice A sites becomes 1 and that for sublattice B becomes 0 at k=π/2ak = \pi/2a. Similar asymmetries in the sublattices arise for states ψ2\psi_2, ψ3\psi_3, and ψ4\psi_4 as soon as we make JJ non-zero. For ψ2\psi_2, the probability of finding a spin ↓\downarrow electron at sublattice site B becomes more than that for sublattice A. For the states of E+(k)E_{+}(k) band (ψ3\psi_3 and ψ4\psi_4), the roles of sublattices A and B are interchanged with respect to spin.

Near the Brillouin zone boundary (BZB) of band E−(k)E_{-}(k) (i.e. ψ1\psi_1 and ψ2\psi_2), all the spin ↑\uparrow electrons get localized to the sites of sublattice A and all the spin ↓\downarrow electrons get localized to the sites of sublattice B. For the band E+(k)E_{+}(k), sublattices A and B interchange their roles and we find that spin ↑\uparrow electrons get localized at sublattice B and spin ↓\downarrow electrons get localized at sublattice A near the BZB. This property of the electronic states near the BZB is the reason why interband pairing is the dominant mechanism for the formation of on-site opposite-spin Cooper pairs in an AF/S bilayer when the Fermi level lies within the antiferromagnetic band gap E+−E−E_{+} - E_{-}. On the other hand, intraband pairing is the dominant mechanism when the Fermi level lies within one of the two bands.

Bogoliubov-de Gennes calculation

The Hamiltonian in Eq. (1) can be written as

H=−μN+∑j∣Δj∣2U+12∑i,jΨi†H~i,jΨj(19)H=-\mu N+\sum_j\frac{|\Delta_j|^2}{U}+\frac{1}{2}\sum_{i,j}\Psi_i^\dagger\widetilde{H}_{i,j}\Psi_j \tag*{(19)}

where Ψj†=(cj,↑† cj,↓† cj,↑ cj,↓)\Psi_j^\dagger=(c_{j,\uparrow}^\dagger\ c_{j,\downarrow}^\dagger\ c_{j,\uparrow}\ c_{j,\downarrow}) and NN is the total number of sites. The matrix H~\widetilde{H} is diagonalized by solving the BdG equations

∑jH~i,jϕj,n=Enϕi,n,(20)\sum_j\widetilde{H}_{i,j}\phi_{j,n}=E_n\phi_{i,n}, \tag*{(20)}

where

ϕi,n=(ui,n,↑ ui,n,↓ vi,n,↑ vi,n,↓)T,(21)\phi_{i,n}=\left(u_{i,n,\uparrow}\ u_{i,n,\downarrow}\ v_{i,n,\uparrow}\ v_{i,n,\downarrow}\right)^{T}, \tag*{(21)}

is the eigenvector and EnE_n the eigenenergies of H~\widetilde{H}. Now the Hamiltonian can be written as

H=−μN+∑j∣Δj∣2U−12∑~nEn+∑~nΓn†EnΓn,(22)H=-\mu N+\sum_j\frac{|\Delta_j|^2}{U}-\frac{1}{2}\widetilde{\sum}_n E_n+\widetilde{\sum}_n\Gamma_n^\dagger E_n\Gamma_n, \tag*{(22)}

where ∑~n\widetilde{\sum}_n represents sum over positive eigenenergies, and Γn\Gamma_n’s are Bogoliubov fermionic operators related to the old fermionic operators by cj,σ=∑~n(uj,n,σΓn+vj,n,σ∗Γn†)c_{j,\sigma}=\widetilde{\sum}_n(u_{j,n,\sigma}\Gamma_n+v_{j,n,\sigma}^{*}\Gamma_n^\dagger). The superconducting order parameter is calculated self-consistently using the relation

Δj=−U∑~n[uj,n,↓vj,n,↑∗[1−fFD(En)]+vj,n,↓∗uj,n,↑fFD(En)],(23)\Delta_j=-U\widetilde{\sum}_n\left[u_{j,n,\downarrow}v_{j,n,\uparrow}^{*}[1-f_{FD}(E_n)]+v_{j,n,\downarrow}^{*}u_{j,n,\uparrow}f_{FD}(E_n)\right], \tag*{(23)}

where fFD(En)=⟨Γn†Γn⟩=1/(eEn/kBT+1)f_{FD}(E_n)=\langle\Gamma_n^\dagger\Gamma_n\rangle=1/(e^{E_n/k_BT}+1) is the Fermi-Dirac distribution.

