Bardeen's 1947 theory attributed the observed near-independence of rectification barriers on the metal work function to a high density of electronic surface states in the semiconductor's forbidden gap, whose charge forms a compensating double layer that pins the surface Fermi level. The paper left open the energy distribution of these states (discrete or continuous), the quantitative conditions for the free-surface double layer in specific materials such as Si, Ge and Se, the effect of surface films, and — most lastingly — the barrier height at atomically intimate contacts where the metal broadens the surface levels. Later literature resolved the surface-science questions: direct photoelectric and photoemission measurements on clean cleaved Si and Ge confirmed the picture and measured surface-state bands, and Heine's theory of metal-induced gap states (the continuum of electronic states that metal wave functions create inside the semiconductor's forbidden gap at an intimate contact) reframed the intimate-contact question. The surviving open core is the modern Schottky-barrier-height problem: no generally accepted, parameter-free theory quantitatively predicts barrier heights and Fermi-level pinning for real metal–semiconductor interfaces, as authoritative reviews and first-principles work through 2026 confirm.

Background

Rectification at a metal–semiconductor contact — the strongly asymmetric, diode-like current–voltage behavior first systematically studied in the 1870s — was among the earliest known semiconductor phenomena and underpinned practical diodes before the p–n junction (historical account in the Mönch reprint volume). Its classical theory was formulated by Mott (1938) and in parallel by Schottky (1938). When a metal touches an n-type semiconductor, electrons flow until the Fermi levels (the electrochemical potentials of the electrons) of the two sides align, leaving a positively charged depletion layer — a Schottky exhaustion layer — extending into the semiconductor. The equilibrium barrier height ϕ0\phi_0 for electron flow from the metal into the semiconductor should then equal the difference between the metal work function (the energy needed to take an electron from the metal's Fermi level to vacuum) and the semiconductor electron affinity (the energy from the conduction-band edge to vacuum). This is the Schottky–Mott rule: barrier heights should track the metal work function, metal by metal. Experiments of the 1940s on rectifying contacts to silicon, germanium, cuprous oxide and selenium (analyzed in Bardeen's paper) contradicted this rule: barrier heights deduced from rectification characteristics were nearly independent of the metal used.

Bardeen (1947) supplied the now-classic explanation. Theory in the 1930s had shown that the mere termination of the crystal potential at a surface can introduce localized electronic states with energies in the forbidden gap between the valence and conduction bands — Bardeen's paper calls them Tamm levels, today simply surface states. He showed that if the density of such states is large enough, the net charge of the (partially filled) surface states, together with an equal and opposite space charge extending into the semiconductor, forms a double layer at the free surface that pins the position of the semiconductor Fermi level relative to the band edges at the surface, making the work function nearly independent of bulk doping (which sets the interior Fermi level). At a metal contact, the contact potential difference is then accommodated by charge in the surface states rather than by the semiconductor space charge, so the barrier height and rectification become nearly independent of the metal. The quantitative treatment assumed a uniform continuous energy distribution of surface states across the forbidden gap, a uniform exhaustion layer, and an interfacial gap of atomic dimensions.

The paper itself closed with unresolved issues: the true energy distribution of surface states (discrete versus continuous) was unknown; quantitative material-specific predictions for the free-surface double layer were unavailable; surface films and adsorbed impurities were not analyzed; for contacts so intimate that the metal significantly broadens the surface levels the barrier height was left as an open theoretical problem; only the uniform exhaustion layer was treated; and direct experimental evidence for the surface states was lacking. Each thread was taken up by later work. Photoelectric measurements on atomically clean cleaved silicon (Allen and Gobeli 1962) and germanium (Gobeli and Allen 1964) supplied the direct experimental evidence and confirmed the pinning picture; angle-resolved photoemission spectroscopy (ARPES, which maps the energy and momentum distribution of electrons emitted from a surface) then measured the surface-state bands of clean ordered surfaces directly, starting with Si(111)2×1 (Hansson, Uhrberg and Nicholls 1983), showing that on such surfaces the states form two-dimensional bands rather than discrete levels. Heine (1965) attacked the intimate-contact question: strictly, localized surface states cannot persist once a metal is deposited, because the metal's wave functions tail into the semiconductor gap as a continuum of metal-induced gap states (MIGS); this insight, together with Cowley and Sze's (1965) formalization of Bardeen's model with an atomically thin interfacial layer, launched the modern theory of the Schottky barrier height (SBH).

The following decades produced a sequence of competing and partially complementary models: the covalent-to-ionic pinning transition of Kurtin, McGill and Mead (1969); the first realistic first-principles (density-functional theory, DFT) interface calculations of Louie, Chelikowsky and Cohen (1977); Brillson's chemical-reactivity transition (1978); Spicer et al.'s unified defect model (1979); Freeouf and Woodall's effective work function model (1981); and Tersoff's charge-neutrality-level account of the gap-state continuum (1984).

