Population-inversion-induced band-structure instability in long-period one-dimensional superlattices: survival beyond the idealized static model and experimental observability

Kroemer (Phys. Rev. B 15, 880 (1977)) predicted that a long-period (order 100 Å) quasi-one-dimensional superlattice with a nearly filled lowest miniband, driven into strong population inversion across the first minigap, exhibits a band-structure instability: carriers transferred across the minigap reverse the sign of their screening contribution (antiscreening), drive the free-electron dielectric function at the superlattice wave number through zero, render the self-consistent superlattice potential multivalued, and at a critical inversion level collapse the first minigap abruptly — a first-order transition of the electron-hole gas with predicted optical hysteresis, saturable far-infrared absorption, and tunable emission, requiring cryogenic temperatures and Q-switched-laser-level inversion. The prediction rests on an idealized static one-dimensional model — an imposed double-step inverted distribution, a sinusoidal potential, twofold nearly-free-electron perturbation theory, neglected transverse motion, and electrostatic-only branch energetics — and has never been tested. The later literature leaves the prediction unresolved: a 1981 paper by Ferry extended the same mechanism to two-dimensional lateral superlattices, the 1980s–90s lateral-superlattice program observed miniband transport effects but never the inversion instability. The open problem is to determine whether the instability survives realistic conditions — finite temperature, kinetically determined occupations, multiband and non-sinusoidal potentials, transverse dispersion — to map the parameter region and thresholds where it occurs, to settle branch selection with proper total energies, to analyze the transition dynamics, and to establish whether any real quasi-one-dimensional or lateral superlattice can be inverted far enough, for long enough, for the predicted signatures to be observed.

Background

An artificial superlattice is a crystal in which a periodic perturbation with a period $\Lambda$ much longer than the atomic lattice constant is imposed on a semiconductor, for example by alternating layers of two materials or by periodic doping. Because the Brillouin zone is folded by the long period, the carrier spectrum splits into narrow minibands separated by minigaps. Esaki and Tsu (1970) proposed that miniband transport in such structures yields negative differential conductivity (NDC), a decrease of current with increasing electric field, which motivated much early superlattice work. Kroemer's 1977 paper starts from a different observation. In a quasi-one-dimensional conductor — a material whose electrons move effectively along parallel chains — the free-electron contribution to the static dielectric function has a sharp maximum, the Kohn anomaly, when the perturbation wave number equals twice the Fermi wave number $2k_F$. This is the same response that underlies the Peierls instability, the tendency of a one-dimensional metal to spontaneously distort into a lattice of period $\pi/k_F$ that opens a gap exactly at the Fermi level. Consequently, in a long-period (order 100 Å) quasi-one-dimensional superlattice produced, for instance, by periodic doping, the minigap is strongly screened, and the screening depends on how the electrons are distributed in energy.

Kroemer asked what happens under strong population inversion, a nonequilibrium state in which carriers are promoted across the minigap so that, in the language of quasi-Fermi levels (separate effective Fermi levels for electrons and holes out of equilibrium), the states just above the gap are more filled and those just below it more empty than at equilibrium. Using the one-dimensional effective-mass approximation (the periodic crystal potential absorbed into an effective mass $m^*$, and motion transverse to the chains ignored), a purely sinusoidal intrinsic potential $V_i=V_{i0}\cos gx$ with $g=2\pi/\Lambda$, and the nearly-free-electron twofold-degenerate perturbation result for states near the zone boundary $k=\pm g/2$, he wrote the total potential $V(x)=V_0\cos gx$ as the sum of intrinsic and electronic parts and imposed electrostatic self-consistency. The result is a closed relation between the intrinsic and total amplitudes, $V_{e0}=V_0-V_{i0}=-(\lambda/\lambda_0)V_0 S(V_0)$, where $S$ is an occupation-dependent susceptibility, the scaling parameter $\lambda=(\Lambda^3 m^*/m_0)/(A\epsilon_c/\epsilon_0)$ collects the period, effective mass, chain cross-section $A$, and crystal permittivity $\epsilon_c$, and $\lambda_0=16.4$ Å. States below the minigap contribute screening charge; states above it contribute antiscreening charge, a space-charge modulation with the opposite phase. For an inverted double-step distribution in which the states bracketing the minigap are more than half transferred, $S(V_0)$ turns negative and diverges logarithmically as $V_0\to 0$: the free-electron dielectric function at the superlattice wave number passes through zero, and the relation $V_0(V_{i0})$ becomes triple-valued. An electrostatic potential-energy argument selects the stable branch; beyond a critical inversion the selected branch jumps and the first minigap collapses abruptly — in Kroemer's words, essentially a first-order phase transition of the electron-hole gas. For an illustrative case ($\Lambda=100$ Å, $m^*=0.1m_0$, kinetic scale $\hbar^2g^2/2m^*\approx 0.15$ eV), laser-illumination-level carrier transfer was estimated to shift the gap by about 0.01 eV, with predicted observable consequences: saturable far-infrared absorption compounded by the Burstein shift (the blue shift of the absorption edge caused by band filling), optical hysteresis and cross-band triggering under Q-switched laser pulses (intense pulses obtained by quality-factor switching of the laser cavity), shielding of thick-film interiors by absorbing and reflecting surfaces, and possibly tunable emission.

