Iridium-manganese (1/3) is a compound characterized by its unique stoichiometry, consisting of one iridium atom to three manganese atoms, represented as IrMn. This intermetallic compound exhibits distinct physical and chemical properties due to the combination of iridium, a rare and highly corrosion-resistant metal, and manganese, known for its magnetic and catalytic properties. The compound typically appears as a gray metallic solid and is notable for its high melting point and density. Iridium itself is the second densest element after osmium and is renowned for its resistance to oxidation and corrosion, while manganese contributes to the compound's catalytic efficiency in various reactions .
The synthesis of iridium-manganese (1/3) can be achieved through various methods:
Recent studies have focused on the interactions between iridium and manganese oxides. These interactions enhance catalytic properties significantly compared to using iridium alone. For instance, when individual iridium atoms are dispersed on manganese oxide surfaces, they maintain high catalytic activity while using substantially less iridium—up to 95% less—without compromising performance . The unique interaction between these two metals is crucial for developing more sustainable catalysts.
Iridium-manganese (1/3) stands out due to its optimized ratio that enhances catalytic activity while minimizing the use of rare materials like iridium, making it a promising candidate for sustainable energy solutions.
The exchange bias phenomenon in ferromagnetic/iridium--manganese (1/3) heterostructures originates from the interfacial exchange coupling between the ferromagnetic layer and the antiferromagnetic iridium--manganese (1/3) film [1]. This coupling mechanism introduces unidirectional anisotropy at the interface, which manifests as a horizontal shift in the magnetization hysteresis loop of the ferromagnetic layer [2]. The unidirectional anisotropy emerges due to the pinning of ferromagnetic spins by the ordered antiferromagnetic moments at the interface, creating an energetic preference for one magnetization direction over the other [1] [3].
The strength of interface coupling depends critically on the magnetic structure of the iridium--manganese (1/3) layer [1] [4]. In chemically ordered L12-iridium--manganese (1/3), the triangular magnetic structure with manganese magnetic moments aligned at 120 degrees to each other in the (111) plane creates specific interfacial spin configurations that determine the coupling strength [5] [1]. The interface exchange energy can be expressed as the product of the interfacial exchange constant and the cosine of the angle between ferromagnetic and antiferromagnetic moments at the interface [6].
Experimental studies have demonstrated that the exchange bias field varies significantly depending on the ferromagnetic layer composition and thickness [7] [8]. In cobalt/iridium--manganese (1/3) systems, exchange bias fields ranging from 158 Oersted in compensated interfaces to over 500 Oersted in optimized configurations have been reported [7]. The unidirectional anisotropy energy density typically reaches values of 10^-2 to 10^-1 erg/cm² at room temperature [8].
Table 1: Structural and Magnetic Properties of Iridium--manganese (1/3)
| Property | Value | Source |
|---|---|---|
| Chemical Formula | IrMn₃ | [9] |
| Crystal Structure (Ordered) | L1₂ (ordered) | [5] [1] |
| Crystal Structure (Disordered) | Face-centered cubic (γ) | [5] [1] |
| Lattice Parameter (L12) | 3.785 Å | [10] |
| Lattice Parameter (γ-phase) | 3.77 ± 0.001 nm | [11] |
| Molecular Weight | 247.16 g/mol | [9] |
| Magnetic Structure (L12) | Triangular (T1) | [5] [1] |
| Magnetic Structure (γ-phase) | Cubic with tilted moments | [5] [1] |
| Néel Temperature (L12) | > 300 K | [11] [1] |
| Magnetic Anisotropy Energy | 3 × 10⁷ J/m³ (theoretical) | [12] [13] |
| Blocking Temperature | > 300 K | [11] [8] |
The interfacial coupling mechanism is further influenced by the degree of compensation at the ferromagnetic/antiferromagnetic interface [4] [14]. In perfectly compensated interfaces, equal numbers of up and down antiferromagnetic spins are present, resulting in zero net interfacial magnetization [6]. However, real interfaces often exhibit partial compensation due to structural imperfections, leading to uncompensated antiferromagnetic moments that directly couple to the ferromagnetic layer [15]. The number of uncompensated spins correlates directly with the exchange bias magnitude, with theoretical models predicting that the exchange bias field is proportional to the uncompensated moment density [15].