The central phenomenon is Fermi-level pinning (FLP): the barrier height varies far more weakly with metal work function than the Schottky–Mott rule predicts, conventionally quantified by the pinning factor S=dϕB/dΦMS=d\phi_B/d\Phi_M (the change of barrier height per change of metal work function; S=1S=1 is the unpinned Schottky–Mott limit, S=0S=0 complete pinning). Tung's 2014 review concludes that existing models are too simple to treat the chemistry exhibited at real interfaces and that no simple equations applicable to all metal–semiconductor interfaces have emerged; Tung and Kronik (2021) achieved a quantitative account of computed barrier heights for epitaxial (crystallographically registered) interfaces between metals and zinc-blende semiconductors (semiconductors with the GaAs-type cubic crystal structure); and first-principles and experimental work through 2026 (Xiang, Luo and Li 2024, 2025; Kim et al. 2017; Dovale-Farelo and Choudhary 2026; Qin and Wong 2026) continues to contest the pinning mechanism and to find general prediction out of reach.

Problem Statement

The problem asks for a quantitative, predictive account of how surface and interface electronic states — electronic states localized at the semiconductor surface or at the metal–semiconductor interface, with energies inside the semiconductor's forbidden gap between the valence and conduction bands — determine the barrier height, the equilibrium energy step that an electron must surmount to pass from the metal into the semiconductor. Among the issues Bardeen's paper left open were: (1) the energy distribution of semiconductor surface states — whether discrete levels or a continuum, and with what density — for specific materials (for example Si, Ge, Se), including how that distribution is modified by surface films and adsorbed impurities; and (2) the quantitative conditions under which a free semiconductor surface carries a compensating double layer (surface-state charge plus an equal and opposite space charge extending into the bulk), including their dependence on the bulk charged-impurity density and on surface preparation, beyond the idealized uniform Schottky exhaustion layer (the fully depleted, uniformly charged region assumed in the classical theory) and for general space-charge distributions such as those involving trapped charge. These two questions were settled by the surface-science literature of 1950–1990 — direct photoelectric and photoemission measurements, ARPES surface-state band mapping, and standard band-bending electrostatics — and they serve as established context and inputs rather than open tasks. The live objective is (3): compute the equilibrium barrier height at a metal–semiconductor contact, including the atomically intimate regime in which the metal significantly broadens and merges the surface levels, so that the barrier height is quantitatively predictable from the properties of the metal, the semiconductor, and the interface structure alone — the surviving modern Schottky-barrier-height / Fermi-level-pinning prediction problem. The target of (3) is a parameter-free account, using no fitted constants, that reproduces — to within 0.1 eV for barrier heights and within 0.1 for pinning factors — the values reported by the cited experimental and first-principles studies for a class of real metal–semiconductor interfaces that spans at least two chemically distinct semiconductor families and at least two contacting metals, includes at least one strongly pinned interface (measured pinning factor S ≤ 0.5), and reaches beyond the interface types already given such an account (such as the epitaxial metal–zinc-blende contacts of Tung and Kronik 2021). The three components are scientifically linked: the settled surface-state spectra of (1) and double-layer conditions of (2) fix the free-surface reference that any contact theory of (3) must reproduce and extend into the intimate-contact limit, while the barrier heights of (3) are the observable against which (1) and (2) were ultimately tested.

