Pairing channel and microscopic coupling constants for the magnetic-field-induced excitonic-insulator transition of a bismuth-type semimetal

Abrikosov's 1973 theory of the metal-insulator transition of a bismuth-type semimetal in a strong magnetic field identified two candidate pairing instabilities of the quantum-limit state - type I electron-hole pairing (an excitonic insulator in which two thirds of the electron spectral weight remains gapless) and type II intragroup electron-electron pairing - but left undetermined which one, if either, actually occurs, because the renormalized interaction constants $\lambda_1$ and $\lambda_2$ were not known in sign or magnitude for any real material. Four further issues are also left open: the unproven assumption that pairing occurs at transverse momentum $\mathbf{P}=0$; the neglect of intervalley and short-range impurity matrix elements; the breakdown of the ladder-chain approximation in the fluctuation-dominated critical region near $T_c$; and the absence of the field-dependent material parameters needed for quantitative comparison with experiment. Later literature treated simplified compensated models, recast the high-field phenomenology of bismuth largely in single-particle terms (unconventional Zeeman coupling, valley emptying), and confirmed excitonic insulators in the quantum limit of graphite and other materials - but not of bismuth. The open problem is the microscopic determination of the interaction-driven fate of the quantum-limit state of bismuth-type semimetals: the signs and magnitudes of the coupling constants $\lambda_1$ and $\lambda_2$ and hence the winning pairing channel, with the fluctuation, impurity, and quantitative-comparison issues listed above entering as background limitations of the source framework.

Background

Bismuth is a semimetal whose Fermi surface consists of one hole pocket at the T point of the Brillouin zone and three symmetry-equivalent, highly anisotropic electron pockets at the L points, with a band overlap of only a few tens of meV. Because the carrier densities are so dilute, a magnetic field of a few tesla along the trigonal axis is enough to reach the quantum limit, in which all carriers of every pocket occupy only the lowest Landau level. Landau levels are the discrete, massively degenerate subbands produced by quantizing the motion transverse to the field; in the quantum limit each pocket keeps a free one-dimensional dispersion along the field, with a logarithmically diverging density of states that amplifies pairing instabilities, while the Coulomb interaction energy grows relative to the kinetic energy.

The excitonic insulator, predicted in the mid-1960s and formulated in modern terms by Jerome, Rice, and Kohn, is an insulating state of a semimetal or narrow-gap semiconductor that forms when the electron-hole Coulomb attraction binds carriers into excitons which condense coherently, much as Cooper pairs condense in a superconductor, opening a gap at the (tiny) Fermi surfaces. Fenton had analyzed this state in a magnetic field already in 1968, and Brandt and Chudinov reported in 1972 that in bismuth-antimony alloys with band overlap $|G|\lesssim 1$ meV a threshold field of 30-40 kOe produces an insulating phase that they identified as the excitonic-insulator phase, with a gap growing with field (about 7.5 K at 70 kOe), a critical temperature $T_c\simeq 0.7\Delta$, and pairing of L-point electrons with T-point holes.

Abrikosov's 1973 paper supplied the theory of this observation for a model of three electron groups and one hole group appropriate to bismuth with the field along the trigonal axis, in which the limiting Fermi momenta are unequal (of order $p_0$ for each electron pocket and $3p_0$ for the hole pocket), so the Landau-level occupations are not compensated. The central object is the two-particle vertex function $\Gamma(x,\mathbf{P})$ - the scattering amplitude - as a function of a logarithmic energy variable $x\propto|\lambda|\ln(\varepsilon_F/\omega)$ and the transverse momentum $\mathbf{P}$ measured in units of the inverse magnetic length. Building on earlier parquet analyses of the one-dimensional limit of this problem, Abrikosov argued that once the transverse Landau-level motion is retained, the parquet diagrammatic resummation - the simultaneous summation of the particle-particle and particle-hole channels that is unavoidable in a strictly one-dimensional system - reduces to ladder-type equations, because the transverse-momentum integrals are nonsingular near the pole; the vertex then takes the pole form $\Gamma\simeq 1/[x-x_0(\mathbf{P})]$, and the pole at $x_0(\mathbf{P})$ signals the pairing instability. Under the assumption that the pole occurs at transverse momentum $\mathbf{P}=0$, two candidate insulating states were derived: type I, electron-hole pairing between the electron and hole groups, which gaps one third of the electron spectral weight and leaves two thirds gapless (the residue at the gapless pole is $2/3$), so that even the 'insulator' retains metallic carriers; and type II, electron-electron pairing within a single electron group, with a BCS-like gap $\Delta_2(0)=2\varepsilon_F\exp(-\pi v_e/|\lambda_2|)$ over three identical electron subsystems. (The type I / type II terminology here is the source paper's own usage and differs from the zero-field excitonic-insulator convention in which the types label equal versus finite pairing momentum.) Impurity scattering was included through a formal mapping onto the Abrikosov-Gor'kov pair-breaking equations - the standard superconductivity formalism describing how scattering suppresses the gap and $T_c$ of a paired state, with the gap vanishing when the scattering rate reaches the pair-breaking scale. The paper computed thermodynamic properties, the dielectric constant at $T=0$, and the low-temperature conductivity, finding for type I $\sigma(0)=(2/3)Ne^2(\tau_1/2)/m_e$, and compared conductivity-ratio predictions with the Brandt-Chudinov data, finding qualitative agreement with the type I channel.