The antiferromagnetic domain structure in iridium--manganese (1/3) plays a crucial role in determining the exchange bias properties of heterostructures [16] [3]. The domain configuration affects both the magnitude and thermal stability of the exchange bias through several mechanisms related to the complex magnetic ordering of the antiferromagnetic layer [1] [17].
In chemically ordered L12-iridium--manganese (1/3), the triangular magnetic structure creates eight equivalent magnetic ground states corresponding to different orientations of the triangular spin arrangement [5] [1]. These ground states give rise to antiferromagnetic domains with different spin orientations, and the population of these domains significantly influences the interfacial exchange coupling [16]. The domain wall energy in iridium--manganese (1/3) is exceptionally high due to the large magnetocrystalline anisotropy, estimated at 3 × 10^7 J/m³ from theoretical calculations [12] [13].
Table 2: Exchange Bias Parameters in Ferromagnetic/IrMn₃ Heterostructures
| System | Exchange Bias Field (Oe) | Blocking Temperature (K) | Interface Type | Reference |
|---|---|---|---|---|
| IrMn₃/Fe | Variable | > 300 | Compensated | [1] [4] |
| IrMn₃/CoFe | 100-1000 | > 300 | Partially compensated | [18] [8] |
| IrMn₃/Co | 158 | > 300 | Compensated | [7] |
| IrMn₃/Permalloy | Variable | > 300 | Partially compensated | [11] [19] |
| Co/IrMn₃ (5 nm) | Up to 500 | > 300 | Compensated | [7] |
| IrMn₃/NiFe | Up to 70 | > 300 | Partially compensated | [19] |
The influence of antiferromagnetic domain configurations becomes particularly evident in the facet-dependent behavior of iridium--manganese (1/3) films [20] [16]. Different crystallographic orientations exhibit distinct domain structures and corresponding exchange bias properties [20]. In (001)-oriented films, the antiferromagnetic domains can be manipulated through perpendicular field annealing, leading to significant changes in the exchange bias magnitude [16]. This manipulation is possible because the cubic anisotropy of iridium--manganese (1/3) creates preferential domain orientations that can be controlled through magnetic field treatments [16].
The antiferromagnetic domain configuration also determines the training effects observed in these systems [17] [7]. The domain structure influences how the antiferromagnetic moments respond to the cycling of the ferromagnetic layer, with some domains being more stable than others during magnetization reversal [17]. First-principles calculations have revealed that the relationship between antiferromagnetic order at the interface and heterostructure stability is fundamental to understanding these domain effects [14].
Experimental observations using advanced characterization techniques have shown that the antiferromagnetic domains in iridium--manganese (1/3) exhibit complex multidomain structures [16]. These structures can be manipulated through current-induced switching, demonstrating the interconnection between domain configurations and electrical control of exchange bias [19] [16]. The ability to control domain populations through electrical means opens possibilities for novel spintronic applications based on antiferromagnetic domain manipulation [19].
Training effects in ferromagnetic/iridium--manganese (1/3) heterostructures represent a fundamental aspect of the exchange bias phenomenon, characterized by the systematic variation of magnetic properties during consecutive magnetization cycles [7]. These effects arise from the complex interplay between the ferromagnetic layer dynamics and the antiferromagnetic domain structure evolution during thermal and magnetic field cycling [7] [21].
The training effect in cobalt/iridium--manganese (1/3) bilayers exhibits both athermal and thermal components [7]. The athermal training effect becomes more prominent as the cobalt layer thickness increases, indicating that thicker ferromagnetic layers approach closer to equilibrium conditions according to relaxation models [7]. The training effect constant γ increases systematically with cobalt layer thickness, ranging from 0.12 for 2 nanometer thick cobalt layers to 0.28 for 20 nanometer thick layers [7]. This dependence suggests that the relaxation dynamics of the antiferromagnetic domains are influenced by the magnetic stiffness of the ferromagnetic layer [7].
