--- title: 'Fe3GaTe2: Layered vdW Ferromagnet' url: https://www.emergentmind.com/topics/fe3gate2 type: topic --- # Fe3GaTe2: Layered vdW Ferromagnet Searching arXiv for recent Fe3GaTe2 papers to ground the article in current literature. arxiv_search query="Fe3GaTe2" max_results=10 sort_by="submittedDate" Fe\(_3\)GaTe\(_2\) is a layered van der Waals ferromagnetic metal whose crystal structure is generally refined in the hexagonal space group \(P6_3/mmc\), with Te-terminated Fe–Ga slabs separated by weak interlayer gaps of about \(0.78\)–\(0.81\) nm. Across bulk crystals, exfoliated flakes, patterned films, and epitaxial heterostructures, it is distinguished by ferromagnetism at or above room temperature, strong perpendicular magnetic anisotropy, metallic transport, pronounced anomalous Hall response, and compatibility with a broad range of spintronic architectures, including magnetic tunnel junctions, current-switched Hall bars, and engineered skyrmion textures [2406.02346] [2303.05163]. ## 1. Crystal chemistry and symmetry Fe\(_3\)GaTe\(_2\) is consistently described as a layered hexagonal compound in \(P6_3/mmc\), built from Te–Fe\(_3\)Ga–Te or equivalently Te–Fe–Ga–Fe–Te structural units stacked along the \(c\)-axis through van der Waals coupling. Reported lattice parameters cluster near \(a=b=3.986\)–\(4.12\) Å and \(c=16.088\)–\(16.41\) Å, with single-layer or quintuple-layer thicknesses near \(0.78\)–\(0.81\) nm [2406.02346] [2412.04872] [2411.16463]. X-ray diffraction on bulk crystals and thin films commonly shows dominant \((00\ell)\) reflections, consistent with \(c\)-axis-oriented layering [2410.18689] [2505.06128]. Several studies resolve two inequivalent Fe environments. One structural description assigns Fe1 to \(2c\) and Fe2 to \(4f\), with Ga at \(2b\) and Te at \(4f\); another identifies Fe1 on \(4e\), Fe\(_2\) on \(2c\), and occasional Fe\(_3\) intercalants in the van der Waals gap [2405.07156] [2507.23068]. Site-resolved electronic structure calculations further indicate different local moments on the inequivalent Fe sites, with calculated values \(m(\mathrm{Fe1})=1.33\,\mu_B\) and \(m(\mathrm{Fe2})=1.98\,\mu_B\) in one study, implying that crystallographic inequivalence is magnetically consequential [2405.07156]. Stoichiometry is close to \(3{:}1{:}2\), but compositional deviations are repeatedly observed. Reported EDS ratios include Fe:Ga:Te \(\simeq 3.14{:}1.00{:}1.97\), \(2.93{:}1.07{:}2.00\), and \(2.8{:}1{:}1.9\) [2406.02346] [2402.14573] [2407.16924]. These deviations are not merely chemical detail: Fe deficiency, interlayer occupation, and site disorder recur in interpretations of Dzyaloshinskii–Moriya interaction, exchange bias, and skyrmion stabilization. A central structural development is the distinction between global and local inversion symmetry. Single-crystal X-ray diffraction refines Fe\(_3\)GaTe\(_2\) as globally centrosymmetric \(P6_3/mmc\), yet HAADF-STEM reveals off-center displacements of Fe\(_2\) and Ga relative to the mirror plane between Te layers, with \(\Delta z_{\mathrm{Fe2}}\simeq 0.16\pm0.06\) Å and \(0.17\pm0.07\) Å on the two sides, and \(\Delta z_{\mathrm{Ga}}\simeq 0.05\pm0.08\) Å and \(0.04\pm0.07\) Å [2507.23068]. This suggests that nanoscale inversion breaking can coexist with average centrosymmetry and provides a structural route for local chiral interactions in a nominally centrosymmetric host. ## 2. Magnetic order, anisotropy, and phase structure The defining magnetic property of Fe\(_3\)GaTe\(_2\) is ferromagnetic order persisting to at least room temperature in many specimens. Reported Curie temperatures depend on sample thickness and probe, ranging from \(T_C \gtrsim 320\) K in anomalous-Hall measurements on tunnel-junction electrodes, to \(333\)–\(340\) K in bulk and exfoliated flakes, to \(\sim 