Letter | Published:

Spin–orbit-driven band inversion in bilayer graphene by the van der Waals proximity effect

Abstract

Spin–orbit coupling (SOC) is the key to realizing time-reversal-invariant topological phases of matter1,2. SOC was predicted by Kane and Mele3 to stabilize a quantum spin Hall insulator; however, the weak intrinsic SOC in monolayer graphene4,5,6,7 has precluded experimental observation in this material. Here we exploit a layer-selective proximity effect—achieved via a van der Waals contact with a semiconducting transition-metal dichalcogenide8,9,10,11,12,13,14,15,16,17,18,19,20,21—to engineer Kane–Mele SOC in ultra clean bilayer graphene. Using high-resolution capacitance measurements to probe the bulk electronic compressibility, we find that SOC leads to the formation of a distinct, incompressible, gapped phase at charge neutrality. The experimental data agree quantitatively with a simple theoretical model in which the new phase results from SOC-driven band inversion. In contrast to Kane–Mele SOC in monolayer graphene, the inverted phase is not expected to be a time-reversal-invariant topological insulator, despite being separated from conventional band insulators by electric-field-tuned phase transitions where crystal symmetry mandates that the bulk gap must close22. Our electrical transport measurements reveal that the inverted phase has a conductivity of approximately e2/h (where e is the electron charge and h Planck’s constant), which is suppressed by exceptionally small in-plane magnetic fields. The high conductivity and anomalous magnetoresistance are consistent with theoretical models that predict helical edge states within the inverted phase that are protected from backscattering by an emergent spin symmetry that remains robust even for large Rashba SOC. Our results pave the way for proximity engineering of strong topological insulators as well as correlated quantum phases in the strong spin–orbit regime in graphene heterostructures.

Access optionsAccess options

Rent or Buy article

Get time limited or full article access on ReadCube.

from$8.99

All prices are NET prices.

Data availability

The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.

Additional information

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  1. 1.

    Hasan, M. Z. & Kane, C. L. Colloquium: topological insulators. Rev. Mod. Phys. 82, 3045–3067 (2010).

  2. 2.

    Qi, X.-L. & Zhang, S.-C. Topological insulators and superconductors. Rev. Mod. Phys. 83, 1057–1110 (2011).

  3. 3.

    Kane, C. L. & Mele, E. J. Quantum spin Hall effect in graphene. Phys. Rev. Lett. 95, 226801 (2005).

  4. 4.

    Min, H. et al. Intrinsic and Rashba spin-orbit interactions in graphene sheets. Phys. Rev. B 74, 165310 (2006).

  5. 5.

    Huertas-Hernando, D., Guinea, F. & Brataas, A. Spin-orbit coupling in curved graphene, fullerenes, nanotubes, and nanotube caps. Phys. Rev. B 74, 155426 (2006).

  6. 6.

    Yao, Y., Ye, F., Qi, X.-L., Zhang, S.-C. & Fang, Z. Spin-orbit gap of graphene: first-principles calculations. Phys. Rev. B 75, 041401(R) (2007).

  7. 7.

    Sichau, J. et al. Resonance microwave measurements of an intrinsic spin-orbit coupling gap in graphene: a possible indication of a topological state. Phys. Rev. Lett. 122, 046403 (2019).

  8. 8.

    Gmitra, M. & Fabian, J. Graphene on transition-metal dichalcogenides: a platform for proximity spin-orbit physics and optospintronics. Phys. Rev. B 92, 155403 (2015).

  9. 9.

    Gmitra, M. & Fabian, J. Proximity effects in bilayer graphene on monolayer WSe2: field-effect spin valley locking, spin-orbit valve, and spin transistor. Phys. Rev. Lett. 119, 146401 (2017).

  10. 10.

    Wang, Z. et al. Strong interface-induced spin–orbit interaction in graphene on WS2. Nat. Commun. 6, 8339 (2015).

  11. 11.

    Wang, Z. et al. Origin and magnitude of ‘designer’ spin-orbit interaction in graphene on semiconducting transition metal dichalcogenides. Phys. Rev. X 6, 041020 (2016).

  12. 12.

    Yang, B. et al. Tunable spin-orbit coupling and symmetry-protected edge states in graphene/WS2. 2D Mater. 3, 031012 (2016).

  13. 13.

    Yang, B. et al. Strong electron-hole symmetric Rashba spin-orbit coupling in graphene/monolayer transition metal dichalcogenide heterostructures. Phys. Rev. B 96, 041409 (2017).

  14. 14.

