betaFeedback

WorksPhysics and Astronomy

Temperature and Thickness Dependent Spin Current Transmission through Epitaxial Antiferromagnetic α-Fe2O3

Abstract

Antiferromagnetic insulator α-Fe2O3 is a potential medium for coherent motion of spin waves over relatively large distances, allowing non-dissipative transfer of information between circuit elements. However, the structural defects and temperature dependent spin reorientations in this oxide impact the propagation. Understanding how the magnetic state of an antiferromagnetic insulator affects spin-current transmission is therefore important for elucidating the mechanisms governing spin transport. Here, we report spin-current transmission through epitaxial α-Fe2O3 thin films sandwiched between Permalloy and Pt. Using spin pumping and inverse spin Hall effect, we observe a finite symmetric dc voltage across the Pt layer over the temperature range 1.6 to 300 K for alpha-Fe2O3 thicknesses of 0, 30, 40, and 100 nm. The symmetric voltage coefficient Ks exhibits two approximately linear temperature regimes, separated by a characteristic temperature T* whose value depends on the thickness of the α-Fe2O3 layer. Further, the dc magnetization measurements on some of the samples suggest that the T* is correlated with the spin reorientation transition (Morin transition) in α-Fe2O3. Interestingly, a significantly larger magnitude of dKs/dT is noted below the characteristic temperature T*. The observed thickness and temperature dependent spin-current transmission indicates a sensitivity to the magnetic state of the insulating hematite layer.

The ability to generate, transport, and manipulate spin angular momentum without an accompanying charge current is a central objective of minimally dissipative spintronics. Magnetic insulators are particularly attractive for this purpose because their collective magnetic excitations can propagate over large distances without loss of coherence[1]. Ferrimagnetic insulators like yttrium iron garnet have provided a platform for studying spin transport[2, 3, 4], while antiferromagnetic (AFM) insulators offer additional possibilities associated with their compensated magnetic structure, high-frequency dynamics, and negligible stray fields. Experiments on oxide-based AFMs like NiO, Cr2_2O3_3 and Fe2_2O3_3 have demonstrated spin-angular-momentum transmission through nanometre-thick layers and subsequent detection in an adjacent heavy metal, establishing AFM insulators as potential media for spin transport[5, 6, 7, 8, 9].

Spin pumping combined with inverse spin Hall effect (ISHE) in a strongly spin-orbit-coupled (SOC) metal like Pt provides a convenient means of investigating spin transmission across insulating layers. During ferromagnetic resonance (FMR), a driven ferromagnet can pump spin current into an adjacent material[10]. When a SOC heavy metal is separated from the ferromagnet by an insulating magnetic layer, the transmitted spin current can be detected in the heavy metal through its ISHE response. Studies of YIG/NiO/Pt structures have shown that spin-pumping signal depends on AFM-layer thickness, while experiments on epitaxial NiO have demonstrated the importance of crystallographic orientation and interfacial properties[5, 11, 12]. These results indicate that spin-current transmission is sensitive not only to magnetic properties of the insulating layer but also to its structural and interfacial characteristics.

Hematite, α\alpha-Fe2_2O3_3 is a particularly interesting AFM-insulating system for spin pumping and spin propagation studies. It has a high Néel temperature (~955 K) and undergoes a Morin transition between different magnetic configurations. Across the transition, the Fe moments reorient from abab plane to cc axis, providing an opportunity to examine how spin transmission evolves with the magnetic state of the antiferromagnet[13, 14, 15, 16]. Spin transport through crystalline hematite has been observed over extended distances and across Morin transition[17]. Recent studies have also demonstrated magnon transport and magnon torque in epitaxial α\alpha-Fe2_2O3_3 films18,19. Importantly, the Morin transition in hematite thin films is sensitive to thickness, strain, structural quality, substrate-induced effects, and magnetic field, and can therefore occur at temperatures substantially different from the bulk value[13, 15] [20]. Such sensitivity raises a fundamental question of how does spin-current transmission change when the magnetic state of an epitaxial hematite layer is altered? Therefore, temperature dependent measurements across this transition are essential for establishing the relationship between the magnetic state of hematite and spin-current transmission.