In order to calculate the spin-triplet correlations, we consider the anomalous Matsubara Green’s function Fjj,σσ′(τ)=−⟨Tτcj,σ(τ)cj,σ′(0)⟩F_{jj,\sigma\sigma^\prime}(\tau)=-\langle T_\tau c_{j,\sigma}(\tau)c_{j,\sigma^\prime}(0)\rangle, where τ=it~\tau=i\widetilde{t} is the imaginary time, t~\widetilde{t} is the time, and TτT_\tau is the ordering operator for τ\tau. Taking its Fourier transform, we get

Fjj,σσ′(iωl)=∫0βeiωlτFjj,σσ′(τ) dτ=∑~n[uj,n,σvj,n,σ′∗iωl−En/ℏ+vj,n,σ∗uj,n,σ′iωl+En/ℏ],(24)F_{jj,\sigma\sigma^\prime}(i\omega_l)=\int_0^\beta e^{i\omega_l\tau}F_{jj,\sigma\sigma^\prime}(\tau)\,d\tau=\widetilde{\sum}_n\left[\frac{u_{j,n,\sigma}v_{j,n,\sigma^\prime}^{*}}{i\omega_l-E_n/\hbar}+\frac{v_{j,n,\sigma}^{*}u_{j,n,\sigma^\prime}}{i\omega_l+E_n/\hbar}\right], \tag*{(24)}

where β=ℏ/kBT\beta=\hbar/k_BT, kBk_B is the Boltzmann constant, TT is the temperature, and ωl=(2l+1)π/β\omega_l=(2l+1)\pi/\beta is a Matsubara frequency for fermions with integer ll. This expression is used in calculating the correlations in Eq. (3) and their relevant components Eqs. (4)-(7).

Decay of triplet correlations with distance in 2-D

In section V, we discussed the spatial variation of triplet correlations in a 2-D superconductor interfaced with a canted-AF and observed that the Néel triplet correlations flip sign on adjacent sites to form a checkerboard pattern. In Figure 8, we study the decay of the ss-wave spin-triplet correlations with the distance from the AF/S interface. We consider half-filling (f=0.5f=0.5) to avoid the influence of Friedel oscillations. We observe that the magnitude of both the ferromagnetic and antiferromagnetic components of the correlations decay exponentially over a length scale determined by the coherence length. The rapid oscillations of the Néel order of the canted-AF over the atomic length scale thus only affects the sign of the triplet correlations.

Figure 8. The average magnitude of the ss-wave spin-triplet correlations normalized by the singlet correlation in each jyj_y layer parallel to the canted-AF/S interface (∣Ft,α/Fs∣jy, α=x,y)\left(\left|F^{t,\alpha}/F^s\right|_{j_y},\ \alpha=x,y\right) is plotted with respect to the distance from the interface. We consider half filling (f=0.5f=0.5) and maximal canting (θt=π/4\theta_t=\pi/4). The sites with index jy=1j_y=1 form the layer adjacent to the AF. Ignoring the edge effects, we show only the first 25 layers of a superconducting sheet with 102×30102 \times30 sites. The Ft,zF^{t,z} component is zero in all the layers. The detailed parameters employed for the numerical evaluation are specified in Appendix E.

Critical temperature of 2-D S/canted-AF

In section IV, we found that the critical temperature vs. canting angle curves of a 1-D superconductor show opposite trends at half-filling (f=0.5f=0.5) and away from half-filling (f=0.6f=0.6). In Figure 9, we show that TcT_c behaves similarly for a 2-D superconductor. The change in TcT_{\mathrm{c}} as we change the canting angle from the collinear AF case (θt=0\theta_{\mathrm{t}}=0) to the F case (θt=π/2\theta_{\mathrm{t}}=\pi/2) however decreases in magnitude. This is because with an increasing S thickness, the effective spin-splitting induced by the AF decreases.