    Völkl, T. et al. Magnetotransport in heterostructures of transition metal dichalcogenides and graphene. Phys. Rev. B 96, 125405 (2017).

  15. 15.

    Wakamura, T. et al. Strong anisotropic spin-orbit interaction induced in graphene by monolayer WS2. Phys. Rev. Lett. 120, 106802 (2018).

  16. 16.

    Zihlmann, S. et al. Large spin relaxation anisotropy and valley-Zeeman spin-orbit coupling in WSe2/graphene/h-BN heterostructures. Phys. Rev. B 97, 075434 (2018).

  17. 17.

    Avsar, A. et al. Spin-orbit proximity effect in graphene. Nat. Commun. 5, 4875 (2014).

  18. 18.

    Dankert, A. & Dash, S. P. Electrical gate control of spin current in van der Waals heterostructures at room temperature. Nat. Commun. 8, 16093 (2017).

  19. 19.

    Ghiasi, T. S. et al. Large proximity-induced spin lifetime anisotropy in transition-metal dichalcogenide/graphene heterostructures. Nano Lett. 17, 7528–7532 (2017).

  20. 20.

    Omar, S. & van Wees, B. J. Spin transport in high-mobility graphene on WS2 substrate with electric-field tunable proximity spin-orbit interaction. Phys. Rev. B 97, 045414 (2018).

  21. 21.

    Benítez, L. A. et al. Strongly anisotropic spin relaxation in graphene–transition metal dichalcogenide heterostructures at room temperature. Nat. Phys. 14, 303–308 (2018).

  22. 22.

    Zaletel, M. P. & Khoo, J. K. The gate-tunable strong and fragile topology of multilayer-graphene on a transition metal dichalcogenide. Preprint at https://arXiv.org/abs/1901.01294 (2019).

  23. 23.

    Gmitra, M., Kochan, D., Hogl, P. & Fabian, J. Trivial and inverted Dirac bands and the emergence of quantum spin Hall states in graphene on transition-metal dichalcogenides. Phys. Rev. B 93, 155104 (2016).

  24. 24.

    Dean, C. R. et al. Boron nitride substrates for high-quality graphene electronics. Nat. Nanotechnol. 5, 722–726 (2010).

  25. 25.

    Zibrov, A. A. et al. Tunable interacting composite fermion phases in a half-filled bilayer-graphene Landau level. Nature 549, 360–364 (2017).

  26. 26.

    McCann, E. & Fal’ko, V. I. Landau-level degeneracy and quantum Hall Effect in a graphite bilayer. Phys. Rev. Lett. 96, 086805 (2006).

  27. 27.

    Khoo, J. Y. & Levitov, L. Tunable quantum Hall edge conduction in bilayer graphene through spin-orbit interaction. Phys. Rev. B 98, 115307 (2018).

  28. 28.

    Hunt, B. M. et al. Direct measurement of discrete valley and orbital quantum numbers in bilayer graphene. Nat. Commun. 8, 948 (2017).

  29. 29.

    Garcia, J. H., Vila, M., Cummings, A. W. & Roche, S. Spin transport in graphene/transition metal dichalcogenide heterostructures. Chem. Soc. Rev. 47, 3359–3379 (2018).

  30. 30.

    Wang, L. et al. One-dimensional electrical contact to a two-dimensional material. Science 342, 614–617 (2013).

  31. 31.

    Gustafsson, M. V. et al. Ambipolar Landau levels and strong band-selective carrier interactions in monolayer WSe2. Nat. Mater. 17, 411–415 (2018).

  32. 32.

    Ashoori, R. C. et al. Single-electron capacitance spectroscopy of discrete quantum levels. Phys. Rev. Lett. 68, 3088–3091 (1992).

  33. 33.

    Spanton, E. M. et al. Observation of fractional Chern insulators in a van der Waals heterostructure. Science 360, 62–66 (2018).