360\)–\(366\) K in magnetometry and local quantum-sensing experiments, and up to \(371\pm0.7\) K with extrapolated \(390\)–\(400\) K in epitaxial films on graphene/SiC [2303.05163] [2402.14573] [2406.02346] [2505.06128]. These values collectively establish Fe\(_3\)GaTe\(_2\) as an above-room-temperature van der Waals ferromagnet, although the exact \(T_C\) is sample-dependent. Perpendicular magnetic anisotropy is equally robust. Square out-of-plane magnetization and anomalous-Hall loops are observed up to room temperature and beyond, while in-plane loops are much weaker, indicating a \(c\)-axis easy axis [2406.02346] [2303.05163]. Quantified anisotropy parameters include \(K_u \simeq 4.15\times10^5\) J/m\(^3\) at \(300\) K from \(B_{\rm sat}\) and \(M_{\rm sat}\), \(K_u>3.9\times10^5\) J/m\(^3\) in thin flakes used for orbital-torque switching, and anisotropy energies \(K\simeq0.40\) meV/Fe at \(2\) K and \(\simeq0.19\) meV/Fe at \(300\) K from thermodynamic analysis [2402.14573] [2412.04872] [2507.23068]. In one device, the effective anisotropy field \(H_K\) increases from \(4.02\) T at \(300\) K to \(5.33\) T at \(225\) K [2412.04872]. Reported magnetization values vary by geometry and sample quality. At \(300\) K, an out-of-plane \(M\)–\(H\) loop yielded \(M_s\simeq(0.3\)–\(0.5)\,\mu_B\) per Fe by graphical estimate in one study, whereas another reported \(M_{\rm sat}(300\,\mathrm{K})\simeq33.26\) emu/g \(\simeq0.98\,\mu_B/\)Fe, and a Raman/spin-phonon study reported \(M_s\simeq48.8\) emu/g \(\simeq1.43\,\mu_B/\)Fe at room temperature [2406.02346] [2402.14573] [2411.19533]. The variation likely reflects differences in stoichiometry, thickness, and analysis protocols. A notable revision of the magnetic ground-state picture was provided by combined magnetometry and neutron diffraction. Under \(H=0.35\) T, a sharp first-order-like drop in magnetization occurs at \(T_{\mathrm{ferri}}\simeq161\) K, and neutron scattering indicates that reducing the field from \(1.0\) T to \(0.35\) T drives a transition from a nearly polarized ferromagnetic state to an intralayer ferrimagnetic configuration with antiparallel Fe1 and Fe2 moments [2507.23068]. This directly challenges the interpretation of the low-temperature state as a glassy mixture of ferro- and ferrimagnetic domains and instead supports globally ferrimagnetic order at low field. ## 3. Electronic transport, Hall response, and correlated scattering Fe\(_3\)GaTe\(_2\) is metallic in both bulk-like and thin-flake transport devices. Longitudinal resistance or resistivity generally decreases on cooling from room temperature, as seen in Hall bars and tunnel-junction electrodes [2303.05163] [2506.02152]. In several samples a low-temperature upturn appears; one work assigns an upturn near \(20\) K to Kondo scattering, whereas another resolves a complete orbital two-channel Kondo sequence in an \(\sim81.4\) nm flake [2406.02346] [2410.18689]. The anomalous Hall effect is a persistent hallmark. In the ferromagnetic regime the Hall response is commonly written as \[ \rho_{xy}=R_0\mu_0H+R_sM, \] with the anomalous term dominating so that \(\rho_{xy}\propto M_z\) in perpendicular geometry [2406.02346]. Square Hall hysteresis survives from cryogenic temperature to about \(340\)–\(360\) K in exfoliated devices and up to \(\sim400\) K in epitaxial films [2406.02346] [2505.06128]. Thickness-controlled transport studies of patterned films identified a giant anomalous Hall effect with \(\rho_{\mathrm{AH}}\) up to \(\sim100\,\mu\Omega\cdot\)cm at low temperature and Hall angle up to \(\sim0.18\) at \(5\) K and \(\sim0.10\) at \(300\) K [2506.02152]. The scaling \[ \rho_{\mathrm{AH}}=a\,\rho_{xx}+b\,\rho_{xx}^2 \] was found to separate a positive, temperature-independent skew-scattering term from negative side-jump and intrinsic contributions. Reported