Although spin pumping and ISHE have been demonstrated in α\alpha-Fe2_2O3_3/Pt heterostructures, these experiments primarily probe spin pumping generated by the magnetic dynamics of hematite itself[17] [21]. Additionally, most of these studies have been conducted at ambient temperatures19,22,23. However, to the best of our knowledge, the effects of the temperature and thickness-dependent magnetic state of epitaxial α\alpha-Fe2_2O3_3 on its spin current transmission characteristics have not been reported. Here, we investigate this phenomenon in epitaxial α\alpha-Fe2_2O3_3 layers sandwiched between the films of permalloy (Py) and Pt. We study heterostructures containing 30, 40, and 100 nm thick α\alpha-Fe2_2O3_3 layers, together with a Py/Pt reference sample, over a wide temperature range spanning from 1.6 to 300 K in an ISHE/FMR spectrometer. These experiments reveal a dc voltage across the Pt layer associated with spin-current transmission through the hematite over entire temperature range investigated. The temperature dependence of symmetric ISHE voltage coefficient, KsK_s exhibits distinct low and high-temperature regimes separated by a characteristic T∗T^* which depends on the hematite thickness. This behaviour of KsK_s indicates a connection between spin-current transmission and reorientation of Fe-moments from ab-plane to along the c-axis at the Morin transition in hexagonal Fe2_2O3_3.

Thin-film heterostructures of the type (0001) Al2_2O3_3/Pt (10 nm)/α\alpha-Fe2_2O3_3(x nm)/Py (30 nm), with xx = 30, 40 and 100 nm were deposited using a load-lock, multi-target high-vacuum magnetron sputtering system with a base pressure of 8 ×\times 10−9^{-9} Torr. Platinum is known to grow epitaxially with (111) orientation on C-plane sapphire due to low lattice mismatch of 0.9%24. Pt seed layer was deposited at 500 ºC, followed by cooling to 350 ºC for deposition of subsequent Pt and α\alpha-Fe2_2O3_3 layers. The initial high-temperature Pt deposition promotes epitaxial Pt (111) nucleation, while subsequent lower-temperature deposition suppresses 3-dimensional island growth and surface roughness associated with prolonged high-temperature Pt growth24. The α\alpha-Fe2_2O3_3 layers were deposited by reactive sputtering of Fe in an Ar-O2_2 atmosphere with Ar:O2_2 flow of ~ 12.5:4 sccm. The Py layer was subsequently deposited in pure argon at room temperature after flushing out oxygen, followed by an Al capping layer to protect the heterostructure from oxidation. The crystallographic structure of the stack was evaluated by X-ray diffraction measurements performed using a Rigaku SmartLab diffractometer. Surface morphology and magnetic textures of the films were evaluated using Park Systems NX10 scanning probe microscope. For in-plane magnetization measurements, we have used the vibrating sample magnetometer (VSM) option of physical property measurement system equipped with a 9 T superconducting magnet. FMR measurements were performed using a custom-built broadband Cryo-FMR spectrometer operated in frequency-modulation mode from 1.6 to 300 K25. The Py/α\alpha-Fe2_2O3_3/Pt/(0001) Al2_2O3_3 samples were placed face down on the signal line of grounded coplanar waveguide (GCPW) inside variable-temperature insert of a superconducting magnet. A frequency-modulated microwave signal was applied to GCPW, and transmitted signal was rectified and detected using a lock-in amplifier. Mutually perpendicular dc and RF magnetic fields were oriented in the sample plane. Simultaneously, FMR-induced dc voltage, including ISHE and spin-rectification contributions, was measured using a nanovoltmeter.