Figure 9. The critical temperature TcT_{\mathrm{c}} vs. canting angle θt\theta_{\mathrm{t}} is plotted for a 2-D superconductor of size 302×3302 \times3 interfaced with a canted-AF at filling factors f=0.5f=0.5 and f=0.6f=0.6. Tc0T_{\mathrm{c}0} and ∣Δ‾0∣|\overline{\Delta}_{0}| are the critical temperature and the average superconducting order parameter at zero-temperature, respectively, in the absence of the magnetic layer. The detailed parameters employed for the numerical evaluation are specified in Appendix § 3.1.

Numerical parameters

The parameters used for each of the figures are

  • Figure 2: A 1-D superconductor with number of sites N=302N=302, hopping parameter t=10t=10, pair potential U/t=1U/t=1, magnetic exchange interaction strength J=0.018t(=1.09Δ0)J=0.018t (=1.09\Delta_{0}), canting angle θt=π/4\theta_{\mathrm{t}}=\pi/4, and filling factor f=0.6f=0.6 (μ/Δ0=37\mu/\Delta_{0}=37), at temperature T=0.1Tc=0.001t/kBT=0.1T_{\mathrm{c}}=0.001t/k_{\mathrm{B}} is taken to calculate ss-wave triplet correlations. Here, 2Δ02\Delta_{0} is the zero-temperature superconducting gap without the adjacent AF.

  • Figure 3: A 1-D superconductor with N=302N=302, t=10t=10, and U/t=1U/t=1 at kBT/t=0.001k_{\mathrm{B}}T/t=0.001 is taken for calculating correlations. For filling factor f=0.5f=0.5 (μ/Δ0=0\mu/\Delta_{0}=0),

J=0.012t=1.09Δ0J = 0.012t = 1.09\Delta_0 is taken, and for f=0.6f = 0.6 (μ/Δ0=37\mu/\Delta_0 = 37), J=0.018t=1.09Δ0J = 0.018t = 1.09\Delta_0, is taken. Here, 2Δ02\Delta_0 represents the superconducting gap in the absence of the canted-AF at zero-temperature.

  • Figure 4: A 1-D superconductor with N=302N = 302, t=10t = 10, U/t=1U/t = 1 is used for the calculation of TcT_c for filling factors f=0.5f = 0.5 (μ/Δ0=0\mu/\Delta_0 = 0) and f=0.6f = 0.6 (μ/Δ0=37\mu/\Delta_0 = 37). The critical temperature of the isolated superconductor Tc0T_{c0} for f=0.6f = 0.6 is 0.0101t/kB0.0101t/k_B and for f=0.5f = 0.5 is 0.0076t/kB0.0076t/k_B.

  • Figure 5: A 2-D superconducting sheet with Nx=102N_x = 102, Ny=14N_y = 14, t=10t = 10, J=0.1t=4.3∣Δ0∣J = 0.1t = 4.3\lvert\Delta_0\rvert, θt=π/4\theta_t = \pi/4, U/t=1U/t = 1, and f=0.6f = 0.6 (μ/∣Δ0∣=18.5\mu/\lvert\Delta_0\rvert= 18.5) at kBT/t=0.001k_BT/t = 0.001 is used to calculate the ss-wave triplet correlations. Here, ∣Δ0∣\lvert\Delta_0\rvert represents the average magnitude of the superconducting order parameter in absence of the canted-AF layer.

  • Figure 6: A 2-D superconductor with Nx=202N_x = 202, Ny=14N_y = 14, t=10t = 10, J=0.1t=4.3∣Δ0∣J = 0.1t = 4.3\lvert\Delta_0\rvert, θt=π/4\theta_t = \pi/4, U/t=1U/t = 1, and f=0.6f = 0.6 (μ/∣Δ0∣=18\mu/\lvert\Delta_0\rvert= 18) at kBT/t=0.001k_BT/t = 0.001 is used to calculate the pyp_y-wave triplet correlations. Here, ∣Δ0∣\lvert\Delta_0\rvert represents the average magnitude of the superconducting order parameter in absence of the AF.

