Parkin et al. (Nature Materials 3, 862–867, 2004) reported sputter-deposited magnetic tunnel junctions with (100)-textured MgO barriers showing ~220% room-temperature tunnelling magnetoresistance (TMR), ~85% tunnelling spin polarization, and thermal stability to 400°C, and raised five issues for further study: the discrepancy between transport-derived barrier heights (~1.1–1.7 eV) and the ~3.5 eV valence-band offset from photoemission; the mechanism by which annealing enhances TMR and spin polarization; the absence of predicted MgO-thickness scaling of TMR; whether the very large TMR values predicted for ideal epitaxial barriers could be achieved in manufacturable sputtered stacks; and the quantitative relation between spectroscopically measured spin polarization and device TMR beyond the Jullière model. Three of these five issues have since been substantially resolved: room-temperature TMR of 604% (single barrier, 1144% at 5 K) and 1056% (double barrier) has been achieved in sputtered CoFeB/MgO stacks; oscillatory and monotonic MgO-thickness dependence has been observed in epitaxial and sputtered junctions; and the annealing mechanism (bcc CoFeB crystallization templated by MgO, boron segregation, interdiffusion control) is largely understood. The surviving open core is twofold: a quantitative microscopic census of defect states in sputtered MgO barriers that reconciles the low transport barrier heights with the photoemission band offset and accounts for their effect on TMR, and a predictive, quantitative theory connecting tunnelling spin polarization and TMR in real MgO junctions, including bias-dependent inelastic processes and finite-temperature magnon effects.

Background

A magnetic tunnel junction (MTJ) is a two-terminal device consisting of two ferromagnetic electrodes separated by a thin insulating tunnel barrier; its resistance depends on the relative magnetization orientation of the electrodes. The tunnelling magnetoresistance (TMR) is conventionally defined as TMR=(RAP−RP)/RP\mathrm{TMR}=(R_{\mathrm{AP}}-R_{\mathrm{P}})/R_{\mathrm{P}}, where RAPR_{\mathrm{AP}} and RPR_{\mathrm{P}} are the resistances in the antiparallel and parallel magnetic configurations. MTJs are the storage cell of magnetoresistive random-access memory (MRAM) and the read element of hard-disk sensors. The simplest theory, due to Jullière (1975), relates TMR to the spin polarizations P1,P2P_1,P_2 of the two electrode–barrier interfaces as TMR=2P1P2/(1−P1P2)\mathrm{TMR}=2P_1P_2/(1-P_1P_2), treating the barrier as spin-neutral.

First-principles calculations by Butler, Zhang, Schulthess and MacLaren (2001) and by Mathon and Umerski (2001) predicted that perfectly ordered epitaxial Fe/MgO/Fe(001) junctions should show TMR of hundreds to thousands of percent: the crystalline MgO(001) barrier acts as a symmetry filter, transmitting the Bloch states of Δ1\Delta_1 symmetry (and, with different efficiency, other symmetries) with strongly spin- and symmetry-dependent evanescent decay lengths, while amorphous alumina barriers of the pre-2004 era were limited to about 70% at room temperature. Simultaneous experiments then established the effect: Yuasa et al. (2004) achieved 180% at room temperature in fully epitaxial single-crystal Fe/MgO/Fe junctions, observing oscillations of TMR with barrier thickness that demonstrate coherent tunnelling, while Parkin et al. (2004) achieved ~220% at room temperature and ~300% at low temperature in sputter-deposited, industrially compatible polycrystalline stacks on amorphous underlayers, whose MgO barriers were nevertheless highly (100)-textured with CoFe electrodes.