Download references

Acknowledgements

Experimental work at UCSB was supported by the ARO under award MURI W911NF-16-1-0361. D.R. and J.C.H. acknowledge support by the US Department of Energy, DE-SC0016703, for synthesis of WSe2 crystals. K.W. and T.T. acknowledge support from the Elemental Strategy Initiative conducted by the MEXT, Japan, and the CREST (JPMJCR15F3), JST. M.P.Z. was supported by the Director, Office of Science, Office of Basic Energy Sciences, Materials Sciences and Engineering Division of the US Department of Energy under contract no. DE-AC02-05-CH11231 (van der Waals heterostructures programme, KCWF16). A.F.Y. acknowledges the support of the David and Lucile Packard Foundation and the Alfred. P. Sloan Foundation. J.O.I. acknowledges the support of the Netherlands Organization for Scientific Research (NWO) through the Rubicon grant, project number 680-50-1525/2474. C.L. and L.S.L. acknowledge support of the STC Center for Integrated Quantum Materials under NSF grant no. DMR-1231319. J.Y.K. acknowledges support by the National Science Scholarship from the Agency for Science, Technology and Research (A*STAR). A portion of this work was performed at the National High Magnetic Field Laboratory, which is supported by National Science Foundation Cooperative Agreement no. DMR-1644779 and the State of Florida. Measurements made use of a dilution refrigerator funded through the Major Research Instrumentation Program of the US National Science Foundation under award no. DMR-1531389, and the MRL Shared Experimental Facilities, which are supported by the MRSEC Program of the US National Science Foundation under award no. DMR-1720256.

Reviewer information

Nature thanks Saroj Prasad Dash and the other anonymous reviewer(s) for their contribution to the peer review of this work.

Author information

J.O.I., X.C. and H.Z. fabricated the devices. J.O.I., X.C., E.M.S. and A.F.Y. performed the measurements. J.Y.K., C.L. and M.P.Z. performed the theoretical simulations, with J.Y.K. and C.L. advised by L.S.L. J.O.I., X.C., J.Y.K., C.L., L.S.L., M.P.Z. and A.F.Y. analysed the data. D.R. and J.C.H. grew the WSe2 crystals used in devices A1, S1 and A2/S2. T.T. and K.W. grew the hBN crystals used in all devices. J.O.I. and A.F.Y. wrote the paper in consultation with X.C., J.Y.K., C.L., L.S.L. and M.P.Z.

Competing interests

The authors declare no competing interests.

Correspondence to A. F. Young.

Extended data figures and tables

Extended Data Fig. 1 Optical images and corresponding models for all the devices studied.

For each device, the optical image is shown at the top and the corresponding model shown under. a, Control device C1. b, Asymmetric device A1. c, Symmetric device S1. d, Another symmetric device with a single-sided region. Details about this device are presented in Extended Data Fig. 9. e, Another asymmetric device, A3. This device showed additional features in the magnetocapacitance measurements that are associated with a moiré superlattice potential due to alignment of the BLG with the top hBN, see Extended Data Fig. 7.

Extended Data Fig. 2 Electrical schematic showing the details of the penetration field capacitance measurements.

The components enclosed in the red dashed box are inside the cryostat, held at base temperature. Voltages are applied to Vtop and Vsamp (at a fixed Vgate) in order to adjust charge density n = ctvt + cbvb and displacement field D = (ctvt − cbvb)/(2ϵ0). See Methods for details.

Extended Data Fig. 3 Comparison between experimental and simulated penetration field capacitance for the three device configurations studied.

a, Penetration field capacitance, CP, as a function of charge density n and displacement field D measured at B = 0 and T ≈ 50 mK in the control device, C1. b, Schematic of device C1, a BLG flake encapsulated with hBN. c, Simulated CP as a function of interlayer bias, u, and charge density, n, from a low-energy continuum model for device C1. d, Low-energy bands for C1 near the K point of the Brillouin zone with ky = 0. Line colour represents the expectation value of the out-of-plane projection of the electron spin, 〈Sz〉. Panels correspond to u = −10 meV (top), u = 0 meV (middle) and u = 10 meV (bottom). e, CP for device A1. Arrows indicate weak features in CP. f, Schematic of device A1, in which the BLG is asymmetrically encapsulated between WSe2 and hBN crystals. g, Simulated CP for the asymmetric geometry with λI = 1.7 meV Ising SOC on the bottom layer. Arrows denote band-edge singularity-associated features arising from spin-split valence (conduction) bands for electron (hole) doping, visible in h, the low-energy band structure. h, Low-energy bands for A1 near the K point of the Brillouin zone with ky = 0. Line colour represents the expectation value of the out-of-plane projection of the electron spin, 〈Sz〉. Panels correspond to u = −10 meV (top), u = 0 meV (middle), and u = 10 meV (bottom). i, CP measured for device S1. Note the incompressible phase centred at D = 0, n = 0, absent in either control or symmetric devices. j, Schematic of device S1, in which the BLG is symmetrically encapsulated between two few-layer WSe2 crystals. k, Simulated CP for the symmetric geometry, with an Ising SOC of equal magnitude (λI = 2.6 meV) but opposite signs on opposite layers. l, Low-energy bands for S1 in the symmetric geometry near the K point of the Brillouin zone with ky = 0. Line colour represents the expectation value of the out-of-plane projection of the electron spin, 〈Sz〉. Panels correspond to u = −10 meV (top), u = 0 meV (middle), and u = 10 meV (bottom).