  • Figure 8: A 2-D superconductor with Nx=102N_x = 102, Ny=30N_y = 30, t=10t = 10, J=0.1t=1.7∣Δ0∣J = 0.1t = 1.7\lvert\Delta_0\rvert, θt=π/4\theta_t = \pi/4, U/t=1U/t = 1, and f=0.5f = 0.5 (μ=0\mu= 0) at kBT/t=0.001k_BT/t = 0.001 is used to calculate the ss-wave triplet correlations, where ∣Δ0∣\lvert\Delta_0\rvert represents the average magnitude of the superconducting order parameter in absence of the AF.

  • Figure 9: A 2-D superconductor with Nx=302N_x = 302, Ny=3N_y = 3, t=10t = 10, and U/t=1U/t = 1 is used to plot TcT_c vs. θt\theta_t. For f=0.5f = 0.5, J=0.047t=1.34∣Δ0∣=2.4Tc0J = 0.047t = 1.34\lvert\Delta_0\rvert= 2.4T_{c0} is taken, and for f=0.6f = 0.6, J=0.07t=1.35∣Δ0∣=2.4Tc0J = 0.07t = 1.35\lvert\Delta_0\rvert= 2.4T_{c0} is taken, where Tc0T_{c0} and ∣Δ0∣\lvert\Delta_0\rvert are the critical temperature and the average magnitude of the superconducting order parameter at zero temperature of the same superconductor without the AF layer.

Canted-AF/S with rotated magnetic moments

To compare the results of this article with [49], one needs to rotate the magnetic moments of the Hamiltonian [Eq. (1)] to $\vec{M}_j = [(-1)^{j+1}\cos\theta_t\hat{z} + \sin\theta_t\hat{x}]$ (see Fig. 10(d)). The correlations for this rotated system is plotted in {xref}Fig. 10`. Here, Ft,zF^{t,z} is the Néel triplet

Figure 10. Spatial variation of normalized triplet correlations for 10 lattice sites considering θt=π/4\theta_t = \pi/4 and μ/Δ0=37\mu/\Delta_0 = 37, corresponding to filling factor f=0.6f = 0.6. We show the imaginary part of the triplet correlations Fjt,zF_j^{t,z} [panel (a)] and Fjt,xF_j^{t,x} [panel (b)], and the real part of the triplet correlation Fjt,yF_j^{t,y} [panel (c)]. The real part of the two former and the imaginary part of the latter are zero. All the correlations are normalized by the spatially averaged magnitude of the singlet correlation ∣Fs‾∣|\overline{F^s}|. (d) Orientation of the magnetic moments of sublattices A and B of the AF in the rotated configuration. Fjt,zF_j^{t,z} correlation coming from the antiferromagnetic component of the canted-AF, Fjt,xF_j^{t,x} comes from the ferromagnetic component, and Fjt,yF_j^{t,y} comes from the noncollinearity in the canted-AF.

Acknowledgments

SC and AK acknowledge financial support from the Spanish Ministry for Science and Innovation – AEI Grant CEX2018-000805-M (through the “Maria de Maeztu” Programme for Units of Excellence in R&D). LJK acknowledges support from the Research Council of Norway through its Centres of Excellence funding scheme, Project No. 262633 ‘QuSpin’. IVB acknowledges support from MIPT, Project FSMG-2023-0014.