coefficients include \(a=0.35\pm0.04\) and \(B\simeq-6.07\times10^{-4}\,\mu\Omega^{-1}\,\mathrm{cm}^{-1}\), while \(\sigma_{\mathrm{AH}}^{\mathrm{int}}\propto \rho_0^{-2}\) was interpreted as characteristic of the dirty-metal regime [2506.02152]. The most detailed correlated-scattering study reported three transport regimes below \(T_C\approx340\) K: a logarithmic Kondo regime between \(T_0\simeq30\) K and \(T_K\simeq9.3\) K, \[ \Delta R(T)\simeq-\alpha\ln(T/T_0), \] a non-Fermi-liquid regime between \(T_D\simeq1\) K and \(T_K\), \[ \Delta R(T)\simeq-\beta T^{1/2}, \] and a Fermi-liquid regime below \(T_D\), \[ \Delta R(T)\simeq-\gamma T^2. \] The invariance of \(T_0\), \(T_K\), \(T_D\), and \(\alpha,\beta,\gamma\) up to \(\mu_0H=9\) T was taken as evidence for an orbital two-channel Kondo effect generated by nonmagnetic two-level systems rather than magnetic impurities [2410.18689]. Hall anomalies do not have a single interpretation across the literature. Some studies resolve a room-temperature topological Hall component \(\rho_{xy}^{T}\sim0.18\)–\(0.2\,\mu\Omega\cdot\)cm near \(0.15\) T in skyrmion-hosting samples [2407.16924], whereas the two-channel Kondo study reports only ordinary and anomalous Hall effects and attributes certain antisymmetric features to geometric AHE artifacts rather than topological Hall response [2410.18689]. This indicates that unconventional Hall signatures in Fe\(_3\)GaTe\(_2\) are strongly contingent on magnetic texture, field geometry, and sample microstructure. ## 4. Skyrmions and real-space topological control Fe\(_3\)GaTe\(_2\) is one of the few van der Waals ferromagnets in which skyrmionic textures have been created, imaged, and manipulated near room temperature. In Lorentz TEM, the field evolution at \(T\simeq295\) K proceeds from labyrinthine Néel-type domains at \(H=0\) Oe, to the appearance of skyrmion-like bubbles around \(1338\) Oe, mixed stripe-worm-skyrmion states near \(1574\) Oe, and a uniform ferromagnetic state near \(1998\) Oe [2402.14573]. In field-cooled MFM experiments, a skyrmion lattice appears for \(0.03\,\mathrm{T}\lesssim H_{\mathrm{FC}}\lesssim0.08\,\mathrm{T}\), whereas lower fields favor labyrinthine domains and higher fields yield single-domain saturation [2407.16924]. Reported skyrmion sizes fall in the submicron range. L-TEM gives typical diameters \(d_{\mathrm{sk}}\simeq150\)–\(200\) nm and lattice constants \(a_{\mathrm{sk}}\simeq250\)–\(350\) nm [2402.14573]. MFM on \(\sim200\) nm flakes reports diameters of \(200\)–\(300\) nm, with density peaking near \(0.5\,\mu\mathrm{m}^{-2}\) at \(H\simeq0.15\) T [2407.16924]. The origin of skyrmion stability remains an active interpretive point. One study models the texture using strong PMA together with an interfacial DMI \(D\simeq0.4\) mJ/m\(^2\) generated by a lightly oxidized surface [2402.14573]. Another attributes skyrmion formation to competition between dipolar stray-field energy and a smaller DMI, \(D\simeq0.1\)–\(0.2\) mJ/m\(^2\), induced by Fe vacancies and local inversion breaking [2407.16924]. A later structural study strengthens the latter picture by directly observing local Fe\(_2\) and Ga off-centering and correlating it with Néel skyrmions in a globally centrosymmetric lattice [2507.23068]. Taken together, these results suggest that the relevant chirality is not imposed by global bulk symmetry but emerges from local inversion-symmetry breaking and defect-sensitive interfaces. Real-space manipulation is unusually advanced. MFM-tip stray fields of \(\sim300\) Oe in one setup and \(\sim0.05\)–\(0.1\) T at \(100\) nm lift height in another were used to write and erase skyrmions reproducibly [2402.14573] [2407.16924]. Under \(H_{\mathrm{bias}}=0.05\) T, a single scan can nucleate \(\sim10\)–\(20\) skyrmions, while \(5\)–\(7\) passes