The following remain undetermined: (i) the renormalized interaction constants $\lambda_1$ (electron-hole) and $\lambda_2$ (intragroup electron-electron) are not known in sign or magnitude for any real material, so which pairing occurs is undecided; (ii) the assumption that the pole sits at $\mathbf{P}=0$ - where the unrenormalized matching value would grow as $e^{P^2/2}$ and thus be minimized - is plausible but unproven once the nonsingular integral terms of the full vertex equations are kept; (iii) intervalley and short-range impurity matrix elements were neglected under a screened-impurity assumption, which may fail in real alloys; (iv) near $T_c$ the ladder-chain treatment breaks down because singular parquet-like corrections re-emerge in a narrow critical region, so fluctuation-dominated thermodynamic singularities are not described; (v) the field-dependent effective masses, carrier densities, and impurity scattering rates needed for quantitative comparison with experiment are not derived.

The subsequent literature split into three streams. Direct continuations in 1976-1981 treated simplified compensated one-electron-band-plus-one-hole-band models: self-consistent Hartree theory of the excitonic fluctuation (which found the static, local approximation unsatisfactory), mean-field phase diagrams with impurity suppression of $T_c$, collective excitation modes, pair condensation versus liquefaction, and the ground-state energy of the electron-hole liquid. Modern parquet renormalization-group work has constructed complete ground-state 'atlases' for a three-dimensional semimetal with a single electron pocket and a single hole pocket in the quantum limit, deciding between excitonic-insulator, spin-density-wave, charge-density-wave, Wigner-crystal, and marginal-Fermi-liquid phases as functions of screening and pocket size - but for a model aimed at graphite, not at bismuth's three-electron-plus-one-hole multivalley structure. Experiments on bismuth in the extreme quantum limit have explained most observed field scales by single-particle physics (unconventional Zeeman coupling, valley emptying, Landau-level crossings), while correlation effects remain unsettled, and the excitonic insulator has been confirmed in the quantum limit of graphite and reported recently in HfTe5, as a possible spin-triplet excitonic insulator - but not in bismuth.

Problem Statement

For a bismuth-type semimetal (bismuth, and Bi-Sb alloys with small overlap) in a strong magnetic field that confines all carriers to their lowest Landau levels, determine the interaction-driven fate of the quantum-limit state, as posed and left open by Abrikosov (1973). The single scientific objective is a quantitative, microscopic identification of which pairing instability is realized in the real material class: determine the signs and magnitudes of the renormalized electron-hole coupling $\lambda_1$ and intragroup electron-electron coupling $\lambda_2$ appropriate to the Landau-quantized, anisotropic, multivalley band structure of bismuth-type semimetals, and thereby establish which of the two pairing channels derived in the source actually occurs - type I electron-hole pairing (an excitonic insulator in which two thirds of the electron spectral weight remains gapless) or type II intragroup electron-electron pairing - or whether neither derived pairing channel is realized in the material. Within the source's ladder framework for this model, the realized channel is the leading instability, namely the one whose attractive coupling has the larger magnitude - type I when $\lambda_1$ is the larger-magnitude attractive coupling, type II when $\lambda_2$ is - and neither channel is realized when neither coupling is attractive and neither channel develops an instability. The source framework additionally leaves open, as background limitations of its analysis: whether the pairing pole indeed occurs at transverse momentum $\mathbf{P}=0$, i.e. whether $|x_0(\mathbf{P})|$ is minimized at $\mathbf{P}=0$ in the full renormalized vertex equations with the nonsingular (nonladder) integral terms retained far from the pole; the role of the neglected intervalley and short-range impurity matrix elements in real Bi-Sb alloys, where the screened-impurity, intravalley-only approximation may fail; the fluctuation-dominated (parquet-type) thermodynamic singularities in the critical region near $T_c$ that the ladder-chain (mean-field-like) treatment does not capture; and the field-dependent effective masses, carrier densities, and impurity scattering rates in the high-field quantum-limit regime needed for quantitative comparison with experiment.