Parkin et al. (2004) measured the tunnelling spin polarization (TSP) of their CoFe/MgO interfaces by superconducting tunnelling spectroscopy — a low-temperature technique in which a superconducting counterelectrode (there an Si-doped Al layer) analyzes the spin-resolved density of states of the tunnelling current — obtaining ~85% after annealing, then a record for conventional ferromagnetic metals. They extracted effective MgO barrier heights of ~1.1–1.7 eV by fitting current–voltage curves with the Simmons formula (a standard model for tunnelling through a rectangular barrier, which assumes free-electron-like carriers), whereas valence-band photoemission placed the MgO valence-band edge ~3.5 eV below the ferromagnet Fermi level. They also reported that annealing to ~350–425°C strongly increased both TMR and TSP, that TMR showed no pronounced dependence on MgO thickness, that the TMR fell to half its zero-bias value at bias voltages of ~0.3–0.6 V, and they posed five questions for further study: the microscopic nature, energy distribution and density of the defect states that reconcile the transport and photoemission barrier heights; the mechanism of the annealing enhancement; the reason the theoretically predicted MgO-thickness scaling of TMR was absent; whether the very large TMR predicted for ideal epitaxial barriers could be realized in manufacturable, thermally robust sputtered stacks; and the quantitative relation between spectroscopically measured TSP and device TMR, including bias-dependent inelastic processes, imperfect antiparallel alignment, and corrections to the Jullière estimate.

Subsequent work transformed the field. Sputtered CoFeB/MgO junctions — in which an amorphous CoFeB electrode is deposited first and crystallizes on annealing into bcc CoFeB epitaxially matched to the (001)-textured MgO, as established by Hayakawa et al. (2005, 2006), Lee et al. (2006), Ikeda et al. (2008) and microstructural studies such as Kodzuka et al. (2012) — became the industrial workhorse, enabling spin-transfer-torque MRAM (STT-MRAM). The theory of oxygen-vacancy (F center, i.e. an oxygen vacancy occupied by one or more electrons) defect states in MgO was developed by Velev et al. (2007), and defect-mediated resonant transport through MgO was demonstrated by Yang, Yang and Parkin (2007).

Problem Statement

Of the five issues raised by Parkin et al. (2004) on MgO(100)-barrier magnetic tunnel junctions, three have been substantially settled by later work (the realization of hundreds-of-percent room-temperature TMR in manufacturable, thermally robust sputtered stacks; the MgO-thickness dependence of TMR; and the mechanism of the annealing enhancement), leaving a precise surviving core that this problem states:

(1) Determine the microscopic defect population of sputtered, (100)-textured MgO tunnel barriers and quantify its transport consequences, establishing (1a) the atomic identity, energy distribution, and volume density of the defect states — such as oxygen vacancies (F centers), impurities, and grain-boundary states; (1b) a quantitative reconciliation of the low effective barrier heights extracted from transport (about 1.1–1.7 eV) with the valence-band offset of about 3.5 eV below the ferromagnet Fermi level found by photoemission; and (1c) a quantitative account of how these defects modify the resistance-area product and the bias dependence of TMR.

(2) Establish a quantitative, predictive relation between the tunnelling spin polarization measured by spectroscopy (superconducting tunnelling spectroscopy, spin-torque measurements) and the TMR of the corresponding MgO-barrier devices, going beyond the Jullière formula to capture the real-device effects it omits — such as bias-dependent inelastic processes (magnon- and phonon-assisted tunnelling), finite-temperature spin filtering, and imperfect antiparallel magnetic alignment — for realistic (textured, defect-bearing, annealed) MgO barriers, and to predict the achievable TMR magnitude of such devices.

