Abstract
Spin-triplet supercurrent spin valves are of practical importance for the realization of superconducting spintronic logic circuits. In ferromagnetic Josephson junctions, the magnetic-field-controlled non-collinearity between the spin-mixer and spin-rotator magnetizations switches the spin-polarized triplet supercurrents on and off. Here we report an antiferromagnetic equivalent of such spin-triplet supercurrent spin valves in chiral antiferromagnetic Josephson junctions as well as a direct-current superconducting quantum interference device. We employ the topological chiral antiferromagnet Mn3Ge, in which the Berry curvature of the band structure produces fictitious magnetic fields, and the non-collinear atomic-scale spin arrangement accommodates triplet Cooper pairing over long distances (>150 nm). We theoretically verify the observed supercurrent spin-valve behaviours under a small magnetic field of <2 mT for current-biased junctions and the direct-current superconducting quantum interference device functionality. Our calculations reproduce the observed hysteretic field interference of the Josephson critical current and link these to the magnetic-field-modulated antiferromagnetic texture that alters the Berry curvature. Our work employs band topology to control the pairing amplitude of spin-triplet Cooper pairs in a single chiral antiferromagnet.
Similar content being viewed by others
Main
Intensive studies in coupling superconducting condensate state with magnetic-exchange spin splitting have opened up a research field of superconducting spintronics1,2,3, which promises to realize dissipationless spin-based logic and memory technologies. Of particular relevance in this research is the theoretical prediction4,5,6,7 and experimental verification8,9,10,11,12 that spin-polarized triplet pairing states can be created via spin-mixing and spin-rotation processes (for example, non-collinear exchange fields4,5,6,8,9,10 in real space and/or spin–orbit fields7,11,12 in reciprocal/k space) at proximity-engineered superconductor/ferromagnet (FM) interfaces1,2,3,4,5,6,7,8,9,10,11,12. With these advances, the field of superconducting spintronics involving spin-polarized triplet Cooper pairs1,2,3 can answer two practical aspects: how to efficiently generate such triplet pairs and how to tune them in a controllable manner. Yet, as previously pointed out4,13,14, fulfilling these two requirements at the same time seems challenging because the preconfigured robust non-collinearity of spin-mixer and spin-rotator magnetizations for a higher singlet-to-triplet pair conversion makes it difficult to control by an external stimulus.
The active and reversible control of spin-polarized triplet supercurrents has so far been mostly achieved in ferromagnetic Josephson junctions (JJs)13,14, where at least three FMs constitute a Josephson barrier whose relative magnetization directions, and therefore the non-collinearity, can be controlled by external magnetic fields4,13,14. This has led to the so-called spin-triplet supercurrent spin valve4,13,14, in which the proximity-created spin-polarized triplet supercurrents can be switched on and off. However, the fabrication of such ferromagnetic spin-triplet JJs13,14 requires delicate interface engineering, for instance, electronic energy band matching between neighbouring layers, selectivity in coercive fields of the spin-mixer and spin-rotator FMs and circumventing the out-of-plane (OOP) component of stray magnetic fields.
In this work, we demonstrate an antiferromagnetic analogue of the spin-triplet supercurrent spin-valve effect4,13,14 via the use of a single topological chiral antiferromagnet (AFM) Mn3Ge (refs. 15,16,17), which—with its lack of stray fields—can be highly advantageous for developing superconducting spintronic logic circuits1. The noteworthy aspect of Mn3Ge is that its non-collinear triangular antiferromagnetic spin arrangement15,16,17 in real space (Fig. 1a,b) and the fictitious magnetic fields derived from Berry curvature18,19 in k space, which are robust to low temperature T (refs. 15,16,17,20), facilitate the spin-mixing and spin-rotation processes1,2,3,4,5,6,7,8,9,10,11,12 required for singlet-to-triplet pair conversion. This enables, as shown in our recent experiment20, the proximity generation of long-range triplet supercurrents through the single chiral antiferromagnetic Josephson barrier.
Chiral antiferromagnetic spin-triplet JJs
Our focus of the present study is on the Berry-curvature-driven fictitious fields18,19 that play an effective role in converting spin-unpolarized singlet Cooper pairs (S = 0) to form long-range triplets (S = 1) in the topological chiral AFM20. Note that how non-collinear six spins on a kagome bilayer15,16,17, constituting a cluster magnetic octupole, are arranged in real space (equivalently, how time-reversal symmetry is broken) determines the Berry curvature profile18,19 in k space. So, an external magnetic field μ0H⊥ applied perpendicular to the kagome plane tilts the overall antiferromagnetic spin arrangement to a certain extent to the field direction (Fig. 1a,b) and subsequently changes the associated Berry curvature18 around the Fermi energy and the resulting fictitious fields19, as theoretically calculated21,22. This offers, as shown below, a radically different approach to control the pair amplitude of triplets by applying an extremely small magnetic field (<2 mT). Our k-space Berry curvature approach is conceptually different from very recent works on controlling spin-triplet critical currents in a single FM with magnetic vortex23 and domain wall24, both of which focus on the real-space magnetic texturing across the FM Josephson barrier.