References

1. A. A. Abrikosov, 'Transition of a bismuth-type semimetal to an excitonic insulator in a strong magnetic field', Journal of Low Temperature Physics 10, 3-34 (1973). DOI: 10.1007/BF00655239. URL: https://doi.org/10.1007/BF00655239
2. D. Jérome, T. M. Rice, and W. Kohn, 'Excitonic Insulator', Physical Review 158, 462-475 (1967). DOI: 10.1103/PhysRev.158.462. URL: https://doi.org/10.1103/PhysRev.158.462
3. E. W. Fenton, 'Excitonic Insulator in a Magnetic Field', Physical Review 170, 816-821 (1968). DOI: 10.1103/PhysRev.170.816. URL: https://doi.org/10.1103/PhysRev.170.816
4. N. B. Brandt and S. M. Chudinov, 'Investigation of the excitonic insulator phase in bismuth-antimony alloys', Journal of Low Temperature Physics 8, 339-365 (1972). DOI: 10.1007/BF00655088. URL: https://doi.org/10.1007/BF00655088
5. Toshio Tsuzuki, 'Theory of the excitonic fluctuation in a strong magnetic field. I. Development of the physical picture and a self-consistent Hartree theory', Journal of Low Temperature Physics 22, 631-651 (1976). DOI: 10.1007/BF00659065. URL: https://doi.org/10.1007/BF00659065
6. Tatsuzo Nagai and Toshio Tsuzuki, 'Excitonic phase in a strong magnetic field. I. Phase diagram', Journal of Low Temperature Physics 25, 9-26 (1976). DOI: 10.1007/BF00654821. URL: https://doi.org/10.1007/BF00654821
7. Yoshio Kuramoto, 'Collective Excitation Modes in an Excitonic Phase under Strong Magnetic Fields', Journal of the Physical Society of Japan 42, 1143-1153 (1977). DOI: 10.1143/JPSJ.42.1143. URL: https://doi.org/10.1143/JPSJ.42.1143
8. Yu. E. Lozovik and V. I. Yudson, 'The phase fixation in the systems with e-h pairing; collective modes and current states', Solid State Communications 22, 117-122 (1977). DOI: 10.1016/0038-1098(77)90414-8. URL: https://doi.org/10.1016/0038-1098(77)90414-8
9. A. V. Klyuchnik and Yu. E. Lozovik, 'Interband transitions and currents in systems with e-h pairing', Journal of Physics C: Solid State Physics 11, L483-L487 (1978). DOI: 10.1088/0022-3719/11/12/002. URL: https://doi.org/10.1088/0022-3719/11/12/002
10. Yoshio Kuramoto and Masamichi Morimoto, 'Electron-Hole Pair Condensation and Liquefaction in Strong Magnetic Fields', Journal of the Physical Society of Japan 44, 1759-1767 (1978). DOI: 10.1143/JPSJ.44.1759. URL: https://doi.org/10.1143/JPSJ.44.1759
11. M. Morimoto, 'The ground-state energy of an electron-hole liquid in a strong magnetic field', Journal of Physics F: Metal Physics 11, 2315-2325 (1981). DOI: 10.1088/0305-4608/11/11/013. URL: https://doi.org/10.1088/0305-4608/11/11/013
12. Lu Li, J. G. Checkelsky, Y. S. Hor, C. Uher, A. F. Hebard, R. J. Cava, and N. P. Ong, 'Phase Transitions of Dirac Electrons in Bismuth', Science 321, 547-550 (2008). DOI: 10.1126/science.1158908. URL: https://doi.org/10.1126/science.1158908
13. Jason Alicea and Leon Balents, 'Bismuth in strong magnetic fields: Unconventional Zeeman coupling and correlation effects', Physical Review B 79, 241101 (2009). DOI: 10.1103/PhysRevB.79.241101. URL: https://doi.org/10.1103/PhysRevB.79.241101