produce a near-perfect hexagonal lattice with density \(\sim0.4\,\mu\mathrm{m}^{-2}\) [2407.16924]. Erasure can be induced thermally, by field cycling, or by repeated tip passes at higher bias field. Topological engineering extends beyond isolated skyrmions. Regions of opposite topological charge, \(S=+1\) and \(S=-1\), have been “painted” to form topological skyrmion junctions. In a \(10\,\mu\mathrm{m}\times34\,\mu\mathrm{m}\) channel, increasing the number of junctions from \(0\) to \(16\) reduces the two-probe resistance from \(\sim780\,\Omega\) to \(\sim520\,\Omega\), with reversible changes upon erasing and rewriting [2407.16924]. The density of skyrmions also exhibits thermal hysteresis that tracks the ferromagnetic-to-ferrimagnetic transition, indicating that skyrmion stability is coupled to the evolving magnetic phase balance [2507.23068]. ## 5. Lattice dynamics, Raman spectroscopy, and local magnetic probes Raman spectroscopy has established Fe\(_3\)GaTe\(_2\) as a magnetically active lattice system, though mode assignments are not identical across studies. Under \(532\) nm excitation, one experiment identified sharp peaks at \(125.8\) cm\(^{-1}\) and \(143.0\) cm\(^{-1}\), assigned to \(A_{1g}\) and \(E_{2g}\), respectively [2406.02346]. A temperature- and pressure-dependent study instead assigned \(126.0\) cm\(^{-1}\) to \(E_{2g}\) and \(143.5\) cm\(^{-1}\) to \(A_{1g}\) [2411.16463]. A thickness-dependent investigation further resolved two out-of-plane modes, \(A_{1g}^1\) and \(A_{1g}^2\), with bulk frequencies \(\sim103.2\) and \(\sim125.5\) cm\(^{-1}\), both blue-shifting as thickness decreases [2411.19533]. The coexistence of these assignments indicates that Raman interpretation is sensitive to sample thickness, spectral window, and the specific vibrational manifold under analysis. Spin-phonon coupling has been reported independently in two forms. One study extracted a room-temperature coupling strength \(\lambda\simeq0.28\) cm\(^{-1}\) from the deviation of the \(A_{1g}^2\) mode from an anharmonic model below \(T_C\), using \(\Delta\omega_{\mathrm{sp-ph}}(T)\simeq \lambda\langle S_i\!\cdot\!S_j\rangle\) [2411.19533]. Another, using the in-plane \(E_{2g}\) mode and a broader temperature-pressure phase space, obtained \(\lambda_{\mathrm{sp}}\approx0.81\) cm\(^{-1}\) at \(300\) K from \(\Delta\omega_E\simeq0.25\) cm\(^{-1}\) and \(\langle S_i\!\cdot\!S_j\rangle\approx0.31\) [2411.16463]. Both results support direct coupling between lattice vibrations and the magnetic order parameter at room temperature. Phonon calculations reinforce this conclusion. First-principles phonon dispersions show no imaginary modes, confirming dynamical stability [2411.16463]. Comparison of ferromagnetic and nonmagnetic interlayer configurations reveals FM-induced phonon softening, reduced band gaps in the phonon spectrum, and red-shifted partial phonon DOS, all of which were interpreted as fingerprints of spin-phonon coupling [2411.19533]. A distinct line of work uses quantum sensing rather than optical spectroscopy. Shallow PL6 divacancy spins in 4H-SiC, approximately \(40\) nm below the surface and protected from direct contact by an hBN capping layer thinner than \(10\) nm, were employed for noninvasive in situ magnetic detection of Fe\(_3\)GaTe\(_2\) at room temperature [2406.02346]. The PL6 center is a spin-1 defect with zero-field splitting \(D\simeq1351\) MHz and ODMR resonances \[ \nu_\pm=D\pm\gamma_e B_\parallel . \] By comparing probe and reference positions under \(B_0\simeq200\) G, the local stray field of the Fe\(_3\)GaTe\(_2\) flake was extracted as a function of temperature and field, yielding \(T_C\sim360\) K and increasing magnetization with external field [2406.02346]. Spin relaxometry, with \[ C(t)\propto e^{-t\Gamma},\qquad \Gamma=1/T_1, \] showed a pronounced