We carry out proof-of-concept experiments based on Nb/Mn3Ge/Nb lateral JJs20 (Fig. 1c). In particular, the edge-to-edge separation distance ds of the adjacent superconducting Nb electrodes through an epitaxial thin film of the triangular chiral AFM Mn3Ge (Fig. 1a,b) is chosen to be comparable with or larger than the characteristic decay length \(\xi _\mathrm{{triplet}}^{{\mathrm{Mn}}_3{\mathrm{Ge}}}\) = 157–178 nm (at T = 2 K; Extended Data Fig. 1) of spin-polarized triplet supercurrents20. Conventional wisdom is that the spin-unpolarized singlets (S = 0) are mostly exchange field filtered within a few nanometres1,2,3,4,5,6,7,8,9,10,11,12 and the surviving spin-polarized triplets (S = 1, ms = ±1), which are immune to magnetic exchange fields1,2,3,4,5,6,7,8,9,10,11,12,25, finally mediate the long-range Josephson coupling. The mechanism for producing triplet Cooper pairs is a priori different in our case and we call this chiral antiferromagnetic spin-triplet JJs.
Hysteretic out-of-kagome-plane magnetic-field interference patterns
Figure 1c,d shows the typical magnetic-field interference pattern of Josephson critical current Ic(μ0H⊥) for the ds = 199 nm JJ at T = 2 K. Here μ0H⊥ is applied along [0001] and thus perpendicular to the kagome plane of our single-phase hexagonal D019-Mn3Ge(0001) layer (Fig. 1a,b and Methods). There exist two distinctively different features from the Ic(μ0H⊥) interference pattern of our prior ds ≤ 115 nm JJs20. First, the zero-order maximum of Ic appears 0.5–1.0 mT away from the zero field (μ0H⊥ = 0) and it is clearly hysteretic (Fig. 1e). As our single-phase D019-Mn3Ge(0001) has a vanishingly small spontaneous magnetization (≤11 emu c.c.–1 at 2 K)20 in the kagome plane15,16,17,20, we ascribe this hysteretic Ic(μ0H⊥) to the OOP-magnetic-field-modulated Berry curvature, as discussed later. Second, we obtain the characteristic Ic(μ0H⊥) oscillation with clear minima for the ds = 199 nm JJ (Fig. 1d), indicating the transverse uniformity of Ic across the whole Mn3Ge barrier and its coherent spatial quantum interference26. This improved magnetic-field interference in a longer JJ is probably due to the reduced effective edge roughness (several nanometres) of Nb electrodes relative to ds, given that the single crystallinity and surface morphology of previous20 and current D019-Mn3Ge(0001) layers are not fundamentally different (Extended Data Fig. 2).
Our theory, considering a chirality-dependent phase \(Qwd_\mathrm{s}\frac{{J\delta M}}{{\hbar v_\mathrm{F}}}\tau\) arising from the antiferromagnetic spin texture of Mn3Ge (Supplementary Section 1 provides the full details), anticipates the unique hysteretic Fraunhofer pattern and reproduces the overall Ic(μ0H⊥) data (Fig. 1d, solid lines):
Here Q, δM and J are the inverse antiferromagnetic domain size, amplitude of the inhomogeneous part and exchange interaction of the antiferromagnetic spin texture of Mn3Ge, respectively; τ is the chirality index (±1); γ represents the transparency at the Nb/Mn3Ge interface; and χ describes the chirality dependence of the Mn3Ge barrier. Here ħ is the reduced Planck constant and vF is the Fermi velocity of Mn3Ge. Also, \({{\varPhi }}_{\mathrm{JJ}} = \mu {_0}H_ \bot A_{\mathrm{JJ}}^{\mathrm{eff}}\) and \(A_{\mathrm{JJ}}^{\mathrm{eff}}\) = (2λL + ds)w is the effective junction area of magnetic flux penetration (Fig. 1d, bottom inset), λL is the London penetration depth (130 nm at 2 K)27 of 50-nm-thick Nb electrodes, w is the width of the Mn3Ge barrier and \({{\varPhi }}_0 = \frac{h}{{2e}}\) = 2.07 × 10−15 T m2 is the magnetic flux quantum. From theoretical reproduction, we get w ≈ 1.1 µm, close to the actual width of our JJ (Fig. 1c), and \(\left| {Qwd_{\mathrm{s}}\frac{{J\delta M}}{{\hbar v_{\mathrm{F}}}}\tau } \right|\) = 0.2 (Supplementary Section 2 provides a quantitative analysis).
Spin-triplet supercurrent spin valves
We now measure the time-averaged voltage V as a function of μ0H⊥ for the d.c. current I-biased JJs with ds = 28–199 nm (Fig. 2a–d), from which, especially at I ≈ Ic(μ0H⊥ = 0), one can straightforwardly see how the supercurrent spin-valve behaviour evolves as a function of ds. All the JJs in the superconducting state (T = 2 K) reveal asymmetric V(μ0H⊥) curves with respect to μ0H⊥ = 0 and their asymmetry is rigorously inverted when reversing the μ0H⊥ sweep direction. Note that since this asymmetric hysteretic behaviour disappears when the junctions are in the normal state (T = 8 K), it is necessarily connected to superconductivity induced in the chiral AFM leading to the Josephson supercurrent. With increasing ds, the centre-to-centre offset Δμ0H⊥ between the sweep-up and sweep-down V(μ0H⊥) curves progressively broaden from 0.1 to 1.5 mT and the consequent asymmetric hysteresis becomes more evident. On reaching ds = 199 nm (Fig. 2d), comparable with or larger than \(\xi _{\mathrm{triplet}}^{{\mathrm{Mn}}_3{\mathrm{Ge}}}\) (157–178 nm; Extended Data Fig. 1) over which the proximity-created spin triplets can primarily mediate the long-range Josephson coupling1,2,3,4,5,6,7,8,9,10,11,12, the complete antiferromagnetic analogue of the spin-triplet supercurrent spin valve4,13,14 is established in our chiral antiferromagnetic JJ. The applied μ0H⊥ ≤ 2.0 mT here to turn the Josephson supercurrent on and off (Fig. 2d) is intriguingly one order of magnitude smaller than that typically required for ferromagnetic spin-triplet JJs13,14, where one should apply a magnetic field larger than the coercive field of the free spin-rotator FM2 (for example, a few tens of millitesla even for soft FM Ni8Fe2)13 to change its magnetization direction relative to the pinned spin-mixer FM1, providing a beneficial route for controlling the pair amplitude of the spin triplets.