We first characterize structural and magnetic properties of the heterostructures used for spin-current transmission measurements. The AFM images in Figure 1(a) and (Figure 1b) show a strong dependence of Pt morphology on the oxygen content in the plasma during sputtering. Without oxygen (Figure 1(a)), the 9-nm Pt film exhibits a discontinuous, maze-like morphology with irregular islands, indicative of Volmer-Weber growth26, with a surface roughness of ~2.1 nm. This suggests limited wetting of Pt on Al2_2O3_3 and high surface diffusion, leading to isolated clusters. In contrast, Pt deposited in oxygen (Figure 1(b)), shows elongated, more continuous features with reduced surface roughness of ~1.7 nm, indicating improved film continuity due to modified surface interactions. The morphology of a 50 nm thick Fe2O3\mathrm{Fe}_2\mathrm{O}_3 film grown on Pt (111) is shown in Figure 1(c), with a grain size of ~80 nm and surface roughness of ~1.2 nm. Figure 1(d) shows the 2θ2\theta-ω\omega X-ray diffraction pattern of the heterostructure grown on cc-plane sapphire. Two prominent features are observed: the Al2O3\mathrm{Al}_2\mathrm{O}_3 (0006) reflection appearing at higher 2θ2\theta and a feature near ~40°. The latter feature is associated with the Pt (111) and α\alpha-Fe2O3\mathrm{Fe}_2\mathrm{O}_3 (006) reflections. Since Pt (111) and α\alpha-Fe2O3\mathrm{Fe}_2\mathrm{O}_3 (006) reflections occur at closely spaced 2θ2\theta positions, the observed reflection cannot be unambiguously assigned to α\alpha-Fe2O3\mathrm{Fe}_2\mathrm{O}_3 (0006) alone. We therefore attribute this feature to the combined diffraction contribution from the Pt (111) and α\alpha-Fe2O3\mathrm{Fe}_2\mathrm{O}_3 (006) planes which is consistent with previous studies19. Figure 2(a) shows the temperature dependence of normalized magnetization of α\alpha-Fe2O3\mathrm{Fe}_2\mathrm{O}_3/Pt heterostructure measured under zero-field-cooled

AFM surface topography and XRD pattern

Figure 1. AFM surface topography of (a) 9 nm Pt deposited without oxygen, (b) 9 nm Pt deposited in oxygen, (c) 50 nm Fe2O3\mathrm{Fe}_2\mathrm{O}_3 grown on a 9 nm Pt-coated Al2O3\mathrm{Al}_2\mathrm{O}_3 substrate and (d) XRD θ\theta–2θ2\theta pattern of Fe2O3\mathrm{Fe}_2\mathrm{O}_3/Pt on (0001) Al2O3\mathrm{Al}_2\mathrm{O}_3. Inset: Schematic of used for spin-current transmission measurements.

Normalized magnetization and resonance frequency versus resonance field

Figure 2. (a) Normalized magnetization of Pt/α\alpha-Fe2O3\mathrm{Fe}_2\mathrm{O}_3(100 nm) under zero-field-cooled (black) and field-cooled (red) conditions. (b) Resonance frequency versus resonance field for Pt/α\alpha-Fe2O3\mathrm{Fe}_2\mathrm{O}_3 (40 nm)/Py heterostructure. The solid line is a fit to Kittel expression.

(black) and field-cooled (red) conditions. As the temperature decreases, magnetization gradually reduces and reaches a minimum near T∼80T \sim80 K. Since the hematite at low-temperatures is a fully compensated antiferromagnet with the Fe moments aligned along the cc-axis[17], this behaviour provides a possible indication of a Morin transition near TM∼80T_M \sim80 K. This value is substantially lower than the bulk Morin transition temperature (~260 K), but such a reduction is plausible in thin films, where TMT_M depends on film thickness, lattice strain, structural defects, and interfacial conditions[13, 15] [20]. It is important to examine whether the change in magnetic configuration influences the transmission of spin angular momentum. Hence, we employ FMR-driven spin pumping from Py layer and detect resulting voltage in Pt layer. The measurement geometry is illustrated schematically in the inset of Figure 1(d).