Table 3: Training Effects and Thermal Stability
| Sample | Training Effect Constant γ | Thermal Component | Temperature Stability | Reference |
|---|---|---|---|---|
| Co(2 nm)/IrMn₃(5 nm) | 0.12 | Dominant | Poor | [7] |
| Co(3 nm)/IrMn₃(5 nm) | 0.15 | Moderate | Fair | [7] |
| Co(5 nm)/IrMn₃(5 nm) | 0.18 | Moderate | Good | [7] |
| Co(8 nm)/IrMn₃(5 nm) | 0.22 | Athermal dominant | Very good | [7] |
| Co(10 nm)/IrMn₃(5 nm) | 0.25 | Athermal dominant | Excellent | [7] |
| Co(20 nm)/IrMn₃(5 nm) | 0.28 | Athermal dominant | Excellent | [7] |
The thermal stability of exchange bias in iridium--manganese (1/3) systems is characterized by exceptionally high blocking temperatures, typically exceeding 300 K for properly optimized heterostructures [11] [8]. The blocking temperature represents the temperature above which the exchange bias disappears due to thermal disruption of the antiferromagnetic order [8] [22]. In iridium--manganese (1/3)/cobalt-iron systems, the blocking temperature shows a monotonic decrease with increasing ferromagnetic layer thickness, reaching up to 29% reduction for 12 nanometer thick ferromagnetic layers [8].
The thermal stability is influenced by several factors including the exchange field from the ferromagnetic layer acting on the antiferromagnetic layer [8]. This exchange field can increase the degree of thermal activation by up to 44%, effectively reducing the energy barriers that stabilize the antiferromagnetic domain structure [8]. The temperature dependence of the exchange bias field follows complex behavior that can be modeled using thermal activation models based on antiferromagnetic grain stability [21].
Finite size effects significantly impact the thermal stability of ultrathin iridium--manganese (1/3) films [23] [24]. Atomistic simulations reveal that antiferromagnetic iridium--manganese (1/3) films show stronger finite size dependence of the Néel temperature compared to equivalent ferromagnets due to spin frustration effects [23]. The disordered γ-iridium--manganese (1/3) phase exhibits dramatic reduction in Néel temperature to below room temperature for films less than 1 nanometer thick [23] [24]. Interfacial intermixing with non-magnetic capping layers further reduces the Néel temperature, with stronger effects observed in the disordered phase compared to the ordered L12 phase [23].
Table 4: Spin-orbit Torque Efficiency Data
| Orientation | Spin Hall Angle | Damping-like Efficiency | Field-like Torque | Reference |
|---|---|---|---|---|
| (001) IrMn₃ | 0.092 ± 0.003 | High | Present | [20] [25] |
| (111) IrMn₃ | ~0.03 | Moderate | Reduced | [20] [25] |
| Polycrystalline IrMn₃ | ~0.05 | Moderate | Moderate | [20] [25] |
| γ-IrMn₃ (001) | 0.16 | Very high | Enhanced | [26] |
| γ-IrMn₃ (111) | 0.05 | Moderate | Moderate | [26] |
The unique triangular antiferromagnetic structure of iridium--manganese (1/3) enables extraordinary spin transport phenomena that are highly dependent on crystalline orientation and current direction. These phenomena arise from the combination of strong spin-orbit coupling, noncollinear magnetic order, and broken time-reversal symmetry in the material's chiral magnetic structure [1] [2].
The spin Hall conductivity in iridium--manganese (1/3) exhibits remarkable dependence on crystalline orientation, with the (001)-oriented films demonstrating significantly enhanced performance compared to (111)-oriented counterparts. Using spin torque ferromagnetic resonance measurements on epitaxial films, researchers have demonstrated effective spin Hall angles reaching 0.35 in (001)-oriented single-crystalline antiferromagnetic iridium--manganese (1/3) thin films [1] [2]. This value represents approximately three times the magnitude observed in (111)-oriented films, where the effective spin Hall angle measures approximately 0.12 [1] [2].
The crystalline orientation dependence originates from the triangular magnetic structure of iridium--manganese (1/3), where manganese atoms are arranged in triangular networks within the (111) planes [2]. In this configuration, manganese magnetic moments align at 120° to each other, forming two distinct antiferromagnetic arrangements that can be transformed into each other through mirror reflection or time reversal operations [2]. The face-centered cubic lattice structure with space group Pm3̄m accommodates this noncollinear arrangement, creating substantial intrinsic spin Hall conductivity that varies dramatically with crystal orientation [2].