14. Benoît Fauqué, Baptiste Vignolle, Cyril Proust, Jean-Paul Issi, and Kamran Behnia, 'Electronic instability in bismuth far beyond the quantum limit', New Journal of Physics 11, 113012 (2009). DOI: 10.1088/1367-2630/11/11/113012. URL: https://doi.org/10.1088/1367-2630/11/11/113012
15. Benoît Fauqué, Huan Yang, Ilya Sheikin, Luis Balicas, Jean-Paul Issi, and Kamran Behnia, 'Hall plateaus at magic angles in bismuth beyond the quantum limit', Physical Review B 79, 245124 (2009). DOI: 10.1103/PhysRevB.79.245124. URL: https://doi.org/10.1103/PhysRevB.79.245124
16. Huan Yang, Benoît Fauqué, Liam Malone, Arlei B. Antunes, Zengwei Zhu, Ctirad Uher, and Kamran Behnia, 'Phase diagram of bismuth in the extreme quantum limit', Nature Communications 1, 47 (2010). DOI: 10.1038/ncomms1039. URL: https://doi.org/10.1038/ncomms1039
17. Zengwei Zhu, Benoît Fauqué, Liam Malone, Arlei Borba Antunes, Yuki Fuseya, and Kamran Behnia, 'Landau spectrum and twin boundaries of bismuth in the extreme quantum limit', Proceedings of the National Academy of Sciences 109, 14813-14818 (2012). DOI: 10.1073/pnas.1209197109. URL: https://doi.org/10.1073/pnas.1209197109
18. R. Küchler, L. Steinke, R. Daou, M. Brando, K. Behnia, and F. Steglich, 'Thermodynamic evidence for valley-dependent density of states in bulk bismuth', Nature Materials 13, 461-465 (2014). DOI: 10.1038/nmat3909. URL: https://www.nature.com/articles/nmat3909
19. Zengwei Zhu, Jinhua Wang, Huakun Zuo, Benoît Fauqué, Ross D. McDonald, Yuki Fuseya, and Kamran Behnia, 'Emptying Dirac valleys in bismuth using high magnetic fields', Nature Communications 8, 15297 (2017). DOI: 10.1038/ncomms15297. URL: https://doi.org/10.1038/ncomms15297
20. Z. Zhu, R. D. McDonald, A. Shekhter, B. J. Ramshaw, K. A. Modic, F. F. Balakirev, and N. Harrison, 'Magnetic field tuning of an excitonic insulator between the weak and strong coupling regimes in quantum limit graphite', Scientific Reports 7, 1733 (2017). DOI: 10.1038/s41598-017-01693-5. URL: https://doi.org/10.1038/s41598-017-01693-5
21. Zhiming Pan, Xiao-Tian Zhang, and Ryuichi Shindou, 'Theory of metal-insulator transitions in graphite under high magnetic field', Physical Review B 98, 205121 (2018). DOI: 10.1103/PhysRevB.98.205121. URL: https://doi.org/10.1103/PhysRevB.98.205121
22. Zhiming Pan and Ryuichi Shindou, 'Ground-state atlas of a three-dimensional semimetal in the quantum limit', Physical Review B 100, 165124 (2019). DOI: 10.1103/PhysRevB.100.165124. URL: https://doi.org/10.1103/PhysRevB.100.165124
23. Andrey Varlamov and Peter Littlewood, 'Alexei Alekseevich Abrikosov. 25 June 1928 - 29 March 2017', Biographical Memoirs of Fellows of the Royal Society 76, 9-26 (2024). DOI: 10.1098/rsbm.2023.0030. URL: https://doi.org/10.1098/rsbm.2023.0030
24. Jinyu Liu, Varsha Subramanyan, Robert Welser, Timothy McSorley, Triet Ho, David Graf, Michael T. Pettes, Avadh Saxena, Laurel E. Winter, Shi-Zeng Lin, and Luis A. Jauregui, 'Possible Spin-Triplet Excitonic Insulator in the Ultraquantum Limit of HfTe5', Physical Review Letters 135, 046601 (2025). DOI: 10.1103/bj2n-4k2w. arXiv:2501.12572 (preprint titled 'Spin-Triplet Excitonic Insulator in the Ultra-Quantum Limit of HfTe5'). URL: https://doi.org/10.1103/bj2n-4k2w