The analysis, however, rests on idealizations the author explicitly flagged: the double-step inverted distribution is imposed rather than derived from carrier kinetics; the twofold-degenerate perturbation formula is applied heuristically to all states, with an ad hoc renormalization of the intrinsic amplitude; the potential is sinusoidal; transverse motion is ignored; the treatment is static and spatially uniform; branch selection uses only electrostatic energy; and no systematic parameter map or experimental threshold analysis is given. The only direct theoretical follow-up, by Ferry (1981), recast the same mechanism for lateral superlattices — periodic potentials imposed in the plane of a two-dimensional electron gas, for instance by an array of surface gates on a semiconductor inversion layer — but within the same static screening framework and again with an imposed nonequilibrium occupation, so the same idealizations remained unresolved.

Problem Statement

Determine whether Kroemer's population-inversion-induced band-structure instability of a long-period quasi-one-dimensional superlattice is a real physical effect and whether it is experimentally observable, by resolving the open issues arising from the idealizations that the 1977 analysis itself states:

(1) Robustness of the mechanism. Replace the idealized model — imposed double-step inversion, sinusoidal intrinsic potential, twofold-degenerate nearly-free-electron perturbation treatment applied to all states, purely one-dimensional motion — with physically consistent treatments: a nonequilibrium occupation derived from generation, recombination, and relaxation kinetics under intense (e.g., Q-switched-laser) pumping, finite temperature, multiband coupling, non-sinusoidal potentials, and the effect of residual transverse dispersion in quasi-one-dimensional materials, or the two-dimensional lateral-superlattice geometry of Ferry (1981). Establish whether the antiscreening-driven sign reversal of the dielectric function at the superlattice wave number, the multivalued self-consistent potential, and the abrupt collapse of the first minigap survive, and at what thresholds.

(2) Parameter-space delineation. Map the material and geometric parameters — period $\Lambda$, effective mass $m^*$, permittivity $\epsilon_c$, chain cross-section $A$, carrier density, and inversion level and duration — inside which the instability occurs, or show that no physically realizable parameter region supports it, including thresholds for the predicted optical effects.

(3) Branch selection and stability. Decide which branch of the multivalued self-consistent relation is realized, using total energies that include kinetic and exchange-correlation contributions — the energy terms beyond the classical electrostatic one that arise from quantum statistics (exchange) and interaction corrections beyond the mean field (correlation) — in addition to the electrostatic potential energy, and characterize the static stability of each branch.

(4) Dynamics and spatial structure. Analyze the dynamics of the predicted first-order transition: switching, nucleation of domains of different branches, spatial inhomogeneity and surface shielding in thick films, and temporal recovery after the pump, at least to the level needed to state on what time scales the inverted state and the instability could manifest.

