Brar, Kroemer and English (J. Cryst. Growth 127, 752 (1993)) reported strong low-temperature photoluminescence from GaSb quantum wells of only 1-5 monolayers embedded in AlSb, and assigned it to 'quasi-direct' transitions: in (100)-oriented wells below about 2 nm, quantum confinement drives the Γ (zone-center) and L (along ⟨111⟩) conduction valleys of GaSb above the X-valley pair (minima along ⟨100⟩) oriented perpendicular to the well plane, whose heavy quantization mass leaves it lowest with an in-plane minimum at wave vector zero; the atomic-scale well width relaxes momentum conservation along the growth axis, so no-phonon recombination with Γ-valence-band holes becomes allowed. The paper left as open problems the quantitative theory of the effect — the oscillator strength (the dimensionless measure of a transition's optical strength, proportional to its radiative recombination rate) or recombination rate of the quasi-direct transition as a function of well width and valley configuration — and the quantification of island-like growth and intensity transfer between monolayer wells, as well as the demonstration of quasi-direct transitions in other systems such as InAs-AlSb or Ge(111) wells with SiGe barriers. No prior study carries out the requested quantitative and experimental program: the group's own InAs-AlSb follow-up could not uniquely identify the recombination mechanism; an independent modern line on type-I indirect-gap wells (wells whose indirect conduction-band minimum and valence-band maximum both lie in the well material) realizes no-phonon emission from indirect minima — its 2008 founding work claims the indirect-valley character of the emitting state, but neither treats GaSb-AlSb nor computes the requested width- and valley-resolved oscillator strength; and Ge-well emitters proceeded via the tensile-strain direct-gap route instead.

Background

In a direct-gap semiconductor such as bulk GaSb, the conduction-band minimum sits at the zone center (the Γ point), band-to-band optical recombination conserves crystal momentum, and emission is strong. Like most III–V compounds, GaSb also has higher-lying conduction-band minima at the L points (along ⟨111⟩) and the X points (along ⟨100⟩); recombination involving these valleys is momentum-indirect, so in bulk it requires a phonon to supply the missing momentum and is correspondingly weak. A quantum well — a thin semiconductor layer sandwiched between wider-gap barriers, grown here by molecular beam epitaxy (MBE) — quantizes carrier motion along the growth direction: each valley's subband energy rises roughly in inverse proportion to its effective mass along that direction and to the square of the well width wzw_z. The barriers (AlSb) lie above the relevant GaSb minima, so the ordering of the confined subbands is set entirely by these confinement energies. Because the Γ valley of GaSb has a very light electron mass while the X valleys are heavy and anisotropic, confinement raises the Γ subband fastest: earlier photoluminescence (PL) studies of wider GaSb/AlSb multiple quantum wells, which also established the band offsets — the energy steps between the conduction- and valence-band edges of GaSb and AlSb at their interface — used in such modeling (Cebulla et al., 1988), had shown that below wz≈4w_z\approx 4 nm the Γ subband is pushed above the four equivalent L-valley subbands, making the well indirect, like Ge.

Brar, Kroemer and English (1993) took this progression to its limit. In (100)-oriented wells below about 2 nm, quantization also lifts the L subbands above the lowest X pair — specifically the X pair whose k-space displacement vector is perpendicular to the well plane, which has the largest effective mass in the quantization direction and therefore the weakest confinement shift. That pair projects onto an in-plane band with its minimum at in-plane wave vector zero. An electron in this X subband and a heavy hole — a valence-band state whose large effective mass along the growth direction leaves it as the lowest confined hole subband — at the Γ point of the valence band then share the same in-plane momentum; only the momentum component along the growth axis differs. In a well just one to five monolayers (ML) thick (1 ML = 0.305 nm, comparable to the GaSb lattice constant of 0.61 nm), the carrier wavefunction is strongly localized along the growth direction, so conservation of that momentum component breaks down, and a no-phonon, momentum-quasi-allowed — 'quasi-direct' — radiative transition becomes possible without phonon assistance.

The 1993 work grew GaSb-AlSb multiple quantum wells with nominal GaSb thicknesses of 1–5 ML on (100) GaAs substrates (with an AlSb buffer accommodating the lattice mismatch) and observed clearly resolved 1.4 K PL peaks for every well width, including a single-monolayer peak that is comparatively weaker than the strong 2-ML peak. The measured peak energies agreed with a simple one-dimensional square-well effective-mass calculation for the perpendicular X pair, and the transition energies shifted only weakly with well width — the expected signature of the heavy X mass, and evidence against Γ–Γ or L–Γ assignments. The authors interpreted the luminescence as the first experimental realization of quasi-direct X→Γ recombination.