Theoretical calculations using density functional theory and Berry curvature analysis reveal the microscopic origin of this orientation dependence [1] [2]. For (111)-oriented films, in-plane electric currents generate considerable spin currents within the (111) plane but produce negligible out-of-plane spin currents. The calculated spin Hall conductivity for in-(111)-plane currents generating out-of-plane spin currents is approximately 10 siemens per centimeter [1]. In contrast, (001)-oriented films exhibit dramatically enhanced out-of-plane spin current generation, with calculated spin Hall conductivity reaching approximately 390 siemens per centimeter when the coordinate system is rotated to align with the [3] direction [1] [2].
Experimental measurements confirm this theoretical prediction through anomalous Hall conductivity measurements, which show values of 8800 siemens per centimeter for the [3] direction compared to 224 siemens per centimeter for the [4] direction [5]. This represents a ratio of approximately 39:1 between the spin Hall conductivities along these crystallographic directions [5]. The enhancement factor of three between (001) and (111) orientations is consistently observed across multiple measurement techniques and sample preparation methods [1] [2].
The intrinsic nature of this orientation dependence is confirmed by the fact that it persists across different film thicknesses and is independent of extrinsic factors such as interface roughness or chemical intermixing [1]. The spin diffusion length in iridium--manganese (1/3) is estimated to be less than 10 angstroms, indicating that the observed effects are primarily interfacial phenomena rather than bulk transport effects [1] [2].
The magnetic spin Hall effect represents a novel transport phenomenon observed specifically in noncollinear antiferromagnets with chiral spin structures. In (001)-oriented iridium--manganese (1/3) films, this effect manifests as additional spin current generation that depends on the magnetic order parameter and exhibits distinct symmetry properties compared to conventional spin Hall effects [6] [5].
Ferromagnetic resonance measurements on iridium--manganese (1/3)/nickel80iron20 bilayers reveal the presence of magnetic spin Hall effects through their characteristic angular dependence [6] [5]. When direct current is applied along the [3] crystallographic direction, significant modulation of ferromagnetic resonance linewidth occurs for magnetic fields applied along the [3] direction, while minimal changes are observed for fields applied along the [4] direction [5]. This anisotropic behavior provides direct evidence for the magnetic spin Hall effect, which exhibits different symmetries compared to conventional charge-to-spin conversion mechanisms [5].
The microscopic origin of the magnetic spin Hall effect lies in the chiral nature of the triangular antiferromagnetic structure [6] [5]. The manganese atoms form Kagome lattices in the {111} planes, with magnetic moments pointing either toward or away from each triangular arrangement [5]. These two configurations, designated as Lattice 1 and Lattice 2, are chiral images of each other and can be interconverted through time reversal operations [5]. Under time reversal symmetry, the magnetic spin Hall effect is odd, meaning that Lattice 1 and Lattice 2 exhibit identical magnetic spin Hall conductivity but opposite anomalous Hall conductivity [5].
Quantitative analysis of the magnetic spin Hall effect yields an effective magnetic spin Hall angle of 0.33 ± 0.02 for currents applied along the [3] direction [5]. This value is extracted from ferromagnetic resonance linewidth analysis using the Landau-Lifshitz-Gilbert-Slonczewski equation, which incorporates additional terms for magnetic spin Hall contributions [5]. The measurement utilizes current densities of 2.5×10⁴ amperes per square centimeter and accounts for the electrical conductivity of 8.2×10³ ohm⁻¹ centimeter⁻¹ along the [3] direction [5].
The angular dependence of the magnetic spin Hall effect follows a characteristic sin(±2φ) relationship, where φ represents the angle between the applied magnetic field and crystallographic axes [5]. This functional form arises from the symmetry breaking associated with the chiral magnetic structure and provides a clear signature distinguishing magnetic spin Hall effects from other transport phenomena [5]. The maximum modulation amplitude reaches 40% of the baseline damping value, demonstrating the substantial magnitude of this effect [5].