References

  1. S. S. P. Parkin, C. Kaiser, A. Panchula, P. M. Rice, B. Hughes, M. Samant, S.-H. Yang (2004). Giant tunnelling magnetoresistance at room temperature with MgO (100) tunnel barriers. Nature Materials 3, 862–867. DOI 10.1038/nmat1256. https://doi.org/10.1038/nmat1256
  2. M. Julliere (1975). Tunneling between ferromagnetic films. Physics Letters A 54, 225–226. DOI 10.1016/0375-9601(75)90174-7. https://doi.org/10.1016/0375-9601(75)90174-7
  3. W. H. Butler, X.-G. Zhang, T. C. Schulthess, J. M. MacLaren (2001). Spin-dependent tunneling conductance of Fe|MgO|Fe sandwiches. Physical Review B 63, 054416. DOI 10.1103/PhysRevB.63.054416. https://doi.org/10.1103/PhysRevB.63.054416
  4. J. Mathon, A. Umerski (2001). Theory of tunneling magnetoresistance of an epitaxial Fe/MgO/Fe(001) junction. Physical Review B 63, 220403. DOI 10.1103/PhysRevB.63.220403. https://doi.org/10.1103/PhysRevB.63.220403
  5. S. Yuasa, T. Nagahama, A. Fukushima, Y. Suzuki, K. Ando (2004). Giant room-temperature magnetoresistance in single-crystal Fe/MgO/Fe magnetic tunnel junctions. Nature Materials 3, 868–871. DOI 10.1038/nmat1257. https://doi.org/10.1038/nmat1257
  6. D. D. Djayaprawira, K. Tsunekawa, M. Nagai, H. Maehara, S. Yamagata, N. Watanabe, S. Yuasa, Y. Suzuki, K. Ando (2005). 230% room-temperature magnetoresistance in CoFeB/MgO/CoFeB magnetic tunnel junctions. Applied Physics Letters 86, 092502. DOI 10.1063/1.1871344. https://doi.org/10.1063/1.1871344
  7. J. Hayakawa, S. Ikeda, F. Matsukura, H. Takahashi, H. Ohno (2005). Dependence of giant tunnel magnetoresistance of sputtered CoFeB/MgO/CoFeB magnetic tunnel junctions on MgO barrier thickness and annealing temperature. Japanese Journal of Applied Physics 44, L587. DOI 10.1143/jjap.44.l587. https://doi.org/10.1143/jjap.44.l587
  8. J. Hayakawa, S. Ikeda, Y. M. Lee, F. Matsukura, H. Ohno (2006). Effect of high annealing temperature on giant tunnel magnetoresistance ratio of CoFeB/MgO/CoFeB magnetic tunnel junctions. Applied Physics Letters 89, 232510. DOI 10.1063/1.2402904. https://doi.org/10.1063/1.2402904
  9. Y. M. Lee, J. Hayakawa, S. Ikeda, F. Matsukura, H. Ohno (2006). Giant tunnel magnetoresistance and high annealing stability in CoFeB/MgO/CoFeB magnetic tunnel junctions with synthetic pinned layer. Applied Physics Letters 89, 042506. DOI 10.1063/1.2234720. https://doi.org/10.1063/1.2234720
  10. G. D. Fuchs, J. A. Katine, S. I. Kiselev, D. Mauri, K. S. Wooley, D. C. Ralph, R. A. Buhrman (2006). Spin torque, tunnel-current spin polarization, and magnetoresistance in MgO magnetic tunnel junctions. Physical Review Letters 96, 186603. DOI 10.1103/PhysRevLett.96.186603. https://doi.org/10.1103/PhysRevLett.96.186603
  11. Y. S. Choi, Y. Nagamine, K. Tsunekawa, H. Maehara, D. D. Djayaprawira, S. Yuasa, K. Ando (2007). Effect of Ta getter on the quality of MgO tunnel barrier in the polycrystalline CoFeB/MgO/CoFeB magnetic tunnel junction. Applied Physics Letters 90, 012505. DOI 10.1063/1.2426902. https://doi.org/10.1063/1.2426902