References

References

  1. [3]M. Sigrist and K. Ueda, Phenomenological theory of unconventional superconductivity, Rev. Mod. Phys. 63, 239 (1991).DOI
  2. [4]M. Eschrig, Spin-polarized supercurrents for spintronics: a review of current progress, Reports On Progress In Physics 78, 104501 (2015).DOI
  3. [5]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
  4. [6]F. S. Bergeret, M. Silaev, P. Virtanen, and T. T. Heikkilä, Colloquium: Nonequilibrium effects in superconductors with a spin-splitting field, Rev. Mod. Phys. 90, 041001 (2018).DOI
  5. [7]A. I. Buzdin, Proximity effects in superconductor-ferromagnet heterostructures, Rev. Mod. Phys. 77, 935 (2005).DOI
  6. [8]J. Linder and J. W. A. Robinson, Superconducting spintronics, Nature Physics 11, 307 (2015).DOI
  7. [9]G. Yang, C. Ciccarelli, and J. W. A. Robinson, Boosting spintronics with superconductivity, APL Materials 9, 050703 (2021).DOI
  8. [10]I. V. Bobkova, A. M. Bobkov, and M. A. Silaev, Spin torques and magnetic texture dynamics driven by the supercurrent in superconductor/ferromagnet structures, Phys. Rev. B 98, 014521 (2018).DOI
  9. [11]X. Waintal and P. W. Brouwer, Magnetic exchange interaction induced by a josephson current, Phys. Rev. B 65, 054407 (2002).DOI
  10. [12]K. Halterman and M. Alidoust, Josephson currents and spin-transfer torques in ballistic SFSFS nanojunctions, Superconductor Science and Technology 29, 055007 (2016).DOI
  11. [13]J. Linder and T. Yokoyama, Supercurrent-induced magnetization dynamics in a josephson junction with two misaligned ferromagnetic layers, Phys. Rev. B 83, 012501 (2011).DOI
  12. [14]L. Fu and C. L. Kane, Superconducting proximity effect and majorana fermions at the surface of a topological insulator, Phys. Rev. Lett. 100, 096407 (2008).
  13. [15]Y. Tanaka, M. Sato, and N. Nagaosa, Symmetry and topology in superconductors –odd-frequency pairing and edge states–, Journal of the Physical Society of Japan 81, 011013 (2012), https://doi.org/10.1143/JPSJ.81.011013.
  14. [16]M. Eschrig and T. Löfwander, Triplet supercurrents in clean and disordered half-metallic ferromagnets, Nature Physics 4, 138 (2008).DOI
  15. [17]R. S. Keizer, S. T. B. Goennenwein, T. M. Klapwijk, G. Miao, G. Xiao, and A. Gupta, A spin triplet supercurrent through the half-metallic ferromagnet CrO2, Nature 439, 825 (2006).DOI
  16. [18]T. S. Khaire, M. A. Khasawneh, W. P. Pratt, and N. O. Birge, Observation of spin-triplet superconductivity in co-based josephson junctions, Phys. Rev. Lett. 104, 137002 (2010).pubmed.ncbi.nlm.nih.gov/20481906
  17. [19]K.-R. Jeon, C. Ciccarelli, A. J. Ferguson, H. Kurebayashi, L. F. Cohen, X. Montiel, M. Eschrig, J. W. A. Robinson, and M. G. Blamire, Enhanced spin pumping into superconductors provides evidence for superconducting pure spin currents, Nature Materials 17, 499 (2018).DOI
  18. [20]K.-R. Jeon, X. Montiel, S. Komori, C. Ciccarelli, J. Haigh, H. Kurebayashi, L. F. Cohen, A. K. Chan, K. D. Stenning, C.-M. Lee, M. Eschrig, M. G. Blamire, and J. W. A. Robinson, Tunable pure spin supercurrents and the demonstration of their gateability in a spin-wave device, Phys. Rev. X 10, 031020 (2020).
  19. [21]S. Diesch, P. Machon, M. Wolz, C. Sürgers, D. Beckmann, W. Belzig, and E. Scheer, Creation of equal-spin triplet superconductivity at the Al/EuS interface, Nature Communications 9, 5248 (2018).DOI