During FMR excitation, spin current JsJ_s is pumped from the precessing Py magnetization toward Pt through the intervening α\alpha-Fe2_2O3_3. The transmitted spin current is converted into a transverse dc charge current via ISHE. FMR-induced voltages can also contain spin-rectification contributions from the ferromagnetic layer27–29. To minimize the latter effect, the voltage contacts were made only on the Pt layer (see inset of Figure 1d). The measured voltage is subsequently decomposed into its symmetric and antisymmetric components, allowing the symmetric contribution associated with spin-current transmission to be extracted for further analysis25. The characteristic resonance fields HrH_r were obtained by fitting the field-dependent VmixV_{\mathrm{mix}} signals and used to determine the frequency dependence of the resonance. Figure 2 (b) presents frf_r as a function of corresponding HrH_r for Py/α\alpha-Fe2_2O3_3(40 nm)/Pt heterostructure. The observed trend is well described by Kittel relation;

fr=γ2πμ0Hr(Hr+Meff)(1)f_r = \frac{\gamma}{2\pi}\mu_0\sqrt{H_r(H_r+M_{\mathrm{eff}})} \tag*{(1)}

where γ=gμB/h\gamma= g\mu_B/h. Here gg, μB\mu_B and hh are Lande’s gg factor, Bohr magneton and Planck’s constant, respectively and MeffM_{\mathrm{eff}} is the effective magnetization30,31. The fit yields g=2.03±0.04g = 2.03 \pm0.04 and Meff=1.02±0.05M_{\mathrm{eff}} = 1.02 \pm0.05 T, consistent with permalloy thin films32,33. This agreement establishes the basis for the subsequent measurements of the FMR-induced voltage and spin-current transmission. We next separate symmetric and antisymmetric components of VmixV_{\mathrm{mix}} to identify the contributions associated with spin-current transmission through the α\alpha-Fe2_2O3_3 layer. Figure 3 shows representative magnetic-field-dependent VmixV_{\mathrm{mix}} signals measured from 1.6 to 300 K for the Py/Pt reference and for Py/α\alpha-Fe2_2O3_3 (t=30t = 30, 40 and 100 nm)/Pt heterostructures.

Field dependence of Vmix for Py/α-Fe2O3/Pt heterostructures

Figure 3. Field dependence of VmixV_{\mathrm{mix}} for Py/α\alpha-Fe2_2O3_3/Pt heterostructures with α\alpha-Fe2_2O3_3 thicknesses of (a) 0, (b) 30, (c) 40, and (d) 100 nm over the temperature range 1.6–300 K, in 20 K steps. Curves are vertically offset for clarity.

The overall line shape and amplitude of VmixV_{\mathrm{mix}} response remains observable over the entire temperature range, providing a reference for the voltage response in the absence of α\alpha-Fe2_2O3_3 layer. Upon insertion of the hematite layer, a clear voltage response remains detectable for all three thicknesses, from 1.6 K to room temperature, as shown in figs. 3 (b-d).

The persistence of voltage after insertion of α\alpha-Fe2_2O3_3 is important because the insulating layer electrically separates the Py injector from the Pt detector. Thus, the measured response cannot be attributed to direct charge conduction through α\alpha-Fe2_2O3_3 layer. Similar FMR-based experiments on insulating AFM spacers have used voltage detection in Pt to probe spin-current transmission[5, 11]. However, VmixV_{\mathrm{mix}} signal may be contaminated by contributions unrelated to spin pumping and the resultant ISHE voltage28,34. A collective effect of such signals is to make the VmixV_{\mathrm{mix}} asymmetric in field. The antisymmetric components can arise from anisotropic magnetoresistance and anomalous Hall effect in Py, together with thermal-gradient contributions, whereas the ISHE contribution exhibits a symmetric Lorentzian line shape35–37. We therefore separately analyse symmetric and antisymmetric parts of VmixV_{\mathrm{mix}}.

Figure 4 (a) and (Figure 4b) show representative field-dependent voltage measurements for Py/Pt reference bilayer and Py/α\alpha-Fe2_2O3_3(100 nm)/Pt heterostructure at 1.6 K. Both samples exhibit pronounced voltage response near the FMR field. The measured voltage is decomposed into symmetric and antisymmetric components using;

Field-dependent voltage measurements for Py/Pt and $\alpha$-Fe$_2$O$_3$ 100 nm

Figure 4. Field dependence of VmixV_{\mathrm{mix}} for (a) Py/Pt and (b) α\alpha-Fe2_2O3_3 100 nm at 1.6 K, respectively. Solid lines are fits using Eq. (2). Insets show the corresponding symmetric Lorentzian components.