Supercurrent spin valve and 0-to-π phase shift in a d.c. SQUID
By taking advantage of this low-field spin-triplet supercurrent spin valve, we next fabricate a direct-current superconducting quantum interference device (d.c. SQUID) to showcase its potential as an on-chip local probe28 of out-of-kagome-plane magnetic moments of chiral AFMs with high sensitivity. Note that because the active superconducting loop of our SQUID (Fig. 3a,b and Methods) contains two chiral antiferromagnetic spin-triplet JJs of ds = 172 and 179 nm (≥\(\xi _{\mathrm{triplet}}^{{\mathrm{Mn}}_3{\mathrm{Ge}}}\)) that are laterally connected through the single layer of D019-Mn3Ge(0001), the SQUID action of this device is available only when the superconducting Nb electrodes are Josephson coupled via spin-triplet Cooper pairs24,29. From the zero-field current–voltage (I–V) curve of the fabricated SQUID (Fig. 3c), we find that the total critical current \(I_\mathrm{c}^{\mathrm{tot}}\) is approximately twice the Ic value of a single JJ with similar ds (Extended Data Fig. 1). This matches the standard theory26 of a d.c. SQUID comprising two overdamped JJs20,26 with a low resistance–capacitance product, that is,
Here we assume a small self-inductance of the SQUID loop for simplicity and consider the low-field regime (ΦJJ ≪ Φ0) such that the \(I_\mathrm{c}^{\mathrm{tot}}\left( {\mu _0H_ \bot } \right)\) curve mostly reflects the SQUID characteristics. Here Ic1 (Ic2) and φ1 (φ2) are the zero-field Josephson critical current and intrinsic phase difference30 for the first (second) JJ of the SQUID, respectively. Also, \(\varPhi _{\mathrm{SQUID}} = \mu _0H_ \bot A_{\mathrm{SQUID}}^{\mathrm{eff}}\) is the magnetic flux threading the SQUID loop given by μ0H⊥ and \(A_{\mathrm{SQUID}}^{\mathrm{eff}}\) = (2λL + Lx)(2λL + Ly) is the effective SQUID area (Fig. 3b). Note that for μ0H⊥ = 0, \(I_\mathrm{c}^{\mathrm{tot}} = I_{\mathrm{c}1} + I_{\mathrm{c}2}\).
Most importantly, the low-field supercurrent spin-valve functionality, that is, the active modulation of Josephson coupling strength as well as the ground-state phase difference, is successfully implemented in the SQUID oscillation (Fig. 3d–f). To the best of our knowledge, only a recent work utilizing multiple FMs has succeeded in the controllable switching between 0- and π-phase states of the ferromagnetic spin-triplet JJs embedded in a d.c. SQUID29. As summarized in Fig. 3g, the implemented spin-valve amplitude ΔR = Rhigh – Rlow increases with increasing I and achieves the maximum value at \(I \approx I_\mathrm{c}^{\mathrm{tot}}\left( {\mu _0H_ \bot = 0} \right)\), followed by a strong drop for larger I. Especially for \(I \lesssim I_\mathrm{c}^{\mathrm{tot}}\left( {\mu _0H_ \bot = 0} \right)\), we obtain an infinite spin-valve magnetoresistance \({{{\mathrm{MR}}}} = \frac{{\Delta R}}{{R_{\mathrm{low}}}} \to \infty\) by definition with rigorous switching between quasiparticle currents and supercurrents (Fig. 3d,g). This demonstration of the SQUID spin-valve oscillation can be used to devise a phase-resolved and magnetization-component-specific detector of antiferromagnetic domain walls, which remains a major challenge in the research field of AFM spintronics28.
For the I-biased SQUID of two overdamped JJs in the limit of small self-inductance and in the low-field limit (ΦJJ ≪ Φ0), the conversion of a magnetic flux into V modulation can be approximated by26
where Rn1 (Rn2) is the normal-state zero-bias resistance of the first (second) JJ. From the measured \(V\left( {\mu _0H_ \bot ,\;I \ge I_\mathrm{c}^{\mathrm{tot}}} \right)\) oscillation with a period of μ0Hosc = 0.16‒0.24 mT (Fig. 3d–f and Extended Data Fig. 3) and using the relationship \(\mu _0H_{\mathrm{osc}} = \frac{{{{\varPhi }}_0}}{{A_{\mathrm{SQUID}}^{\mathrm{eff}}}}\), we obtain \(A_{\mathrm{SQUID}}^{\mathrm{eff}}\) = 9‒13 μm2. This value is 2‒3 times larger than the geometrical area of the SQUID loop ((2λL + Lx)(2λL + Ly) = 4.1 μm2), which we attribute to a flux-focusing effect. Note that if the width of the SQUID loop is much larger than λL, the flux-focusing effect comes into play and effectively widens the SQUID area to be \(L_x^{\mathrm{ctc}}L_y^{\mathrm{ctc}}\) ≈ 11 μm2 (Fig. 3b), indicating the reliable performance of our SQUID. Here \(L_{x,y}^{\mathrm{ctc}}\) is the centre-to-centre spacing between the tracks defining the two opposite sides of the SQUID.