  20. [22]A. Costa, M. Sutula, V. Lauter, J. Song, J. Fabian, and J. S. Moodera, Signatures of superconducting triplet pairing in Ni–Ga-bilayer junctions, New Journal of Physics 24, 033046 (2022).DOI
  21. [23]A. Hijano, V. N. Golovach, and F. S. Bergeret, Quasiparticle density of states and triplet correlations in superconductor/ferromagnetic-insulator structures across a sharp domain wall, Phys. Rev. B 105, 174507 (2022).DOI
  22. [24]J. W. A. Robinson, J. D. S. Witt, and M. G. Blamire, Controlled injection of spin-triplet supercurrents into a strong ferromagnet, Science 329, 59 (2010).DOI
  23. [25]F. Chiodi, J. D. S. Witt, R. G. J. Smits, L. Qu, G. B. Halász, C.-T. Wu, O. T. Valls, K. Halterman, J. W. A. Robinson, and M. G. Blamire, Supra-oscillatory critical temperature dependence of Nb-Ho bilayers, EPL (Europhysics Letters) 101, 37002 (2013).DOI
  24. [26]C. González-Ruano, L. G. Johnsen, D. Caso, C. Tiusan, M. Hehn, N. Banerjee, J. Linder, and F. G. Aliev, Superconductivity-induced change in magnetic anisotropy in epitaxial ferromagnet-superconductor hybrids with spin-orbit interaction, Phys. Rev. B 102, 020405 (2020).
  25. [27]C.-T. Wu, O. T. Valls, and K. Halterman, Reentrant superconducting phase in conical-ferromagnet–superconductor nanostructures, Phys. Rev. Lett. 108, 117005 (2012).pubmed.ncbi.nlm.nih.gov/22540503
  26. [28]N. Banerjee, J. A. Ouassou, Y. Zhu, N. A. Stelmas henko, J. Linder, and M. G. Blamire, Controlling the superconducting transition by spin-orbit coupling, Phys. Rev. B 97, 184521 (2018).DOI
  27. [29]L. G. Johnsen, N. Banerjee, and J. Linder, Magnetization reorientation due to the superconducting transition in heavy-metal heterostructures, Phys. Rev. B 99, 134516 (2019).DOI
  28. [30]S. H. Jacobsen, J. A. Ouassou, and J. Linder, Critical temperature and tunneling spectroscopy of superconductor-ferromagnet hybrids with intrinsic rashba-dresselhaus spin-orbit coupling, Phys. Rev. B 92, 024510 (2015).arxiv.org/abs/1503.06835
  29. [31]S. Tamura, Y. Tanaka, and T. Yokoyama, Generation of polarized spin-triplet cooper pairings by magnetic barriers in superconducting junctions, Phys. Rev. B 107, 054501 (2023).DOI
  30. [32]P. Dutta and A. M. Black-Schaffer, Signature of odd-frequency equal-spin triplet pairing in the josephson current on the surface of weyl nodal loop semimetals, Phys. Rev. B 100, 104511 (2019).DOI
  31. [33]P. V. Leksin, N. N. Garif’yanov, I. A. Garifullin, Y. V. Fominov, J. Schumann, Y. Krupskaya, V. Kataev, O. G. Schmidt, and B. Büchner, Evidence for triplet superconductivity in a superconductor-ferromagnet spin valve, Phys. Rev. Lett. 109, 057005 (2012).DOI
  32. [34]A. Buzdin and M. Kupriyanov, Transition-temperature of a superconductor-ferromagnet superlattice, JETP Letters 52, 487 (1990).
  33. [35]Z. Radović, M. Ledvij, L. Dobrosavljević-Grujić, A. I. Buzdin, and J. R. Clem, Transition temperatures of superconductor-ferromagnet superlattices, Phys. Rev. B 44, 759 (1991).pubmed.ncbi.nlm.nih.gov/9999178
  34. [36]J. J. Hauser, H. C. Theuerer, and N. R. Werthamer, Proximity effects between superconducting and magnetic films, Phys. Rev. 142, 118 (1966).DOI
  35. [37]M. Hübener, D. Tikhonov, I. A. Garifullin, K. Westerholt, and H. Zabel, The antiferromagnet/superconductor proximity effect in Cr/V/Cr trilayers, Journal Of Physics-Condensed Matter 14, 8687 (2002).