Vmix(H)=KsΔH2(H−Hr)2+ΔH2+Kas−2ΔH(H−Hr)(H−Hr)2+ΔH2(2)V_{\mathrm{mix}}(H) = K_{\mathrm{s}} \frac{\Delta H^2}{(H-H_{\mathrm{r}})^2+\Delta H^2} + K_{\mathrm{as}} \frac{-2\Delta H(H-H_{\mathrm{r}})}{(H-H_{\mathrm{r}})^2+\Delta H^2} \tag*{(2)}

Where KsK_{\mathrm{s}} and KasK_{\mathrm{as}} are amplitudes of the symmetric and antisymmetric components, respectively, and ΔH\Delta H is the half width at half maximum of Vmix(H)V_{\mathrm{mix}}(H)38–40. The black curves in Figure 4(a) and Figure 4(b) represent total fits, with corresponding symmetric components shown in insets. For the Py/Pt reference, symmetric component reaches ∼0.39\sim0.39 μ\muV, while ∼0.21\sim0.21 μ\muV is obtained for the Py/α\alpha-Fe2_2O3_3(100 nm)/Pt heterostructure.

The persistence of the symmetric voltage after insertion of 100 nm α\alpha-Fe2_2O3_3 layer is consistent with spin-current transmission from Py to Pt. We associate KsK_{\mathrm{s}} with the spin current related voltage detected in Pt, while considering that a FMR induced voltage can also contain contributions from spin rectification or thermoelectric effects29,38,41,42.

We examine temperature and thickness dependence of KsK_{\mathrm{s}} below. Figure 5 (a–Figure 5d) show KsK_{\mathrm{s}} as a function of temperature for the Py/Pt reference and for the Py/α\alpha-Fe2_2O3_3/Pt heterostructures with α\alpha-Fe2_2O3_3 thicknesses of 30, 40, and 100 nm, respectively. The Py/Pt bilayer exhibits a nearly linear decrease of KsK_{\mathrm{s}} with increasing temperature, with dKs/dT=(−7.24±0.18)×10−10dK_{\mathrm{s}}/dT = (-7.24 \pm0.18) \times10^{-10} V/K.

Temperature dependence of $K_s$ for Pt/$\alpha$-Fe$_2$O$_3$/Py heterostructures with $\alpha$-Fe$_2$O$_3$ thicknesses of (a) 0, (b) 30, (c) 40, and (d) 100 nm

Figure 5. Temperature dependence of KsK_s for Pt/α\alpha-Fe2_2O3_3/Py heterostructures with α\alpha-Fe2_2O3_3 thicknesses of (a) 0, (b) 30, (c) 40, and (d) 100 nm. Solid lines are linear fits over the respective temperature ranges; dashed lines indicate T∗T^*.

In contrast, α\alpha-Fe2_2O3_3 containing samples show a pronounced change in the slope of KsK_s at low temperatures. For 30 and 40 nm α\alpha-Fe2_2O3_3 films, KsK_s decreases more rapidly below T∗∼40T^* \sim40 K and exhibits a substantially weaker temperature dependence above this characteristic temperature. A similar change occurs for 100 nm film, although T∗T^* shifts to ∼80\sim80 K. The dashed lines in Figure 5 indicate the change in slope near T∗T^*. The linear fits to the two regimes quantify the corresponding dKs/dTdK_s/dT and the values are summarized in Table 1. For all three α\alpha-Fe2_2O3_3 thicknesses, KsK_s decreases more rapidly below T∗T^* as indicated by the larger magnitude of dKs/dTdK_s/dT whereas its temperature dependence is weaker above T∗T^*.

α\alpha-Fe2_2O3_3 thicknessT∗T^* (K)dKs/dTdK_s/dT below T∗T^* (V/K)dKs/dTdK_s/dT above T∗T^* (V/K)
0 nm (Py/Pt)--−7.24±0.17×10−10-7.24 \pm0.17 \times10^{-10}
30 nm∼40\sim40−1.75±0.05×10−9-1.75 \pm0.05 \times10^{-9}−2.66±0.27×10−10-2.66 \pm0.27 \times10^{-10}
40 nm∼40\sim40−1.26±0.49×10−9-1.26 \pm0.49 \times10^{-9}−3.32±0.16×10−10-3.32 \pm0.16 \times10^{-10}
100 nm∼80\sim80−1.13±0.15×10−9-1.13 \pm0.15 \times10^{-9}−8.40±0.45×10−10-8.40 \pm0.45 \times10^{-10}

Table 1. Linear-fit parameters for KsK_s (TT) of Py/Pt reference and Pt/α\alpha-Fe2_2O3_3/Py heterostructures.