What the paper did not do was quantify the effect. The intensity comparison was qualitative, and the effective-mass square-well model — an instance of the envelope-function approximation, which treats the crystal electron as a slowly varying envelope multiplying the periodic Bloch function and so breaks down when the well is only a few atomic layers thick — is itself at or beyond its validity limit at monolayer widths. Three issues stand unresolved: no quantitative analysis of the radiative matrix element (the dipole-overlap integral that determines the optical emission probability) — a full theoretical calculation of the oscillator strength or recombination rate as a function of well width and valley configuration; no systematic study of well-thickness uniformity over the wafer, with the weak 1-ML peak beside the strong 2-ML peak suggesting island-like growth that transfers intensity between wells, an effect left unquantified; and no experimental demonstration of the quasi-direct concept in other material systems with suitable band structures, such as InAs-AlSb quantum wells or Ge(111) quantum wells with SiGe barriers.

Problem Statement

The problem is to convert the 1993 qualitative picture of quasi-direct recombination into a quantitative and experimentally validated account, in three linked parts that jointly serve this one objective.

(1) Quantitative theory. Calculate the radiative matrix element — equivalently the oscillator strength or the radiative recombination rate — of the quasi-direct X→Γ transition in GaSb-AlSb (100) quantum wells, as a function of well width across the monolayer range (1-5 monolayers (ML), with 1 ML = 0.305 nm) and as a function of valley configuration: which competing conduction-band valley (Γ, the four equivalent L valleys, or the perpendicular versus in-plane X pairs) forms the ground state at each width, and how the oscillator strength is distributed or transferred among them as the width changes. Two considerations shape this calculation without prescribing its method: the effective-mass envelope approximation used for the 1993 energy fit is at or beyond its validity limit at wells comparable in thickness to one lattice constant, and the relevant material parameters (GaSb/AlSb X-valley band offsets, anisotropic effective masses, interface structure) are poorly constrained, so the validity range of whatever treatment is used and the sensitivity of its predictions to these parameters are themselves part of the problem. Its predictions are to be confronted with the published 1.4 K transition energies and relative photoluminescence intensities of the 1993 monolayer wells.

(2) Disentangling growth from intrinsic recombination. Quantify the island-like growth and thickness nonuniformity of monolayer-scale wells — specifically, how much of the observed pattern of a weak 1-ML peak beside a strong 2-ML peak reflects intensity transfer between nominally different wells rather than intrinsic recombination rates. Because this separation bears directly on the intensity comparison in part (1), part (2) must rest on monolayer-scale structural characterization of the emitting wells.

(3) Extension to other material systems. Determine experimentally whether quasi-direct transitions occur in each of the two material systems proposed here: InAs-AlSb quantum wells and Ge(111) quantum wells with SiGe barriers. This part is confined to those two systems. The item remains open for both: for InAs-AlSb the quasi-direct assignment is one of four candidate mechanisms left undecided by the group's own 1993 follow-up, and for Ge(111)/SiGe it has not been attempted. A demonstration means observing no-phonon luminescence whose emitting state is identified as an indirect-valley state recombining with Γ-symmetry valence-band holes through relaxed momentum conservation along the growth axis, with the valley character of the emitting state established by evidence rather than assumed. The competing explanations make this identification the crux: for InAs-AlSb the 1993 follow-up left undecided among an X-like (quasi-direct) electron ground state, Γ-X mixing with the barrier, valence-band interface Tamm states (hole states localized at the interface itself), and breakdown of the envelope-function approximation; for Ge(111)/SiGe the indirect-valley (L along the [111] growth axis) character of the emitting state has yet to be established.