  36. [38]C. Bell, E. J. Tarte, G. Burnell, C. W. Leung, D.-J. Kang, and M. G. Blamire, Proximity and josephson effects in superconductor/antiferromagnetic Nb/γ−Fe50Mn50 heterostructures, Phys. Rev. B 68, 144517 (2003).DOI
  37. [39]B. L. Wu, Y. M. Yang, Z. B. Guo, Y. H. Wu, and J. J. Qiu, Suppression of superconductivity in Nb by IrMn in IrMn/Nb bilayers, Applied Physics Letters 103, 152602 (2013).DOI
  38. [40]R. L. Seeger, G. Forestier, O. Gladii, M. Leiviskä, S. Auffret, I. Joumard, C. Gomez, M. Rubio-Roy, A. I. Buzdin, M. Houzet, and V. Baltz, Penetration depth of cooper pairs in the IrMn antiferromagnet, Phys. Rev. B 104, 054413 (2021).DOI
  39. [41]A. Mani, T. G. Kumary, D. Hsu, J. G. Lin, and C.-H. Chern, Modulation of superconductivity by spin canting in a hybrid antiferromagnet/superconductor oxide, Applied Physics Letters 94, 072509 (2009).DOI
  40. [42]A. Mani, T. G. Kumary, and J. G. Lin, Thickness controlled proximity effects in c-type antiferromagnet/superconductor heterostructure, Scientific Reports 5, 12780 (2015).DOI
  41. [43]I. V. Bobkova, P. J. Hirschfeld, and Y. S. Barash, Spin-dependent quasiparticle reflection and bound states at interfaces with itinerant antiferromagnets, Phys. Rev. Lett. 94, 037005 (2005).pubmed.ncbi.nlm.nih.gov/15698311
  42. [44]B. M. Andersen, I. V. Bobkova, P. J. Hirschfeld, and Y. S. Barash, 0 − π transitions in josephson junctions with antiferromagnetic interlayers, Phys. Rev. Lett. 96, 117005 (2006).pubmed.ncbi.nlm.nih.gov/16605855
  43. [45]B. M. Andersen, I. V. Bobkova, P. J. Hirschfeld, and Y. S. Barash, Bound states at the interface between antiferromagnets and superconductors, Phys. Rev. B 72, 184510 (2005).DOI
  44. [46]M. F. Jakobsen, K. B. Naess, P. Dutta, A. Brataas, and A. Qaiumzadeh, Electrical and thermal transport in antiferromagnet-superconductor junctions, Phys. Rev. B 102, 140504 (2020).DOI
  45. [47]A. Kamra, A. Rezaei, and W. Belzig, Spin splitting induced in a superconductor by an antiferromagnetic insulator, Phys. Rev. Lett. 121, 247702 (2018).DOI
  46. [48]L. G. Johnsen, S. H. Jacobsen, and J. Linder, Magnetic control of superconducting heterostructures using compensated antiferromagnets, Phys. Rev. B 103, L060505 (2021).DOI
  47. [49]G. A. Bobkov, I. V. Bobkova, A. M. Bobkov, and A. Kamra, Néel proximity effect at antiferromagnet/superconductor interfaces, Phys. Rev. B 106, 144512 (2022).DOI
  48. [50]V. Baltz, A. Manchon, M. Tsoi, T. Moriyama, T. Ono, and Y. Tserkovnyak, Antiferromagnetic spintronics, Rev. Mod. Phys. 90, 015005 (2018).DOI
  49. [51]E. H. Fyhn, A. Brataas, A. Qaiumzadeh, and J. Linder, Superconducting proximity effect and long-ranged triplets in dirty metallic antiferromagnets (2022), arXiv:2210.09325 [cond-mat.supr-con].DOI
  50. [52]S. I. Pekar and E. I. Rashba, Combined Resonance in Crystals in Inhomogeneous Magnetic Fields, Zh. Eksperim. Teor. Fiz. 47, 1927 (1964).
  51. [53]K.-R. Jeon, B. K. Hazra, J.-K. Kim, J.-C. Jeon, H. Han, H. L. Meyerheim, T. Kontos, A. Cottet, and S. S. P. Parkin, Chiral antiferromagnetic josephson junctions as spin-triplet supercurrent spin valves and d.c. squids, Nature Nanotechnology 10.1038/s41565-023-01336-z (2023).DOI
  52. [54]J. Zhu, Bogoliubov-de Gennes Method and Its Applications, Lecture Notes in Physics (Springer International Publishing, 2016).DOI
  53. [55]G. Tang, C. Bruder, and W. Belzig, Magnetic field-induced “mirage” gap in an ising superconductor, Phys. Rev. Lett. 126, 237001 (2021).DOI