Interestingly, from a comparison of the sweep-up and sweep-down VSQUID(μ0H⊥) data (Fig. 3f and Extended Data Fig. 4), a finite phase shift of φ1 + φ2 ≈ π is evident, which does not exist in a normal-metal Cu-JJ-based SQUID (Fig. 4 and Extended Data Fig. 4). In ferromagnetic spin-triplet JJs29, whether the JJ will be a 0 junction or a π junction depends on the sum of the rotational chirality from left spin-mixer FM1 to central spin-rotator FM2 and that from central spin-rotator FM2 to right spin-mixer FM3. If the JJ has the same rotational chirality across the entire FM1/FM2/FM3 Josephson barrier, then the junction will be a 0 junction, whereas if it has the opposite rotational chirality across the FM1/FM2/FM3 barrier, then the junction will be a π junction. This suggests that the OOP rotational chirality and ground-state phase difference of our chiral antiferromagnetic spin-triplet JJs seem to be controlled by external OOP magnetic fields. In fact, our theoretical modelling (Methods and Supplementary Section 1) assures that both Josephson critical current and phase shift crucially depend on the chiral antiferromagnetic spin structure (or spin textures in the chiral AFM), which can change when the OOP magnetic field is swept. As presented by equation (23) in Supplementary Section 1, our theory predicts that for \(d_{\mathrm{s}}\frac{{2JM_0}}{{\hbar v_{\mathrm{F}}}}\tau > \uppi /2\), the JJ can transition to a π junction from a 0 junction. Given our theory (equation (28) in Supplementary Section 1) that \(\frac{{2J\delta M}}{{\hbar v_{\mathrm{F}}}}\) is the inverse decay length of triplet supercurrents \(\left({\xi _{\mathrm{triplet}}^{{\mathrm{Mn}}_3{\mathrm{Ge}}}}\right)^{ - 1}\) and δM is in the same order as M0, we theoretically expect the 0-to-π transition appearing for \(d_{\mathrm{s}} > \frac{\uppi }{2}\left( {\frac{{\hbar v_{\mathrm{F}}}}{{2JM_0}}} \right)\approx \frac{\uppi }{2}\xi _{\mathrm{triplet}}^{{\mathrm{Mn}}_3{\mathrm{Ge}}}\) ≈ 200 nm (whose value is taken from Extended Data Fig. 1). This agrees with what we observe (Fig. 3f and Extended Data Fig. 4). We also emphasize that for ds ≈ 80 nm(<\(\xi _{\mathrm{triplet}}^{{\mathrm{Mn}}_3{\mathrm{Ge}}}\)) Mn3Ge JJ-based SQUID (Extended Data Fig. 5), none of the supercurrent spin-valve behaviour and the 0-to-π phase shift as a function of μ0H⊥ clearly emerge, which is again consistent with our theoretical prediction (equation (28) in Supplementary Section 1).
By substituting the chiral AFM Mn3Ge with normal-metal Cu (Fig. 4a,b), we also perform a control experiment to check a possible contribution of OOP Abrikosov vortex nucleation under μ0H⊥. The Cu-JJ-based SQUID reveals higher \(I_\mathrm{c}^{\mathrm{tot}}\left( {\mu _0H_ \bot = 0} \right)\) even with larger ds (201 and 205 nm; Fig. 4c) than the Mn3Ge JJ-based SQUID, as would be expected for the long singlet superconducting proximity effect (a few hundred nanometres) in the highly conductive normal-metal Cu barrier31. Its well-defined \(I_\mathrm{c}^{\mathrm{tot}}(\mu _0H_ \bot )\) interference pattern in modest fields further supports a good Josephson property (Extended Data Fig. 4). When biasing \(I \ge I_\mathrm{c}^{\mathrm{tot}}\left( {\mu _0H_ \bot = 0} \right)\) to the Cu-JJ-based SQUID (Fig. 4d–f), we do not observe any asymmetric hysteretic behaviour in the V(μ0H⊥) curves (Fig. 4g) but do measure the clear V(μ0H⊥) oscillation of μ0Hosc = 0.16‒0.17 mT (Extended Data Fig. 3). This result indicates that OOP Abrikosov vortices are unlikely to be the source of the found supercurrent spin-valve behaviour.
Conclusions
Recent theories21,22 have suggested that the out-of-kagome-plane overall tilting of the triangular non-collinear AFM spin arrangement by a few degrees in Mn3X (X = Ir, Sn, Ge) indeed visibly changes the k-space Berry curvature near the Fermi energy and the associated anomalous Hall response. In addition, vanishingly small but finite OOP canted magnetization of our D019-Mn3Ge(0001) film (Extended Data Fig. 6) enables the hysteretic behaviour of Josephson supercurrents as a function of μ0H⊥. These further support our claim that magnetic-field-controlled triplet pairing states through the Berry curvature modification can lead to the spin-triplet supercurrent spin-valve effect even in a single topological chiral AFM. In fact, our theory, which takes both real-space magnetic texture (under an OOP magnetic field) and k-space Weyl nodes into account (Supplementary Section 1), reproduces the observed hysteric Fraunhofer pattern (Fig. 1d). Although the theory has to be further developed, especially regarding full boundary conditions for quasiclassical Green’s functions and their chirality dependence, our present model reasonably captures the physics behind our experimental findings (that is, hysteresis in the Fraunhofer pattern and 0-to-π phase shift in the SQUID data). How microscopic details of antiferromagnetic spin texturing and out-of-kagome-plane titling affect the chirality-dependent phase also need to be systematically studied in the future. We believe that our result facilitates a better understanding of the role of the Berry curvature in singlet-to-triplet pair conversion, inspires future theoretical studies on the interplay of Berry curvature and spin-triplet pairing in a chiral non-collinear AFM in more detail and provides a radical route for controlling the triplet-pair amplitude by an extremely small magnetic field—an essential prerequisite for logic circuit1 or AFM domain-wall sensor28 applications of spin-triplet supercurrents.
