The first electrically driven light-emitting diode (LED) based on a single ZnO nanowire (Bao, Zimmler, Capasso, Wang, and Ren, Nano Lett. 2006) demonstrated reliable electrical injection into an individual nanowire but left several issues explicitly unresolved: hole injection into the nanowire is strongly suppressed by a ~2.5 eV2.5\ \mathrm{eV} valence-band discontinuity at the ZnO/p-Si interface; the broad sub-bandgap electroluminescence was only tentatively attributed to defects and surface states; reverse-bias emission was only tentatively attributed to impact ionization; and the observed spatially nonuniform, spot-like emission lacked a microscopic explanation. This problem asks for the carrier-injection and recombination physics of single ZnO-nanowire LEDs to be established: efficient hole injection from a properly band-aligned substrate, a definitive microscopic attribution of the broadband and reverse-bias electroluminescence, and the identification of the origins of the emission nonuniformity. Later literature demonstrated the proposed substrate fix (single ZnO nanowires on p-GaN and p-AlGaN with ultraviolet electroluminescence) and advanced the defect physics, but the microscopic attribution of the broad sub-bandgap emission and of the spatial nonuniformity, and efficient hole injection into a single ZnO nanowire, remain open.

Background

Zinc oxide (ZnO) is a wide-band-gap semiconductor (band gap Eg≈3.35 eVE_g \approx 3.35\ \mathrm{eV} at room temperature) with a large exciton binding energy (~60 meV60\ \mathrm{meV}), which makes it attractive for ultraviolet (UV) light emitters. An exciton is the bound electron–hole pair that dominates near-band-edge (NBE) optical recombination in such materials. Because stable, reproducible p-type doping of ZnO is a long-standing difficulty, early ZnO nanowire LEDs were built as heterojunctions: an n-type (electron-conducting) ZnO nanowire is placed on a p-type (hole-injecting) substrate, with a metallic film as the top electron-injecting contact. Electroluminescence (EL) is light emitted under electrical bias, as distinct from photoluminescence (PL), which is excited optically. In these devices the EL extended from 350 nm to beyond 850 nm — far below the ZnO band-edge energy — and such broad sub-bandgap emission is commonly ascribed to deep levels: defect states and surface states whose recombination energies lie well inside the band gap. Emission well below the band gap in a UV material is a signature that the injected carriers are recombining through defects rather than producing useful near-band-edge UV light. In the 2006 single-nanowire device (Bao, Zimmler, Capasso, Wang, and Ren, Nano Lett. 2006), the nanowire was contacted by a new fabrication route: poly(methyl methacrylate) (PMMA), a polymer resist, was used as a high-resolution negative electron-beam (e-beam) resist — at high electron dose the exposed polymer chains cross-link into an insoluble film — so that direct e-beam writing could insulate and planarize the sides of an individual wire while leaving its top surface exposed for a titanium/gold (Ti/Au) metal contact, with the heavily doped p-type silicon (p-Si) wafer below acting as the hole-injecting bottom contact. The proximity effect (spreading of deposited energy by electron scattering in the resist) and the beam spot size set the smallest achievable exposed gap (50 nm50\ \mathrm{nm}) and, together with film thickness, dose, and development time, control unwanted residual cross-linked PMMA on the wire top. The device physics problem is set by the band alignment: at a heterojunction, the valence bands of the two materials are offset by a discontinuity, and a report on antimony (Sb)-doped ZnO heterojunction photodiodes (Mandalapu, Xiu, Yang, Zhao, and Liu, 2006) inferred a staggered lineup at the ZnO/Si interface with a valence-band discontinuity of approximately 2.5 eV2.5\ \mathrm{eV}, which acts as a large barrier blocking hole injection from silicon into ZnO. This is why the authors proposed substrates with band gaps larger than that of ZnO, such as p-type GaN (gallium nitride, a III–V wide-band-gap semiconductor), as an alternative. Two further observations demanded explanation: under reverse bias (substrate negative with respect to the top contact) light was also emitted, which the authors believed arose from impact ionization — carrier acceleration in the high field to energies sufficient to create electron–hole pairs that then recombine radiatively; and the forward-bias emission came from a few bright spots along the wire rather than uniformly, which the authors attributed to local nonuniformities, noting that the emission pattern reflects the waveguiding (internal optical guiding along the wire) and the position-dependent out-coupling of light. Prior ZnO nanowire LEDs had been made only from ensembles of many wires; the single-wire geometry isolates these questions but also makes them acute, since each wire's defects, contacts, and morphology differ. Later work built a substantial subfield of n-ZnO nanowire/p-GaN heterojunction LEDs (Jeong, Oh, Ham, Lee, and Myoung, 2007; Lupan, Pauporté, and Viana, 2010), demonstrated single ZnO nanowire/p-GaN and single n-ZnO/p-AlGaN (p-type aluminium gallium nitride, a wider-gap alloy) nanowire devices with UV electroluminescence (Bie et al., 2010; Tang, Li, and Zhou, 2013), identified bound- and free-exciton recombination as the origin of the near-band-edge EL in single nanowire LEDs at low temperature (Zimmler, Voss, Ronning, and Capasso, 2009), treated visible EL as trap-assisted recombination (Kim, Jeon, Cho, and Kim, 2016), measured position-dependent native point defects in ZnO nanowires by depth-resolved cathodoluminescence — cathodoluminescence being light emission excited by a focused electron beam, so that resolving it in depth localises the emitting defects — (Cox, Foster, Jarjour, von Wenckstern, Grundmann, and Brillson, 2018), and controlled interfaces to suppress the defect band (You et al., 2018). These results define what is already known and what remains unexplained in the original question.