References

  1. John Bardeen, Surface States and Rectification at a Metal Semi-Conductor Contact, Physical Review 71, 717–727 (1947), DOI: 10.1103/PhysRev.71.717, https://doi.org/10.1103/PhysRev.71.717
  2. John Bardeen, Surface States and Rectification at a Metal Semi-Conductor Contact, in Winfried Mönch (ed.), Electronic Structure of Metal-Semiconductor Contacts, Perspectives in Condensed Matter Physics vol. 4, pp. 63–73, Springer, Dordrecht (1999; original 1947), DOI: 10.1007/978-94-009-0657-0_7, https://doi.org/10.1007/978-94-009-0657-0_7
  3. Winfried Mönch (ed.), Electronic Structure of Metal-Semiconductor Contacts (reprint collection of the classic papers from Braun 1874 to the 1980s), Perspectives in Condensed Matter Physics vol. 4, Springer, Dordrecht (1990; eBook 1999), DOI: 10.1007/978-94-009-0657-0, https://doi.org/10.1007/978-94-009-0657-0
  4. W. Schottky, Halbleitertheorie der Sperrschicht, Die Naturwissenschaften 26, 843 (1938), DOI: 10.1007/bf01774216, https://doi.org/10.1007/bf01774216
  5. N. F. Mott, Note on the contact between a metal and an insulator or semi-conductor, Mathematical Proceedings of the Cambridge Philosophical Society 34, 568–572 (1938), DOI: 10.1017/s0305004100020570, https://doi.org/10.1017/s0305004100020570
  6. F. G. Allen and G. W. Gobeli, Work Function, Photoelectric Threshold, and Surface States of Atomically Clean Silicon, Physical Review 127, 150–158 (1962), DOI: 10.1103/PhysRev.127.150, https://doi.org/10.1103/PhysRev.127.150
  7. G. W. Gobeli and F. G. Allen, Photoelectric properties and work function of cleaved germanium surfaces, Surface Science 2, 402–408 (1964), DOI: 10.1016/0039-6028(64)90081-0, https://doi.org/10.1016/0039-6028(64)90081-0
  8. G. V. Hansson, R. I. G. Uhrberg and J. M. Nicholls, Photoemission studies of surface states on Si(111)2×1, Surface Science 132, 31–39 (1983), DOI: 10.1016/0039-6028(83)90526-5, https://doi.org/10.1016/0039-6028(83)90526-5
  9. A. M. Cowley and S. M. Sze, Surface States and Barrier Height of Metal-Semiconductor Systems, Journal of Applied Physics 36, 3212–3220 (1965), DOI: 10.1063/1.1702952, https://doi.org/10.1063/1.1702952
  10. Volker Heine, Theory of Surface States, Physical Review 138, A1689–A1696 (1965), DOI: 10.1103/PhysRev.138.A1689, https://doi.org/10.1103/PhysRev.138.A1689
  11. S. Kurtin, T. C. McGill and C. A. Mead, Fundamental Transition in the Electronic Nature of Solids, Physical Review Letters 22, 1433–1436 (1969), DOI: 10.1103/PhysRevLett.22.1433, https://doi.org/10.1103/PhysRevLett.22.1433
  12. Steven G. Louie, James R. Chelikowsky and Marvin L. Cohen, Ionicity and the theory of Schottky barriers, Physical Review B 15, 2154–2162 (1977), DOI: 10.1103/PhysRevB.15.2154, https://doi.org/10.1103/PhysRevB.15.2154
  13. L. J. Brillson, Transition in Schottky Barrier Formation with Chemical Reactivity, Physical Review Letters 40, 260–263 (1978), DOI: 10.1103/PhysRevLett.40.260, https://doi.org/10.1103/PhysRevLett.40.260
  14. W. E. Spicer, P. W. Chye, P. R. Skeath, C. Y. Su and I. Lindau, New and unified model for Schottky barrier and III–V insulator interface states formation, Journal of Vacuum Science and Technology 16, 1422–1433 (1979), DOI: 10.1116/1.570215, https://doi.org/10.1116/1.570215
  15. J. L. Freeouf and J. M. Woodall, Schottky barriers: An effective work function model, Applied Physics Letters 39, 727–729 (1981), DOI: 10.1063/1.92863, https://doi.org/10.1063/1.92863
  16. J. Tersoff, Schottky Barrier Heights and the Continuum of Gap States, Physical Review Letters 52, 465–468 (1984), DOI: 10.1103/PhysRevLett.52.465, https://doi.org/10.1103/PhysRevLett.52.465
  17. Raymond T. Tung, The physics and chemistry of the Schottky barrier height, Applied Physics Reviews 1, 011304 (2014), DOI: 10.1063/1.4858400, https://doi.org/10.1063/1.4858400
  18. Raymond T. Tung and Leeor Kronik, Quantitative explanation of the Schottky barrier height, Physical Review B 103, 035304 (2021), DOI: 10.1103/PhysRevB.103.035304, https://doi.org/10.1103/PhysRevB.103.035304
  19. Changsik Kim, Inyong Moon, Daeyeong Lee, Min Sup Choi, Faisal Ahmed, Seunggeol Nam, Yeonchoo Cho, Hyeon-Jin Shin, Seongjun Park and Won Jong Yoo, Fermi Level Pinning at Electrical Metal Contacts of Monolayer Molybdenum Dichalcogenides, ACS Nano 11, 1588–1596 (2017), DOI: 10.1021/acsnano.6b07159, https://doi.org/10.1021/acsnano.6b07159
  20. Ziying Xiang, Jun-Wei Luo and Shu-Shen Li, Self-passivation reduces the Fermi level pinning in the metal-semiconductor contacts, arXiv:2411.14220 (2024), https://arxiv.org/abs/2411.14220
  21. Ziying Xiang, Jun-Wei Luo and Shu-Shen Li, Unified interface dipole theory for Fermi level pinning effect at metal-semiconductor contacts, arXiv:2511.21494 (2025), https://arxiv.org/abs/2511.21494
  22. Jimmy Qin and H.-S. Philip Wong, Contacts to Low-Dimensional Semiconductors: Physical Theory and Analytical Model, arXiv:2606.16030 (2026), https://arxiv.org/abs/2606.16030
  23. Viviana Dovale-Farelo and Kamal Choudhary, Effect of Exchange-Correlation Functionals on Schottky Barriers at Si/Metal Interfaces, arXiv:2603.06725 (2026), https://arxiv.org/abs/2603.06725