These results demonstrate that the temperature dependence of KsK_s is strongly modified by the presence and thickness of the α\alpha-Fe2_2O3_3 layer in these heterostructures. The characteristic temperatures extracted from the behaviour of KsK_s can be compared with the magnetic response.

T∗∼80T^* \sim80 K observed for the 100-nm α\alpha-Fe2_2O3_3 film is close to the temperature at which magnetization of the Pt/α\alpha-Fe2_2O3_3 heterostructure approaches zero. This correspondence suggests that the change in KsK_s is associated with the magnetic reorientation of α\alpha-Fe2_2O3_3 in the vicinity of the Morin transition regime. The large drop in T∗T^* for the 30 and 40 nm films appears to be an indication of a significant change in the TMT_M of these thinner films. As stated earlier, the Morin transition TMT_M in hematite films depend on thickness, strain, and magnetic field, and it may shift depending on growth conditions and strain[13]. Taken together, these observations suggest that the temperature dependence of spin-current transmission is sensitive to the magnetic state of α\alpha-Fe2_2O3_3 layer. To further characterize spin-current transfer, we analyse the ISHE resonance linewidth ΔH\Delta H within the Landau-Lifshitz-Gilbert model

ΔH=ΔH0+2πγαf(3)\Delta H = \Delta H_0 + \frac{2\pi}{\gamma}\alpha f \tag*{(3)}

Where ΔH0\Delta H_0 is the zero-field linewidth resulting from frozen-in defects in materials and α\alpha is Gilbert damping constant43. The fitted values of α\alpha are found to be 6.21 x 10−310^{-3}, 6.59 x 10−310^{-3}, 6.58 x 10−310^{-3} for Py/Pt reference, 30, and 40 nm Fe2O3\mathrm{Fe}_2\mathrm{O}_3 films, respectively. The enhancement in damping relative to the Py/Pt reference is indicative of spin pumping that allows transfer of spin angular momentum from Py to Fe2O3\mathrm{Fe}_2\mathrm{O}_3 layer. Further, the spin pumping efficiency of interface can be evaluated by calculating real part of the spin mixing conductance g↑↓g^{\uparrow\downarrow} using the following relation;

g↑↓=4πMstPygμB(αPy/Fe2O3(t)/Pt−αPy/Pt)(4)g^{\uparrow\downarrow} = \frac{4\pi M_s t_{\mathrm{Py}}}{g\mu_{\mathrm{B}}}\left(\alpha_{\mathrm{Py}/\mathrm{Fe}_2\mathrm{O}_3(t)/\mathrm{Pt}}-\alpha_{\mathrm{Py}/\mathrm{Pt}}\right) \tag*{(4)}

Where MsM_s and tPyt_{\mathrm{Py}} are the saturation magnetization and thickness of the Py layer, respectively34,44. Using μ0Ms=1\mu_0 M_s = 1 T44,45 and tPy=30t_{\mathrm{Py}} = 30 nm, we obtain geff↑↓∼6.15×1018g_{\mathrm{eff}}^{\uparrow\downarrow} \sim6.15 \times10^{18} and 5.98×10185.98 \times10^{18} m−2^{-2} for 30 and 40 nm α\alpha-Fe2O3\mathrm{Fe}_2\mathrm{O}_3 layers, respectively. The reduction in geff↑↓g_{\mathrm{eff}}^{\uparrow\downarrow} with increasing α\alpha-Fe2O3\mathrm{Fe}_2\mathrm{O}_3 thickness is consistent with reduced spin-current transfer through the spacer.