Problem Statement

For a light-emitting diode built from a single n-type ZnO nanowire on a hole-injecting substrate (the original demonstration used heavily doped p-type silicon, with a metallic top contact and cross-linked PMMA insulating the wire sides), establish the carrier-injection and recombination physics that the 2006 demonstration left open, as a coherent set of linked questions toward one objective — a physically understood, efficiently hole-injected single-ZnO-nanowire LED whose emission mechanism is identified rather than 'believed': (1) Hole injection: the valence-band discontinuity of approximately 2.5 eV2.5\ \mathrm{eV} at the ZnO/Si interface strongly suppresses hole injection from p-Si. Single-wire p-GaN and p-AlGaN devices are demonstrated as proof-of-concept leads (Bie et al., 2010; Tang, Li, and Zhou, 2013), but the substrate choice is not thereby settled and efficient hole injection remains open; determine what substrate choice and band alignment (substrates with band gaps larger than that of ZnO) actually yields efficient hole injection into an individual ZnO nanowire, and quantify the resulting injection-limited diode behavior. (2) Emission mechanism: identify the microscopic origin of the broad sub-bandgap electroluminescence (extending roughly from 350 nm to beyond 850 nm, with only weak ~380 nm near-band-edge emission) — specifically, which defect and surface states, located where in the wire or at its interfaces, recombine with the injected carriers — rather than merely attributing it. (3) Reverse-bias emission: light is also observed under reverse bias; determine whether impact ionization is the actual mechanism, or rule it out in favor of alternatives. (4) Emission uniformity: the forward-bias light originates from a few bright spots along the wire rather than uniformly; determine the microscopic origin of this spatial nonuniformity (local structural nonuniformities, contact inhomogeneity, waveguiding and out-coupling, or localized defect concentrations). The four preceding questions are independently answerable: settling any one of them resolves only that part while the others remain open. The associated fabrication limit — scaling the direct-write contact geometry below the ~50 nm50\ \mathrm{nm} separation set by the e-beam spot size and proximity effect, while eliminating proximity-induced residual cross-linked PMMA on the nanowire top — is acknowledged scope context rather than a separate deliverable, since it bounds which wires can be contacted and how uniformly.