(5) Experimental feasibility. Establish whether appropriate quasi-one-dimensional or lateral superlattice materials can be inverted to the required degree on the required time scales at the necessary low temperatures, and identify the observable signatures — saturable far-infrared absorption, optical hysteresis, abrupt minigap collapse, tunable emission — together with the conditions under which they could be detected.

These components jointly serve one overarching objective: deciding whether the 1977 prediction describes a real, observable nonequilibrium transition of an inverted superlattice electron gas.

References

1. H. Kroemer, 'Band-structure instability in long-period one-dimensional superlattices under strong population inversion', Physical Review B 15, 880–884 (1977). DOI: 10.1103/PhysRevB.15.880, https://doi.org/10.1103/PhysRevB.15.880
2. L. Esaki and R. Tsu, 'Superlattice and negative differential conductivity in semiconductors', IBM Journal of Research and Development 14, 61–65 (1970). DOI: 10.1147/rd.141.0061, https://doi.org/10.1147/rd.141.0061
3. D. K. Ferry, 'Charge Instabilities in Lateral Super-Lattices under Conditions of Population Inversion', physica status solidi (b) 106, 63–71 (1981). DOI: 10.1002/pssb.2221060107, https://doi.org/10.1002/pssb.2221060107
4. R. K. Reich, R. O. Grondin, and D. K. Ferry, 'Transport in lateral surface superlattices', Physical Review B 27, 3483–3493 (1983). DOI: 10.1103/PhysRevB.27.3483, https://doi.org/10.1103/PhysRevB.27.3483
5. R. O. Grondin and J. C. Wang, 'Superlattices as millimeter-wave devices', Superlattices and Microstructures 2, 197–200 (1986). DOI: 10.1016/0749-6036(86)90019-4, https://doi.org/10.1016/0749-6036(86)90019-4
6. G. Bernstein and D. K. Ferry, 'Observation of negative differential conductivity in a FET with structured gate', Zeitschrift für Physik B – Condensed Matter 67, 449–452 (1987). DOI: 10.1007/BF01304112, https://doi.org/10.1007/BF01304112
7. K. Ismail, W. Chu, A. Yen, D. A. Antoniadis, and H. I. Smith, 'Negative transconductance and negative differential resistance in a grid-gate modulation-doped field-effect transistor', Applied Physics Letters 54, 460–462 (1989). DOI: 10.1063/1.100952, https://doi.org/10.1063/1.100952
8. W. Hansen, J. P. Kotthaus, and U. Merkt, 'Electrons in Laterally Periodic Nanostructures', Semiconductors and Semimetals, Chapter 5, pp. 279–380 (1992). DOI: 10.1016/S0080-8784(08)62395-9, https://doi.org/10.1016/S0080-8784(08)62395-9
9. D. K. Ferry, 'Quantum magnetotransport in lateral surface superlattices', Progress in Quantum Electronics 16, 251–317 (1992). DOI: 10.1016/0079-6727(92)90005-G, https://doi.org/10.1016/0079-6727(92)90005-G
10. W. Chen and H. Ahmed, 'Nanofabrication for Electronics', Advances in Imaging and Electron Physics, pp. 87–185 (1997). DOI: 10.1016/S1076-5670(08)70122-8, https://doi.org/10.1016/S1076-5670(08)70122-8
11. D. Golež, L. Boehnke, H. U. R. Strand, M. Eckstein, and P. Werner, 'Nonequilibrium GW+EDMFT: Antiscreening and Inverted Populations from Nonlocal Correlations', Physical Review Letters 118, 246402 (2017). DOI: 10.1103/PhysRevLett.118.246402, https://doi.org/10.1103/PhysRevLett.118.246402 (preprint arXiv:1702.04952, https://arxiv.org/abs/1702.04952)
12. B. A. Glavin, V. A. Kochelap, T. L. Linnik, and K. W. Kim, 'Emission of monochromatic terahertz phonons in layered semiconductor structures: physics and anticipated applications', Proceedings of the 2nd International Conference on Advanced Optoelectronics and Lasers (CAOL 2005), vol. 2, p. 27 (2005). DOI: 10.1109/CAOL.2005.1553907, https://doi.org/10.1109/CAOL.2005.1553907