  54. [56]A. Cottet, D. Huertas-Hernando, W. Belzig, and Y. V. Nazarov, Spin-dependent boundary conditions for isotropic superconducting green’s functions, Phys. Rev. B 80, 184511 (2009).DOI
  55. [57]M. Eschrig, A. Cottet, W. Belzig, and J. Linder, General boundary conditions for quasiclassical theory of superconductivity in the diffusive limit: application to strongly spin-polarized systems, New Journal of Physics 17, 083037 (2015).
  56. [58]M. Kiwi, Exchange bias theory, Journal of Magnetism and Magnetic materials 234, 584 (2001).
  57. [59]K. Maki and T. Tsuneto, Pauli paramagnetism and superconducting state, Progress of Theoretical Physics 31, 945 (1964).DOI
  58. [60]N. Kopnin, Theory of Nonequilibrium Superconductivity, International Series of Monographs (Clarendon Press, 2001).DOI
  59. [61]W. Belzig, F. K. Wilhelm, C. Bruder, G. Schön, and A. D. Zaikin, Quasiclassical green’s function approach to mesoscopic superconductivity, Superlattices and Microstructures 25, 1251 (1999).DOI
  60. [62]M. Eschrig, T. Löfwander, T. Champel, J. C. Cuevas, J. Kopu, and G. Schön, Symmetries of pairing correlations in superconductor–ferromagnet nanostructures, Journal of Low Temperature Physics 147, 457 (2007).
  61. [63]T. Champel and M. Eschrig, Effect of an inhomogeneous exchange field on the proximity effect in disordered superconductor-ferromagnet hybrid structures, Phys. Rev. B 72, 054523 (2005).DOI
  62. [64]S. Takei, B. I. Halperin, A. Yacoby, and Y. Tserkovnyak, Superfluid spin transport through antiferromagnetic insulators, Phys. Rev. B 90, 094408 (2014).DOI
  63. [65]R. Cheng, J. Xiao, Q. Niu, and A. Brataas, Spin pumping and spin-transfer torques in antiferromagnets, Phys. Rev. Lett. 113, 057601 (2014).DOI
  64. [66]A. Kamra and W. Belzig, Spin pumping and shot noise in ferrimagnets: Bridging ferro- and antiferromagnets, Phys. Rev. Lett. 119, 197201 (2017).DOI
  65. [67]Q. Liu, H. Y. Yuan, K. Xia, and Z. Yuan, Mode-dependent damping in metallic antiferromagnets due to intersublattice spin pumping, Phys. Rev. Mater. 1, 061401 (2017).DOI
  66. [68]B. S. Chandrasekhar, A note on the maximum critical field of high-field superconductors, Applied Physics Letters 1, 7 (1962).DOI
  67. [69]A. M. Clogston, Upper limit for the critical field in hard superconductors, Phys. Rev. Lett. 9, 266 (1962).DOI
  68. [70]L. A. B. Olde Olthof, L. G. Johnsen, J. W. A. Robinson, and J. Linder, Controllable enhancement of p-wave superconductivity via magnetic coupling to a conventional superconductor, Phys. Rev. Lett. 127, 267001 (2021).DOI
  69. [71]As the p_y-wave triplet correlations are not defined at the edges, we plot these starting at the second row of lattice sites from the edge.
  70. [72]A. H. Morrish, Canted Antiferromagnetism: Hematite (WORLD SCIENTIFIC, 1995).
  71. [73]T. Wimmer, A. Kamra, J. Gü ck elhorn, M. Opel, S. Gepr ägs, R. Gross, H. Huebl, and M. Althammer, Observation of antiferromagnetic magnon pseudospin dynamics and the hanle effect, Phys. Rev. Lett. 125, 247204 (2020).DOI
  72. [74]H. T. Simensen, A. Kamra, R. E. Troncoso, and A. Brataas, Magnon decay theory of gilbert damping in metallic antiferromagnets, Phys. Rev. B 101, 020403 (2020).DOI

Paper details

vlnc:W_teqszp7j
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

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