References

  1. Bao, Zimmler, Capasso, Wang, Ren (2006). Broadband ZnO Single-Nanowire Light-Emitting Diode. Nano Lett. 6, 1719–1722. DOI 10.1021/nl061080t. https://doi.org/10.1021/nl061080t
  2. Mandalapu, Xiu, Yang, Zhao, Liu (2006). p-type behavior from Sb-doped ZnO heterojunction photodiodes. Appl. Phys. Lett. 88, 112108. DOI 10.1063/1.2186516. https://doi.org/10.1063/1.2186516
  3. Jeong, Oh, Ham, Lee, Myoung (2007). ZnO-Nanowire-Inserted GaN/ZnO Heterojunction Light-Emitting Diodes. Small 3, 568–572. DOI 10.1002/smll.200600479. https://doi.org/10.1002/smll.200600479
  4. Zimmler, Voss, Ronning, Capasso (2009). Exciton-related electroluminescence from ZnO nanowire light-emitting diodes. Appl. Phys. Lett. 94, 241120. DOI 10.1063/1.3157274. https://doi.org/10.1063/1.3157274
  5. Bie, Liao, Wang, Zhou, Han, Ye, Zhao, Wu, Dai, Xu, Sang, Deng, Laurent, Leprince-Wang, Yu (2010). Single ZnO Nanowire/p-type GaN Heterojunctions for Photovoltaic Devices and UV Light-Emitting Diodes. Adv. Mater. 22, 4284–4287. DOI 10.1002/adma.201000985. https://doi.org/10.1002/adma.201000985
  6. Lupan, Pauporté, Viana (2010). Low-Voltage UV-Electroluminescence from ZnO-Nanowire Array/p-GaN Light-Emitting Diodes. Adv. Mater. 22, 3298–3302. DOI 10.1002/adma.201000611. https://doi.org/10.1002/adma.201000611
  7. Tang, Li, Zhou (2013). Ultraviolet Electroluminescence of Light-Emitting Diodes Based on Single n-ZnO/p-AlGaN Heterojunction Nanowires. Nano Lett. 13, 5046–5050. DOI 10.1021/nl401941g. https://doi.org/10.1021/nl401941g
  8. Mo, Fang, Long, Li, Wang, Chen, Huang, Zeng, Zhang, Pan (2014). Unusual electroluminescence from n-ZnO@i-MgO core-shell nanowire color-tunable light-emitting diode at reverse bias. Phys. Chem. Chem. Phys. 16, 9302–9308. DOI 10.1039/c3cp55505d. https://doi.org/10.1039/c3cp55505d
  9. Schuster, Laumer, Zamani, Magén, Morante, Arbiol, Stutzmann (2014). p-GaN/n-ZnO Heterojunction Nanowires: Optoelectronic Properties and the Role of Interface Polarity. ACS Nano 8, 4376–4384. DOI 10.1021/nn406134e. https://doi.org/10.1021/nn406134e
  10. Shi, Zhang, Cui, Zhuang, Wu, Jiang, Chu, Dong, Zhang, Du (2015). Epitaxial growth of vertically aligned ZnO nanowires for bidirectional direct-current driven light-emitting diodes applications. CrystEngComm 17, 40–49. DOI 10.1039/c4ce01788a. https://doi.org/10.1039/c4ce01788a
  11. Kim, Jeon, Cho, Kim (2016). Enhancement of Trap-Assisted Green Electroluminescence Efficiency in ZnO/SiO2/Si Nanowire Light-Emitting Diodes on Bendable Substrates by Piezophototronic Effect. ACS Appl. Mater. Interfaces 8, 2764–2773. DOI 10.1021/acsami.5b11053. https://doi.org/10.1021/acsami.5b11053
  12. You, Xu, Qin, Zhu, Manohari, Xu, Zhao, Liu (2018). Interface control for pure ultraviolet electroluminescence from nano-ZnO-based heterojunction devices. Science Bulletin 63, 38–45. DOI 10.1016/j.scib.2017.12.006. https://doi.org/10.1016/j.scib.2017.12.006
  13. Cox, Foster, Jarjour, von Wenckstern, Grundmann, Brillson (2018). Defect Manipulation To Control ZnO Micro-/Nanowire-Metal Contacts. Nano Lett. 18, 6974–6980. DOI 10.1021/acs.nanolett.8b02892. https://doi.org/10.1021/acs.nanolett.8b02892