  12. S. Yuasa, D. D. Djayaprawira (2007). Giant tunnel magnetoresistance in magnetic tunnel junctions with a crystalline MgO(001) barrier. Journal of Physics D: Applied Physics 40, R337–R354. DOI 10.1088/0022-3727/40/21/R01. https://doi.org/10.1088/0022-3727/40/21/R01
  13. R. Matsumoto, A. Fukushima, T. Nagahama, Y. Suzuki, K. Ando, S. Yuasa (2007). Oscillation of giant tunneling magnetoresistance with respect to tunneling barrier thickness in fully epitaxial Fe/MgO/Fe magnetic tunnel junctions. Applied Physics Letters 90, 252506. DOI 10.1063/1.2750398. https://doi.org/10.1063/1.2750398
  14. J. P. Velev, K. D. Belashchenko, S. S. Jaswal, E. Y. Tsymbal (2007). Effect of oxygen vacancies on spin-dependent tunneling in Fe/MgO/Fe magnetic tunnel junctions. Applied Physics Letters 90, 072502. DOI 10.1063/1.2643027. https://doi.org/10.1063/1.2643027
  15. H. Yang, S.-H. Yang, S. S. P. Parkin (2007). Crossover from Kondo-assisted suppression to co-tunneling enhancement of tunneling magnetoresistance via ferromagnetic nanodots in MgO tunnel barriers. Nano Letters 8, 340–344. DOI 10.1021/nl072930n. https://doi.org/10.1021/nl072930n
  16. M. Mizuguchi, Y. Suzuki, T. Nagahama, S. Yuasa (2007). In situ scanning tunneling microscopy observations of polycrystalline MgO(001) tunneling barriers grown on amorphous CoFeB electrode. Applied Physics Letters 91, 012507. DOI 10.1063/1.2754372. https://doi.org/10.1063/1.2754372
  17. S. Ikeda, J. Hayakawa, Y. Ashizawa, Y. M. Lee, K. Miura, H. Hasegawa, M. Tsunoda, F. Matsukura, H. Ohno (2008). Tunnel magnetoresistance of 604% at 300K by suppression of Ta diffusion in CoFeB/MgO/CoFeB pseudo-spin-valves annealed at high temperature. Applied Physics Letters 93, 082508. DOI 10.1063/1.2976435. https://doi.org/10.1063/1.2976435
  18. S. G. Wang, R. C. C. Ward, G. X. Du, X. F. Han, C. Wang, A. Kohn (2008). Temperature dependence of giant tunneling magnetoresistance in epitaxial Fe/MgO/Fe magnetic tunnel junctions. Physical Review B 78, 180411. DOI 10.1103/PhysRevB.78.180411. https://doi.org/10.1103/PhysRevB.78.180411
  19. L. Jiang, H. Naganuma, M. Oogane, Y. Ando (2009). Large tunnel magnetoresistance of 1056% at room temperature in MgO based double barrier magnetic tunnel junction. Applied Physics Express 2, 083002. DOI 10.1143/apex.2.083002. https://doi.org/10.1143/apex.2.083002
  20. S. Ikeda, K. Miura, H. Yamamoto, K. Mizunuma, H. D. Gan, M. Endo, S. Kanai, J. Hayakawa, F. Matsukura, H. Ohno (2010). A perpendicular-anisotropy CoFeB–MgO magnetic tunnel junction. Nature Materials 9, 721–724. DOI 10.1038/nmat2804. https://doi.org/10.1038/nmat2804
  21. M. Kodzuka, T. Ohkubo, K. Hono, S. Ikeda, H. D. Gan, H. Ohno (2012). Effects of boron composition on tunneling magnetoresistance ratio and microstructure of CoFeB/MgO/CoFeB pseudo-spin-valve magnetic tunnel junctions. Journal of Applied Physics 111, 043913. DOI 10.1063/1.3688039. https://doi.org/10.1063/1.3688039
  22. T. Scheike, Q. Xiang, Z. Wen, H. Sukegawa, T. Ohkubo, K. Hono, S. Mitani (2021). Exceeding 400% tunnel magnetoresistance at room temperature in epitaxial Fe/MgO/Fe(001) spin-valve-type magnetic tunnel junctions. Applied Physics Letters 118, 042411. DOI 10.1063/5.0037972. https://doi.org/10.1063/5.0037972