Methods
Sample growth and characterization
Single-phase hexagonal D019-Mn3Ge(0001) (ref. 20) and Cu thin films were epitaxially grown on a Ru-buffered Al2O3(0001) substrate by d.c. magnetron sputtering in an ultrahigh-vacuum system with a base pressure of 1 × 10−9 torr. A 5-nm-thick Ru buffer layer was first sputtered at 450 °C with a sputtering power of 15 W and Ar pressure of 3 mtorr. Subsequently, Mn and Ge were co-deposited from elemental sputter targets on the Ru(0001) buffer layer at 500 °C and Ar pressure of 3 mtorr, where the sputter powers of 31 and 10 W for Mn and Ge, respectively, were used. Note that these growth conditions are essentially the same as those used for our recent study20. On the other hand, the epitaxial Cu layer (Extended Data Fig. 2) was sputtered at 27 °C with a sputter power of 15 W and Ar pressure of 3 mtorr. All these single-phase hexagonal D019-Mn3Ge(0001) (ref. 20) and Cu epitaxial films were capped with a 1-nm-thick AlOx layer to prevent oxidation. We performed the structural and magnetic characterizations of the prepared thin films using X-ray diffraction and SQUID vibrating-sample magnetometer, respectively. To investigate the Berry-curvature-driven anomalous Hall effect15,16,17,20, we also carried out magnetotransport measurements on the unpatterned D019-Mn3Ge(0001) film in the van der Pauw geometry.
Lithography patterning and device fabrication
As the fabrication procedure of lateral Nb/Mn3Ge/Nb JJs (Fig. 1b) was previously discussed20, here we only describe the fabrication steps for the d.c. SQUID (Figs. 3a and 4a). A central track of either D019-Mn3Ge/Ru (Fig. 3a) or Cu/Ru epitaxial layers (Fig. 4a) with lateral dimensions of 1.5 × 50.0 μm2 was first defined using optical lithography and Ar-ion-beam etching, and then Au (80 nm)/Ru (2 nm) electrical leads and bonding pads were deposited by Ar-ion-beam sputtering. We subsequently defined the SQUID loop with an inner area of 1.0 × 3.0 μm2 (Figs. 3a and 4a), in which two constituent JJs were formed on top of the Mn3Ge/Ru (Fig. 3a) or Cu/Ru (Fig. 4a) track via electron-beam lithography and lift-off steps. The 50-nm-thick Nb electrodes were grown by Ar-ion-beam sputtering at a pressure of 1.5 × 10–4 mbar and the two constituent JJs are edge-to-edge separated by ≥\(\xi _{\mathrm{triplet}}^{{\mathrm{Mn}}_3{\mathrm{Ge}}}\) (157–178 nm; Extended Data Fig. 1). For direct metallic electrical contacts, the Al2O3 capping layer and Au surface were etched away by an Ar-ion beam before sputtering the Nb electrodes.
Low-temperature transport measurement
We measured the I–V curves of the fabricated JJs (Fig. 1d) and d.c. SQUID (Figs. 3c and 4c) with a four-probe configuration in a Quantum Design physical property measurement system using a Keithley 6221 current source and Keithley 2182A nanovoltmeter. The Josephson critical current Ic and normal-state zero-bias resistance Rn of each JJ (Extended Data Fig. 1) were determined by fitting the measured I–V curves with the standard formula for overdamped junctions26, namely, \(V(I) = \frac{I}{{\left| I \right|}}R_\mathrm{n}\sqrt {I^2 - I_\mathrm{c}^2}\). We obtained the magnetic-field interference pattern Ic(μ0H) (Fig. 1d,e and Extended Data Fig. 4) by repeating the I–V measurements at T = 2 K with varying magnetic fields μ0H⊥, applied perpendicular to the kagome plane of D019-Mn3Ge(0001), from negative to positive values, and vice versa. We subsequently measured the V(μ0H⊥) curves for the I-biased JJs (Fig. 2a–d) and d.c. SQUID (Figs. 3d–g and 4d–g) by sweeping μ0H⊥ up and down.
Quasiclassical theory of superconducting proximity effect and spin-triplet supercurrent spin-valve effect
As presented in Supplementary Section 1, we developed the quasiclassical theory of the superconducting proximity effect in a conventional AFM and chiral AFM. We derived the equations describing the propagation of superconducting correlations in the diffusive limit—Usadel equations—that are relevant to the devices studied experimentally here. We found that all the superconducting correlations of spin-unpolarized singlets (S = 0) and spin-zero (S = 1, ms = 0) and spin-polarized (S = 1, ms = ±1) triplets are strongly damped by exchange spin-splitting fields in the conventional AFM, leading to a short-ranged proximity effect25,32,33. However, in case of chiral AFM, spin-momentum locking along with the Weyl node structure turned out to cause a distinctively different superconducting proximity effect. The spin-momentum locking implies that spin texturing in the chiral AFM plays the role of a vector potential, thereby phase shifting the superconducting order parameter and inducing a φ-junction (or π-junction) behaviour, which is controlled by modulating the chiral antiferromagnetic spin texturing. Note also that the Weyl node structure (directly relevant to the Berry curvature) imposes the existence of spin-triplet correlations inside the chiral AFM (equation (16) in Supplementary Section 1). These correlations are expected to propagate over a long distance (equation (17) in Supplementary Section 1).