We note that the prominent feature of our measurements is not simply the presence of transmitted spin current but change in its temperature dependence. As shown in Figure 5, KsK_s for all three α\alpha-Fe2O3\mathrm{Fe}_2\mathrm{O}_3 thicknesses exhibits two linear regimes with different slopes. The magnitude of dKs/dTdK_s/dT is larger at lower temperatures, while characteristic temperature T∗T^* depends on α\alpha-Fe2O3\mathrm{Fe}_2\mathrm{O}_3 thickness as seen from its value of ∼40\sim40 K for the 30 and 40 nm films and ∼80\sim80 K for 100 nm film. The latter is comparable to the temperature range in which the magnetization of Pt/α\alpha-Fe2O3\mathrm{Fe}_2\mathrm{O}_3 sample becomes nearly zero, suggesting that change in KsK_s is associated with magnetic reorientation regime of α\alpha-Fe2O3\mathrm{Fe}_2\mathrm{O}_3. This interpretation is consistent with previous observations that spin transport depends on the magnetic state and thickness of antiferromagnetic insulators46. Such sensitivity is not unexpected in epitaxial antiferromagnetic films, where strain, crystallinity, interfaces, and magnetic configuration can influence spin dependent properties[13] [20]. Hence, these results motivate experiments that directly resolve the magnetic excitations contributing to spin transport. For example, broadband GHz spin-wave spectroscopy or Brillouin light scattering could determine spin-wave dispersion, damping, and propagation characteristics23, while time-resolved THz spectroscopy could probe ultrafast spin-current dynamics and distinguish contributions from coherent antiferromagnetic excitations and spin-caloritronic processes47.

In summary, we investigated spin-current transmission through epitaxial α\alpha-Fe2O3\mathrm{Fe}_2\mathrm{O}_3 thin films incorporated between Py and Pt. The observation of a symmetric dc voltage related to ISHE in Pt demonstrates spin-current transmission through the insulating α\alpha-Fe2O3\mathrm{Fe}_2\mathrm{O}_3 spacer over the temperature range 1.6 to 300 K. The temperature dependence of the spin pumping coefficient KsK_s varies systematically with the thickness of α\alpha-Fe2O3\mathrm{Fe}_2\mathrm{O}_3 layer, showing a larger dKs/dTdK_s/dT below a thickness dependent characteristic temperature. The correspondence between T∗T^* and the magnetic response of the Pt/α\alpha-Fe2O3\mathrm{Fe}_2\mathrm{O}_3 sample suggests that this change is associated with magnetic reorientation of α\alpha-Fe2O3\mathrm{Fe}_2\mathrm{O}_3. These results motivate further studies across the Morin-transition regime in epitaxial hematite.

Acknowledgements

This work has been supported in part by the National Science Foundation through grant No. 2309672. Part of this work was also supported by the Center for Energy Efficient Magnonics, an Energy Frontier Research Center funded by the U.S. Department of Energy, Office of Science, Basic Energy Sciences, under Contract No. DE-AC02-76SF00515.