Extended Data Fig. 4 Comparison between experimental data and numerical simulations including λI and λR.

a, Measured CP of device A1 as a function of n and D. b–d, Simulated CP from a low-energy continuum model with SOC as follows: b, a one-sided Ising SOC of λI = 1.7 meV; c, a one-sided Rashba SOC of λR = 15 meV; and d, a one-sided Ising SOC of λI = 1.7 meV and a Rashba SOC of λR = 15 meV. e, Linecut taken at the location of the dashed white line (D = 0.1 V nm−1) in a. The symbols mark dips in CP indicated with the same symbols in a. f, Linecut of the simulated data in b taken at u = −11.8 meV, equivalent to a displacement field of 0.1 V nm−1 for device A1. The symbols mark dips in CP indicated with the same symbols in b. g, Linecut of the calculated density of states (DOS) in h for device A1 taken at u = −11.8 meV, equivalent to a displacement field of 0.1 V nm−1 for device A1. The symbols mark peaks in DOS which correspond to dips in CP indicated with the same symbols in a, b, e, f and g. h, Calculated DOS for device A1. i, Low-energy bands (specifically bands 3–6) near the K point of the Brillouin zone with ky = 0, u = −10 meV and λI = 1.7 meV. A clear band splitting is observed in the conduction band associated with the addition of an Ising SOC. jl, Fermi contours at E = −10 meV and u = −10 meV (j), E = 5 meV and u = −10 meV (k), and E = 6 meV and u = −10 meV (l). m, Low-energy bands near the K point of the Brillouin zone with ky = 0, u = −10 meV and λR = 15 meV. n–p, Fermi contours at E = −10 meV and u = −10 meV (n), E = 5 meV and u = −10 meV (o) and E = 6 meV and u = −10 meV (p).

Extended Data Fig. 5 Comparison between the ν = ±3 phase transitions in the control device C1 and the asymmetric device A1.

a, Energy level diagram of the zero-energy LL in the absence of SOC. The ν = ±3 transitions are occurring between ground states with identical spin polarization. Note that offsets from u* = 0 are possible due to differing on-site energies within the BLG unit cell, which can arise from coupling to the hBN substrate, but that these offsets do not influence the spin degree of freedom. The solid and dashed lines differentiate spin orientation. b, Measured \({D}_{\nu =\pm 3}^{\ast }\) as a function of the total magnetic field (BT) for fixed B = 4 T in control device C1. No Zeeman dependence is observed, consistent with expectations from an SOC-free model. The red dashed line is the average value of \({D}_{\nu =\pm 3}^{\ast }\). c, Energy level diagram of the zero-energy LL with a layer-selective Ising SOC of λI = 5 meV, with sign chosen so that the effect of the SOC opposes the external field (reproduced from Fig. 2f of the main text). Note that the ν = ±3 transitions now occur between ground states with opposite spin polarization. d, Measured \({D}_{\nu =\pm 3}^{\ast }\) as a function of BT for fixed B = 4 T in device A1, reproduced from the main text. The red dashed line is a two-parameter fit with λI = 1.7 meV and ϵBLG = 2.8, with the latter needed for the conversion between experimentally measured D and theoretically calculated u. e, Schematic of the effect of BT in an asymmetric device. The red curve plots the dot product of the spin orientation on the top layer and the magnetic field, and the blue curve plots the product of the spin orientation on the bottom layer and the magnetic field. Whereas the LL in the unaffected layer always aligns its spin polarization with the external magnetic field (see the red arrows in the dashed boxes for total external magnetic fields of 5, 10 and 20 T, respectively), the spin polarization of LLs in the SOC-proximitized bottom layer result from a competition between SOC-induced Zeeman field (out of plane) and the changing direction of the physical Zeeman field (see the blue arrows in the dashed boxes). The affected spin cants only slightly for \({E}_{{\rm{Z}}}\ll {\lambda }_{{\rm{I}}}\), but eventually the Zeeman energy overwhelms the SOC and the two spins align as EZ/λI → ∞.

Extended Data Fig. 6 Measured penetration field capacitance for devices A1, C1 and S1 at 18 T.

a–c, CP measured as a function of D and n at B = 18 T for devices C1 (a), A1 (b) and S1 (c). The red dashed line in c shows the location of the linecut plotted in Extended Data Fig. 7a.