When the OOP magnetic field is applied and swept, the amplitude IChiral and phase φ0,Chiral of the Josephson triplet supercurrent through the ds > 150 nm Mn3Ge barrier can both visibly change because these values scale directly with ds and are determined by how the antiferromagnetic spin texture of Mn3Ge is configured (Supplementary Section 1). This is the theoretical insight that reasonably explains all the experimental findings of the present paper. Full details of our quasiclassical theory, which reproduces the hysteretic Fraunhofer pattern (Fig. 1d) and explains the 0-to-π phase shift in the SQUID data (Extended Data Fig. 3a,b), can be found in equations (24)–(28) in Supplementary Section 1.
Data availability
The data used in this Article are available from the corresponding authors upon reasonable request.
References
Linder, J. & Robinson, J. W. A. Superconducting spintronics. Nat. Phys. 11, 307–315 (2015).
Eschrig, M. Spin-polarized supercurrents for spintronics: a review of current progress. Rep. Prog. Phys. 78, 104501 (2015).
Birge, N. O. Spin-triplet supercurrents in Josephson junctions containing strong ferromagnetic materials. Philos. Trans. R. Soc. A 376, 20150150 (2018).
Houzet, M. & Buzdin, A. I. Long range triplet Josephson effect through a ferromagnetic trilayer. Phys. Rev. B 76, 060504(R) (2007).
Bergeret, F. S., Volkov, A. F. & Efetov, K. B. Long-range proximity effects in superconductor-ferromagnet structures. Phys. Rev. Lett. 86, 4096–4099 (2001).
Cottet, A. Inducing odd-frequency triplet superconducting correlations in a normal metal. Phys. Rev. Lett. 107, 177001 (2011).
Bergeret, F. S. & Tokatly, I. V. Spin–orbit coupling as a source of long-range triplet proximity effect in superconductor–ferromagnet hybrid structures. Phys. Rev. B 89, 134517 (2014).
Robinson, J. W. A., Witt, J. D. S. & Blamire, M. G. Controlled injection of spin-triplet supercurrents into a strong ferromagnet. Science 329, 59–61 (2010).
Khaire, T. S., Khasawneh, M. A., Pratt, W. P. Jr. & Birge, N. O. Observation of spin-triplet superconductivity in Co-based Josephson junctions. Phys. Rev. Lett. 104, 137002 (2010).
Keizer, R. S. et al. A spin triplet supercurrent through the half-metallic ferromagnet CrO2. Nature 439, 825–827 (2006).
Jeon, K.-R. et al. Enhanced spin pumping into superconductors provides evidence for superconducting pure spin currents. Nat. Mater. 17, 499–503 (2018).
Banerjee, N. et al. Controlling the superconducting transition by spin-orbit coupling. Phys. Rev. B 97, 184521 (2018).
Banerjee, N., Robinson, J. W. A. & Blamire, M. G. Reversible control of spin-polarized supercurrents in ferromagnetic Josephson junctions. Nat. Commun. 5, 4771 (2014).
Martinez, W. M., Pratt, W. P. Jr. & Birge, N. O. Amplitude control of the spin-triplet supercurrent in S/F/S Josephson junctions. Phys. Rev. Lett. 116, 077001 (2016).
Nayak, A. K. et al. Large anomalous Hall effect driven by a nonvanishing Berry curvature in the noncolinear antiferromagnet Mn3Ge. Sci. Adv. 2, e1501870 (2016).
Kiyohara, N., Tomita, T. & Nakatsuji, S. Giant anomalous Hall effect in the chiral antiferromagnet Mn3Ge. Phys. Rev. Appl. 5, 064009 (2016).
Soh, J.-R. et al. Ground-state magnetic structure of Mn3Ge. Phys. Rev. B 101, 140411(R) (2020).
Xiao, D., Chang, M.-C. & Niu, Q. Berry phase effects on electronic properties. Rev. Mod. Phys. 82, 1959–2007 (2010).
Chen, T. et al. Anomalous transport due to Weyl fermions in the chiral antiferromagnets Mn3X, X = Sn, Ge. Nat. Commun. 12, 572 (2021).
Jeon, K.-R. et al. Long-range supercurrents through a chiral non-collinear antiferromagnet in lateral Josephson junctions. Nat. Mater. 20, 1358–1363 (2021).
Chen, H., Niu, Q. & MacDonald, A. H. Anomalous Hall effect arising from noncollinear antiferromagnetism. Phys. Rev. Lett. 112, 017205 (2014).
Busch, O., Göbel, B. & Mertig, I. Microscopic origin of the anomalous Hall effect in noncollinear kagome magnets. Phys. Rev. Research 2, 033112 (2020).
Fermin, R. et al. Superconducting triplet rim currents in a spin-textured ferromagnetic disk. Nano Lett. 22, 2209–2216 (2022).
Bhatia, E. et al. Nanoscale domain wall engineered spin-triplet Josephson junctions and SQUID. Nano Lett. 21, 3092–3097 (2021).
Komori, S. et al. Spin-orbit coupling suppression and singlet-state blocking of spin-triplet Cooper pairs. Sci. Adv. 7, eabe0128 (2021).