Temperature dependence studies confirm that the magnetic spin Hall effect persists at room temperature, making it relevant for practical spintronic applications [5]. The effect shows strong correlation with the blocking temperature of the antiferromagnetic structure, indicating its intimate connection to the magnetic order parameter [5]. This temperature dependence provides additional confirmation that the observed phenomena arise from magnetic rather than purely electronic origins [5].
Current-induced modulation of Gilbert damping in iridium--manganese (1/3)-based systems provides a powerful mechanism for electrical control of magnetization dynamics. This phenomenon manifests as significant changes in the magnetic damping parameter of adjacent ferromagnetic layers when direct current flows through the antiferromagnetic iridium--manganese (1/3) [6] [5].
Systematic ferromagnetic resonance studies on iridium--manganese (1/3)/nickel80iron20 bilayers demonstrate remarkable damping modulation capabilities [5]. The baseline Gilbert damping parameter for nickel80iron20 in contact with iridium--manganese (1/3) measures (3.94 ± 0.02)×10⁻² for both [3] and [4] crystallographic directions, representing a substantial enhancement factor of approximately 4.5 compared to isolated nickel80iron20 films with damping parameter (8.66 ± 0.03)×10⁻³ [5]. This enhancement arises from interfacial spin mixing between the antiferromagnetic and ferromagnetic layers, characterized by an effective spin mixing conductance of (1.6 ± 0.1)×10¹⁶ inverse square centimeters [5].
When direct current is applied along the [3] direction, the Gilbert damping exhibits pronounced current dependence that varies significantly with magnetic field orientation [5]. For magnetic fields applied along the [3] direction, current-induced damping modulation reaches a maximum of 41% of the baseline value [5]. This modulation follows a symmetric dependence on current polarity, with both positive and negative currents of ±1 milliampere producing equivalent magnitude changes but opposite signs [5]. The current density threshold for maximum modulation corresponds to approximately 2.5×10⁴ amperes per square centimeter [5].
The crystallographic direction dependence of damping modulation provides crucial insight into the underlying mechanisms [5]. When magnetic fields are applied along the [4] direction while maintaining current flow along [3], minimal damping modulation occurs [5]. This anisotropic behavior confirms that the observed effects arise from the magnetic spin Hall effect rather than conventional spin-orbit torques or heating effects [5]. The angular dependence of damping modulation follows the relationship {[α [4] - α(φ)]/α [4]}×100% = ΔαMSHE sin(±2φ), where ΔαMSHE = (40 ± 1)% [5].
Three distinct physical mechanisms contribute to the current-induced damping modulation [6] [5]. The first mechanism involves facet-dependent contributions from antiferromagnetic domains containing uncompensated spins at the interface [6]. These domains create local magnetic fields that couple to the ferromagnetic layer magnetization, modulating the effective damping through exchange spring effects [6]. The second mechanism arises from facet-independent bulk spin-orbit coupling within the iridium--manganese (1/3) layer, generating conventional spin currents through the standard spin Hall effect [6]. The third mechanism, which dominates the observed anisotropy, stems from the magnetic spin Hall effect associated with the chiral triangular magnetic structure [6] [5].
The temporal dynamics of damping modulation reveal additional complexity in the underlying physics [5]. Frequency-swept measurements demonstrate that the current-induced changes persist across the entire ferromagnetic resonance spectrum, indicating that the modulation affects the intrinsic damping parameter rather than just specific resonance modes [5]. The linewidth variation follows the standard relationship ΔH = ΔH₀ + (α/γ)f, where the current-dependent contribution appears as an additive term to the Gilbert damping parameter [5].
Interfacial transparency effects play a crucial role in determining the magnitude of damping modulation [6]. Studies with thin copper or gold spacer layers inserted between iridium--manganese (1/3) and nickel80iron20 demonstrate systematic reduction in damping modulation as spacer thickness increases [6]. For gold spacers, the effect decreases to facet-independent values when the spacer thickness reaches approximately 10 angstroms [6]. Copper spacers produce more rapid decay due to stronger exchange interlayer coupling, but some facet dependence persists even at 10 angstrom thickness [6]. These observations confirm that the magnetic spin Hall effect requires intimate interfacial contact between the antiferromagnetic and ferromagnetic layers [6].