peak in \(\Gamma_{\mathrm{FGT}}(T)\) near \(T_C\), consistent with enhanced magnetic fluctuations and longitudinal susceptibility at the Curie point. ## 6. Devices, heterostructures, and materials tuning Fe\(_3\)GaTe\(_2\) has rapidly become a device material rather than only a model magnet. The range of reported platforms is summarized below. | Platform | Key reported result | Source | |---|---|---| | Fe\(_3\)GaTe\(_2\)/WS\(_2\)/Fe\(_3\)GaTe\(_2\) MTJ | TMR up to \(213\%\) at \(10\) K and \(\sim11\%\) at \(300\) K | [2303.05163] | | Fe\(_3\)GaTe\(_2\)/hBN/Fe\(_3\)GaTe\(_2\) MTJ | TMR inversion at \(V_{\mathrm{inv}}\approx0.625\pm0.025\) V | [2501.06063] | | Single-material Fe\(_3\)GaTe\(_2\) Hall bar | \(J_{\mathrm{sw}}=4.7\times10^5\)–\(7.7\times10^5\) A/cm\(^2\) at \(300\) K | [2308.12710] | | Fe\(_3\)GaTe\(_2\)/Ti | Orbital-torque switching with \(J_{\mathrm{sw}}(300\,\mathrm{K})\simeq1.6\times10^6\) A/cm\(^2\) | [2412.04872] | | Epitaxial Fe\(_3\)GaTe\(_2\)/graphene/SiC | Continuous films with \(T_C\) reaching \(\sim400\) K | [2505.06128] | All-2D tunnel junctions demonstrate that the material retains useful spin polarization in vertical transport. In Fe\(_3\)GaTe\(_2\)/WS\(_2\)/Fe\(_3\)GaTe\(_2\) heterojunctions, the TMR obeys \[ \mathrm{TMR}=\frac{R_{AP}-R_P}{R_P}, \] reaching \(213\%\) for a \(3.5\) nm barrier at \(10\) K, corresponding to \(P\simeq72\%\) in the modified Jullière analysis, and remaining observable at \(300\) K with bias currents down to \(10\) nA [2303.05163]. In hBN-barrier junctions, the TMR changes sign reproducibly at \(0.625\pm0.025\) V across devices and temperatures, consistent with bias access to higher-energy spin-resolved DOS structure; a second inversion occurs near \(2.5\)–\(3\) V in some devices [2501.06063]. Lateral devices reveal unusually efficient electrical switching. In single-material Hall bars, intrinsic spin-orbit torque lowers the effective anisotropy barrier according to \[ \Delta E(J)=\tfrac12 M_s\left(K_z-|\Gamma_0J|\right), \] and the coercive field decreases approximately linearly with current density. At \(300\) K, room-temperature nonvolatile switching is achieved with \(J_{\mathrm{sw}}=4.7\times10^5\)–\(7.7\times10^5\) A/cm\(^2\), writing currents around \(\pm1\) mA, and power densities \(1.5\times10^{13}\)–\(2.0\times10^{13}\) W/m\(^3\) [2308.12710]. In Fe\(_3\)GaTe\(_2\)/Ti, the current-induced torque is interpreted as orbital torque rather than conventional spin Hall torque, aided by the large orbital Hall conductivity of Ti, \(\sigma_{\mathrm{OH}}(\mathrm{Ti})\approx4.6\times10^3(\hbar/e)\cdot(\mathrm{S/cm})\), and strong spin–orbit correlation in Fe\(_3\)GaTe\(_2\) [2412.04872]. Exchange bias adds another control axis. Thin-layer devices exhibit robust exchange bias with blocking temperature \(T_B\simeq280\) K, substantially above previous 2D-magnet records. The bias sign and magnitude can be tuned isothermally by changing the field sweep range rather than relying on a conventional FM/AFM interface [2406.02260]. The proposed mechanism is an exchange spring between soft ferromagnetic regions and hard ferromagnetic defect centers with higher coercivity. Large-area growth is no longer limited to stacked flakes. Molecular beam epitaxy on epitaxial graphene/SiC produces continuous Fe\(_3\)GaTe\(_2\) films of \(6\), \(10\), and \(32\) nm thickness, with atomically sharp vdW interfaces, rms roughness \(\simeq0.7\) nm over \(1\times1\,\mu\mathrm{m}^2\), and room-temperature magnetic domains visible by scanning NV microscopy [2505.06128]. Anomalous Hall and XMCD measurements in these films support \(T_C\gtrsim400\) K and strong PMA. Materials tuning is correspondingly broad. Ni substitution proceeds in an ordered sequence: interlayer Ni\(_3\) occupation for \(x<0.10\), progressive replacement of Fe2 for \(0.10