Barone, A. & Paterno, G. Physics and Applications of the Josephson Effect 2nd edn (John Wiley & Sons, 1982).
Gubin, A. I., Il’in, K. S. & Vitusevich, S. A. Dependence of magnetic penetration depth on the thickness of superconducting Nb thin films. Phys. Rev. B 72, 064503 (2005).
Jungwirth, T., Marti, X., Wadley, P. & Wunderlich, J. Antiferromagnetic spintronics. Nat. Nanotechnol. 11, 231–241 (2016).
Glick, J. A. et al. Phase control in a spin-triplet SQUID. Sci. Adv. 4, eaat9457 (2018).
Guichard, W. et al. Phase sensitive experiments in ferromagnetic-based Josephson junctions. Phys. Rev. Lett. 90, 167001 (2003).
Blum, Y., Tsukernik, A., Karpovski, M. & Palevski, A. Oscillations of the superconducting critical current in Nb-Cu-Ni-Cu-Nb junctions. Phys. Rev. Lett. 89, 187004 (2002).
Bell, C. et al. Proximity and Josephson effects in superconductor/antiferromagnetic Nb/γ-Fe50Mn50 heterostructures. Phys. Rev. B 68, 144517 (2003).
Weides, M., Disch, M., Kohlstedt, H. & Bürgler, D. E. Observation of Josephson coupling through an interlayer of antiferromagnetically ordered chromium. Phys. Rev. B 80, 064508 (2009).
Schroder, D. K. Semiconductor Material and Device Characterization 2nd edn (Wiley-Blackwell, 1998).
Acknowledgements
This work was supported by the Alexander von Humboldt Foundation and the National Research Foundation (NRF) of Korea (Grant No. 2020R1A5A1016518).
Funding
Open access funding provided by Max Planck Society.
Author information
Authors and Affiliations
Contributions
K.-R.J. and S.S.P.P. conceived and designed the experiments with the help of T.K. B.K.H. grew the thin films of the chiral non-collinear AFM Mn3Ge and normal-metal Cu, and characterized their crystallinity and magnetic properties. H.L.M. and H.H. helped with the structural analysis. K.-R.J. fabricated the lateral JJs and d.c. SQUID with the help of J.-K.K. and carried out the transport measurements with the help from J.-C.J. K.-R.J. performed the data analysis with the help of T.K. A.C. and T.K. developed the quasiclassical theory of superconducting proximity effect in both conventional AFM and chiral AFM systems. K.-R.J., B.K.H., T.K., A.C. and S.S.P.P. wrote the manuscript with input from all the other co-authors.
Corresponding authors
Ethics declarations
Competing interests
The authors declare no competing interests.
Peer review
Peer review information
Nature Nanotechnology thanks the anonymous reviewers for their contribution to the peer review of this work.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Extended data
Extended Data Fig. 1 Long-range supercurrents and spin-valve effect in chiral antiferromagnetic JJs.
a, Normal-state zero-bias resistance Rn (top) and Josephson critical current Ic (bottom) of the Nb/Mn3Ge/Nb JJs versus the barrier spacing ds. From Rn(ds) in a, we extract the resistance-area product ri of Nb/Mn3Ge interfaces to be 1.0–1.2 mΩ µm2 and the effective resistivity ρch for the Mn3Ge(40 nm)/Ru(5 nm) track to be 25–27 µΩ cm, employing a standard transmission line (TL) theory34, \(R_n = 2R_i + R_{ch} = \frac{{\rho _{ch}}}{{tw}}\left( {2l_t + d_s} \right)\). Here \(R_i = \frac{{r_i}}{{wl_t}}\) and \(l_t = \sqrt {\frac{{r_i}}{{\frac{{\rho _{ch}}}{t}}}}\) is the charge transfer length (13 nm). t and w is the thickness and width of the Mn3Ge/Ru track. b, Characteristic voltage Vc = IcRn as a function of ds, from which we determine the decay length \(\xi _{triplet}^{Mn_3Ge}\) of the Josephson supercurrents in the Mn3Ge barrier to be 157–178 nm using an exponential decay function20, \(exp\left( { - \frac{{\xi _{triplet}^{Mn_3Ge}}}{{d_s}}} \right)\) (black curves). Note that for ds ≥ \(\xi _{triplet}^{Mn_3Ge}\), the complete supercurrent spin-valve effect appears (see Fig. 2d). All Ic values in a,b are obtained at a fixed temperature T = 2 K. Note that data with circle symbols in a,b are taken from Ref. 20. The error bars in a and b represent the standard deviation.
Extended Data Fig. 2 Structural analysis of the chiral non-collinear AFM Mn3Ge and the normal-metal Cu.
a, Specular θ-2θ XRD pattern of single-crystalline Ru(0001) (black symbol) and single-phase Mn3Ge(0001)/Ru (red symbol) films grown on Al2O3(0001) substrates. b, Phi ϕ scan across the Al2O3 {\(10\bar 14\)}, Ru {\(10\bar 13\)}, and Mn3Ge {\(20\bar 21\)} reflections for the same film. c, Diffraction pattern of the epitaxial Cu film grown on Ru/Al2O3(0001) substrate. d, Atomic force micrographs of the single-phase Mn3Ge(0001) and epitaxial Cu films over 5 × 5 µm2 scan area (from left to right respectively). The root-mean-square roughness Rrms of the single-phase Mn3Ge (epitaxial Cu) is estimated to be 0.7 (0.6) nm. These data confirm that the single crystallinity and surface morphology of previous20 and current D019-Mn3Ge(0001) films are almost identical.