References

  1. B. Brar, H. Kroemer, J. H. English, "Quasi-direct" narrow GaSb-AlSb (100) quantum wells, Journal of Crystal Growth 127(1-4), 752-754 (1993). DOI: 10.1016/0022-0248(93)90725-c, https://doi.org/10.1016/0022-0248(93)90725-c
  2. B. Brar, H. Kroemer, J. Ibbetson, J. H. English, Photoluminescence from narrow InAs-AlSb quantum wells, Applied Physics Letters 62(25), 3303-3305 (1993). DOI: 10.1063/1.109053, https://doi.org/10.1063/1.109053
  3. B. Brar, J. Ibbetson, H. Kroemer, J. H. English, Effects of the interface bonding type on the optical and structural properties of InAs-AlSb quantum wells, Applied Physics Letters 64, 3392-3394 (1994). DOI: 10.1063/1.111285, https://doi.org/10.1063/1.111285
  4. U. Cebulla, G. Tränkle, U. Ziem, A. Forchel, G. Griffiths, H. Kroemer, S. Subbanna, Spectroscopic determination of the band discontinuity in GaSb/AlSb multiple-quantum-well structures, Physical Review B 37, 6278-6284 (1988). DOI: 10.1103/PhysRevB.37.6278, https://doi.org/10.1103/PhysRevB.37.6278
  5. A. G. Milnes, A. Y. Polyakov, Gallium antimonide device related properties, Solid-State Electronics 36, 803-818 (1993). DOI: 10.1016/0038-1101(93)90002-8, https://doi.org/10.1016/0038-1101(93)90002-8
  6. P. S. Dutta, H. L. Bhat, V. Kumar, The physics and technology of gallium antimonide: An emerging optoelectronic material, Journal of Applied Physics 81, 5821-5870 (1997). DOI: 10.1063/1.365356, https://doi.org/10.1063/1.365356
  7. T. Wang, F. Kieseling, A. Forchel, Transition from direct to indirect band structure induced by the AlSb layer inserted in the GaSb/AlSb quantum well, Physical Review B 58, 3594-3596 (1998). DOI: 10.1103/PhysRevB.58.3594, https://doi.org/10.1103/PhysRevB.58.3594
  8. F. Fuchs, J. Schmitz, H. Obloh, J. D. Ralston, P. Koidl, Photoluminescence of InAs/AlSb single quantum wells, Applied Physics Letters 64, 1665-1667 (1994). DOI: 10.1063/1.111824, https://doi.org/10.1063/1.111824
  9. T. S. Shamirzaev, A. M. Gilinsky, A. K. Kalagin, K. S. Zhuravlev, Luminescence and energy structure of ultrathin InAs/AlAs quantum wells, physica status solidi (c) 5, 2408-2411 (2008). DOI: 10.1002/pssc.200777648, https://doi.org/10.1002/pssc.200777648
  10. T. S. Shamirzaev, J. Debus, D. R. Yakovlev, M. M. Glazov, E. L. Ivchenko, M. Bayer, Dynamics of exciton recombination in strong magnetic fields in ultrathin GaAs/AlAs quantum wells with indirect band gap and type-II band alignment, Physical Review B 94, 045411 (2016). DOI: 10.1103/PhysRevB.94.045411, https://doi.org/10.1103/PhysRevB.94.045411
  11. D. S. Abramkin, T. S. Shamirzaev, Type-I indirect-gap semiconductor heterostructures on (110) substrates, Semiconductors 53, 703-710 (2019). DOI: 10.1134/S1063782619050026, https://doi.org/10.1134/S1063782619050026
  12. T. S. Shamirzaev, D. R. Yakovlev, A. K. Bakarov, N. E. Kopteva, D. Kudlacik, A. K. Gutakovskii, M. Bayer, Recombination and spin dynamics of excitons in thin (Ga,Al)(Sb,As)/AlAs quantum wells with an indirect band gap and type-I band alignment, Physical Review B 102, 165423 (2020). DOI: 10.1103/PhysRevB.102.165423, https://doi.org/10.1103/PhysRevB.102.165423
  13. Y.-H. Kuo, Y.-S. Li, Direct-gap exciton and optical absorption in the Ge/SiGe quantum well system, Applied Physics Letters 94, 121101 (2009). DOI: 10.1063/1.3106621, https://doi.org/10.1063/1.3106621
  14. S. Kikuoka, R. Kanesawa, M. Yamada, K. Hamaya, K. Sawano, Enhancement of room temperature electroluminescence from strained SiGe/Ge(111) multiple quantum wells light emitting diodes, Materials Science in Semiconductor Processing 176, 108299 (2024). DOI: 10.1016/j.mssp.2024.108299, https://doi.org/10.1016/j.mssp.2024.108299
  15. R. Kanesawa, S. Kikuoka, Y. Shibahara, Y. Wagatsuma, M. Yamada, K. Hamaya, K. Sawano, Fabrication of crack-free strained SiGe/Ge multiple quantum wells on Ge-on-Si(111) by the patterning method, Materials Science in Semiconductor Processing 177, 108300 (2024). DOI: 10.1016/j.mssp.2024.108300, https://doi.org/10.1016/j.mssp.2024.108300
  16. M. Aikawa, S. Kikuoka, M. Yamada, K. Hamaya, K. Sawano, EL and PL emissions from SiGe/Ge(111) MQWs and their dependence on the structural parameters, ECS Meeting Abstracts MA2025-02, 1754 (2025). DOI: 10.1149/MA2025-02361754mtgabs, https://doi.org/10.1149/MA2025-02361754mtgabs