Extended Data Fig. 7 Fractional quantum Hall and Chern insulator states at high magnetic field.

a, Fractional quantum Hall states (black labels) observed at 18 T in device S1. CP/Cref taken at D = 1.5 V nm−1 in Extended Data Fig. 6c (red dashed line), corresponding to a range of −4 < ν < −2. In the N = 0 orbital, fractional quantum Hall states up to sevenths are clearly observed. In the N = 1 orbital, an incompressible state is observed at half-filling. b, Fractional Chern insulator states in asymmetric device A3 at high magnetic fields with the BLG and hBN perfectly aligned. CP (normalized by Cref) is shown as a function of nominal electron density n0/c (where c is the geometric capacitance) and applied perpendicular magnetic field B, at a fixed polarizing electric field p0 \(\left(\frac{{p}_{0}}{c}=\frac{2{\epsilon }_{0}}{c}D=-6{\rm{V}}\right)\). c, Schematic of the observed insulating states in units normalized to the moiré unit cell area (Amoiré): these are the number of flux quanta per moiré unit cell nΦ (= B/AmoiréΦ0) and the number of electrons per moiré unit cell ne = n/Amoiré, where Φ0 = h/e is a flux quantum and n is the electron density. The insulating states are characterized by their inverse slope and intercept in these units, t and s, respectively. We observe a topological Chern band with δt = C = 1 and δs = 1, which originates at nΦ = 1 between insulating states (t, s) = (1, 1) and (2, 0) (black lines). We observe fractional Chern insulating states at 1/3, 2/5, 3/5, 2/3 filling of the band (blue lines) with quantum numbers t, s = (4/3, 2/3), (7/5, 3/5), (8/5, 2/5), (5/3, 1/3), labelled respectively.

Extended Data Fig. 8 Phase transitions as a function of magnetic field for the rest of the zeroth LL and the N = 2, 3 excited states.

Following the analysis for \({D}_{\nu =\pm 3}^{\ast }\) in Fig. 2h, we plot the dips in CP at corresponding polarizations (p0/c D) as a function of perpendicular magnetic field (B) in device A1 for ν = 0 (a), ν = ±1 (b), ν = ±2 (c), ν = ±5 (d), ν = ±6 (e), ν = ±7 (f), ν = ±9 (g), ν = ±10 (h) and ν = ±11 (i). The labels (−, +) indicate the different phase transitions for each integer gap.

Extended Data Fig. 9 Summary of data from device A2/S2.

a, Optical image of device A2/S2. b, Rxx as a function of D and n at B = 0 T for the S2 portion of the device. The inverted phase is evident at charge neutrality and zero displacement field. c, Rxx as a function of B and D for the S2 portion. d, CP as a function of n and D at B = 4 T for device A2/S2. Two sets of ν = ±3 transitions are evident, indicated by the white arrows. e, ν = ±3 transitions for device A2/S2. The crossing between ν = −3 and ν = +3 coming from the one-sided portion of the device (A2) is consistent with the crossing found in the asymmetric device A1. No crossing is evident in the symmetric portion which is consistent with transitions in S1.

Supplementary information

Supplementary Information

This file contains Supplementary Text and Data, Supplementary Fig. 1 and Supplementary References.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark
Fig. 1: Inverted phase in bilayer graphene from proximity-induced SOC.
Fig. 2: Layer-selective spin–orbit proximity effect.
Fig. 3: Magnetoconductance and edge state transport in the inverted phase.
Extended Data Fig. 1: Optical images and corresponding models for all the devices studied.
Extended Data Fig. 2: Electrical schematic showing the details of the penetration field capacitance measurements.
Extended Data Fig. 3: Comparison between experimental and simulated penetration field capacitance for the three device configurations studied.
Extended Data Fig. 4: Comparison between experimental data and numerical simulations including λI and λR.
Extended Data Fig. 5: Comparison between the ν = ±3 phase transitions in the control device C1 and the asymmetric device A1.
Extended Data Fig. 6: Measured penetration field capacitance for devices A1, C1 and S1 at 18 T.
Extended Data Fig. 7: Fractional quantum Hall and Chern insulator states at high magnetic field.
Extended Data Fig. 8: Phase transitions as a function of magnetic field for the rest of the zeroth LL and the N = 2, 3 excited states.
Extended Data Fig. 9: Summary of data from device A2/S2.

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.