Extended Data Fig. 3 Magnetic-flux-to-voltage conversion in Mn3Ge JJ- or Cu JJ-based SQUID.
SQUID voltage VSQUID oscillation as a function of normalized magnetic flux \({\it{\Phi }}_{SQUID}/{\it{\Phi }}_0\) for the I = +0.30 mA-biased Mn3Ge JJ-based SQUID, taken at T = 2 K when sweeping μ0H⊥ up (a) and down (b). Note that we set \(A_{SQUID}^{eff}\) = 9 µm and subtract background low-order polynomial and asymmetric voltage signals are from Fig. 3f for clarity. c,d, Data equivalent to a,b but for the I = +1.15 mA-biased Cu JJ-based SQUID. Here \(A_{SQUID}^{eff}\) = 12 µm is set and background low-order polynomial voltage signals are subtracted from Fig. 4f. Unlike the Cu JJ-based SQUID, there exists a finite phase shift φ1 + φ2 ≈ π in the sweep-up and sweep-down \(V_{SQUID}\left( {{\it{\Phi }}_{SQUID}/{\it{\Phi }}_0} \right)\) data of the Mn3Ge JJ-based SQUID. This implies that the rotational chirality29 and ground-state phase difference of our chiral antiferromagnetic spin-triplet JJs seems to be controlled by μ0H (see main text for details).
Extended Data Fig. 4 Magnetic-field interference pattern of the Cu JJ-based SQUID.
a, Estimated electrical resistivity versus temperature T plot for the Cu(40 nm)/Ru(5 nm) reference Hall-bar device. b, Total critical current \(I_c^{tot}\) versus μ0H⊥ plot for the Cu JJ-based SQUID, taken at T = 2 K. Here μ0H⊥ is applied perpendicular to the interface plane of Nb electrodes, as illustrated in the bottom inset. The top left inset displays the zero-field current-voltage I-V curve at 2 K whereas the top right inset shows the I-V curve near the zero-order minimum of \(I_c^{tot}\)(μ0H⊥). To focus on a full modulation of \(I_c^{tot}\) caused by the magnetic-field quantum interference of two Nb/Cu/Nb lateral JJs, we measure the \(I_c^{tot}\)(μ0H⊥) curves at a rather large interval (1 mT) of μ0H⊥. The black solid lines in b are fitting curves20 to determine the supercurrent non-uniformity γ and the effective junction area AeffJJ. The error bars in b represent the standard deviation.
Extended Data Fig. 5 No supercurrent spin-valve effect and 0-to-π phase shift detectable in the \({\it{d}}_{\it{s}} < {\it{\xi }}_{{\it{triplet}}}^{{\it{Mn}}_3{\it{Ge}}}\) Mn3Ge JJ-based SQUID.
a, Typical magnetic-field interference pattern Ic(μ0H⊥) of the ds = 84 nm (<\(\xi _{triplet}^{Mn_3Ge}\) = 157–178 nm) JJ. The inset displays the magnified plot around the zero field μ0H⊥ = 0. b, Time-averaged voltage V as a function of external magnetic field μ0H⊥ for the dc current I-biased ds = 84 nm JJ, taken at 2 K. In this measurement, μ0H⊥ (≤|3 mT|) is applied perpendicular to the Kagome plane of the Mn3Ge barrier. c-e, Time-averaged voltage V as a function of perpendicular magnetic field μ0H⊥ for the I-biased SQUID, taken at T = 2 K. From the periodic \(V\left( {\mu _0H_ \bot ,\;I \ge I_c^{tot}} \right)\) modulation in c-e, we find a characteristic period of μ0Hosc ≈ 0.30 mT. None of the supercurrent spin-valve behaviour and the 0-to-π phase shift as a function of μ0H⊥ clearly emerge in the \(d_s \approx 80\;nm\;\left( { < \xi _{triplet}^{Mn_3Ge}} \right)\) Mn3Ge JJ-based SQUID, which are well consistent with our theoretical modelling [Eq. (S28)]. The error bars in a represent the standard deviation.
Extended Data Fig. 6 Magnetic properties and AHE of the chiral non-collinear AFM Mn3Ge.
a,b, Typical magnetic hysteresis M(μ0H||) curves of the single-phase 40 nm Mn3Ge(0001) film at 300 K and 2 K, where μ0H|| is applied parallel to the Kagome ab plane. Black represents the raw measured data and blue represents the corrected data after subtracting the paramagnetic contribution. These data are taken from Ref. 20. c,d, Data equivalent to a,b but for the M(μ0H⊥) curves, where μ0H⊥ is applied perpendicular to the Kagome ab plane. e, Representative Hall resistivity versus magnetic field R-H curves of the single-phase Mn3Ge(0001) at 2 and 300 K. These data, taken from Ref. 20, suggest that there exists vanishingly small but finite OOP canted magnetization of our D019-Mn3Ge(0001) film, which is in line with the case of bulk Mn3Ge15.
Supplementary information
Supplementary Information
Supplementary Sections 1–3, Fig. 1 and refs. 1–14.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Jeon, KR., Hazra, B.K., Kim, JK. et al. Chiral antiferromagnetic Josephson junctions as spin-triplet supercurrent spin valves and d.c. SQUIDs. Nat. Nanotechnol. 18, 747–753 (2023). https://doi.org/10.1038/s41565-023-01336-z
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1038/s41565-023-01336-z
This article is cited by
-
Phase biasing of a Josephson junction using Rashba–Edelstein effect
Nature Communications (2023)