References

  1. [1]A. Brataas, B. van Wees, O. Klein, G. de Loubens, and M. Viret, “Spin insulatronics,” Phys. Rep. 885, 1 (2020).DOI
  2. [2]D. Wesenberg, T. Liu, D. Balzar, M. Wu, and B.L. Zink, “Long-distance spin transport in a disordered magnetic insulator,” Nat. Phys. 13, 987 (2017).DOI
  3. [3]K.S. Rathore, A. Swain, J. Lim, A. Mishra, P. Gupta, L. Yong Heng, J. Luwen, R. Mahendiran, A. Hoffmann, and S. Bedanta, “Magnon-magnon coupling in an all-oxide insulator garnet Y₃Fe₅O₁₂/Tm₃Fe₅O₁₂ heterostructure,” Phys. Rev. Appl. 25, 024036 (2026).
  4. [4]J. Li, Y. Xu, M. Aldosary, C. Tang, Z. Lin, S. Zhang, R. Lake, and J. Shi, “Observation of magnon-mediated current drag in Pt/yttrium iron garnet/Pt(Ta) trilayers,” Nat. Commun. 7, 10858 (2016).DOI
  5. [5]H. Wang, C. Du, P.C. Hammel, and F. Yang, “Antiferromagnetic Spin Transport from Y₃Fe₅O₁₂ into NiO,” Phys. Rev. Lett. 113, 097202 (2014).pubmed.ncbi.nlm.nih.gov/25216003
  6. [6]Y. Pu, G. Shi, C. Zhang, X. Chen, H. Park, and H. Yang, “Crystalline-Dependent Magnon Torques in All-Sputtered Hf/Cr₂O₃/Ferromagnet Heterostructures,” Advanced Materials 37, 2501989 (2025).
  7. [7]Z. Qiu, D. Hou, J. Barker, K. Yamamoto, O. Gomonay, and E. Saitoh, “Spin colossal magnetoresistance in an antiferromagnetic insulator,” Nat. Mater. 17, 577 (2018).
  8. [8]A. Zou, Z. Zhang, Y. Xu, M. Du, X. Wang, S. Chang, and Z. Qiu, “Study of interfacial spin transport with an antiferromagnetic Cr₂O₃ interlayer,” Phys. Rev. B 105, 014427 (2022).DOI
  9. [9]⁹ G.R. Hoogeboom, G.-J.N.S. Nicolaas, A. Alexander, O. Kuschel, J. Wollschläger, I. Ennen, B.J. van Wees, and T. Kuschel, “Role of NiO in the nonlocal spin transport through thin NiO films on Y₃Fe₅O₁₂,” Phys. Rev. B 103, 144406 (2021).DOI
  10. [10]¹⁰ Y. Wang, R. Ramaswamy, and H. Yang, “FMR-related phenomena in spintronic devices,” J. Phys. D Appl. Phys. 51, 273002 (2018).DOI
  11. [11]¹¹ L. Jin, K. Jia, D. Zhang, B. Liu, H. Meng, X. Tang, Z. Zhong, and H. Zhang, “Effect of Interfacial Roughness Spin Scattering on the Spin Current Transport in YIG/NiO/Pt Heterostructures,” ACS Appl. Mater. Interfaces 11, 35458 (2019).DOI
  12. [12]¹² S. Chen, A. Pomar, L. Balcells, Z. Konstantinovic, B. Bozzo, C. Frontera, C. Magen, N. Mestres, and B. Martinez, “Crystallographic Engineering of Spin Transport in Antiferromagnetic NiO Thin Films,” ACS Nano 19, 32170 (2025).DOI
  13. [13]¹³ S. Park, H. Jang, J.-Y. Kim, B.-G. Park, T.-Y. Koo, and J.-H. Park, “Strain control of Morin temperature in epitaxial α-Fe₂O₃ (0001) film,” Europhysics Letters 103, 27007 (2013).
  14. [14]¹⁴ N. Shimomura, S.P. Pati, Y. Sato, T. Nozaki, T. Shibata, K. Mibu, and M. Sahashi, “Morin transition temperature in (0001)-oriented α-Fe₂O₃ thin film and effect of Ir doping,” J. Appl. Phys. 117, 17C736 (2015).DOI
  15. [15]¹⁵ L. Huang, C.F. Li, Y.S. Tang, L. Lin, W.J. Zhai, X.M. Cui, G.Z. Zhou, J.H. Zhang, Z.B. Yan, C. Chen, X.P. Jiang, C.L. Lu, and J.-M. Liu, “Magnetotransport around the Morin transition in α-Fe₂O₃ single crystals,” J. Appl. Phys. 132, 163903 (2022).DOI
  16. [16]¹⁶ J. Lielmezs, and A.C.D. Chaklader, “Magnetic Transitions in Alpha Hematite,” Science (1979). 160, 1137 (1968).
  17. [17]¹⁷ R. Lebrun, A. Ross, O. Gomonay, V. Baltz, U. Ebels, A.-L. Barra, A. Qaiumzadeh, A. Brataas, J. Sinova, and M. Kläui, “Long-distance spin-transport across the Morin phase transition up to room temperature in ultra-low damping single crystals of the antiferromagnet α-Fe₂O₃,” Nat. Commun. 11, 6332 (2020).DOI

Paper details

vlnc:W_mbm7hbtj
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC0-1.0
Usage

views · 1

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