Polarization dependence of ZnO Schottky barriers revealed by photoelectron spectroscopy

Philipp Wendel, Shanmugapriya Periyannan, Wolfram Jaegermann, and Andreas Klein

Phys. Rev. Materials 4, 084604 – Published 31 August 2020

Abstract

In order to answer the question of whether Schottky barriers on polar ZnO surfaces are different at Zn- and O-terminated surfaces, the interface formation of $n$ -type ZnO and different high work function metals and metal oxides (Pt, ${PtO}_{x}$ , and ${RuO}_{2}$ ) with Schottky barrier heights of up to 1.5 eV has been studied using photoelectron spectroscopy with in situ sample preparation. The experiments are designed to exclude the effects of substrate reduction and consequent Fermi level pinning by high concentrations of oxygen vacancies. Moreover, by including the Zn LMM Auger emission in the analysis, it is demonstrated that an accurate extraction of barrier heights needs to take into account that the screening of the photoelectron core hole can change in the course of contact formation. The polarization dependence of Schottky barriers, which is important for piezotronic applications, is in most cases dominated by the influence of defects. Reducing the influence of defects, up to $\sim 240$ meV higher Schottky barriers are revealed on oxygen-terminated surfaces. This is opposite to what has been reported in the literature but agrees with the dependence of barrier heights expected for an incomplete screening of the polarization of ZnO by the electrode as for ferroelectric materials.

Received 27 June 2020
Accepted 11 August 2020

DOI:https://doi.org/10.1103/PhysRevMaterials.4.084604

Physics Subject Headings (PhySH)

Defects Electric polarization Electrical conductivity

Diodes Piezoelectrics Solid-solid interfaces Wide band gap systems

Photoemission spectroscopy

Condensed Matter, Materials & Applied Physics

Authors & Affiliations

Philipp Wendel , Shanmugapriya Periyannan, Wolfram Jaegermann, and Andreas Klein ^*

Technische Universität Darmstadt, Institute of Materials Science, 64287 Darmstadt, Germany

^*andreas.klein@tu-darmstadt.de

Article Text (Subscription Required)

Click to Expand

Supplemental Material (Subscription Required)

Click to Expand

References (Subscription Required)

Click to Expand

Issue

Vol. 4, Iss. 8 — August 2020

Reuse & Permissions

Access Options

Author publication services for translation and copyediting assistance advertisement

Images

Figure 1
XP spectra recorded during deposition of ${PtO}_{x}$ onto ZnO. The spectra at the bottom show the bare substrate, the following ones after deposition of 1, 2, 4, and 8 nm of ${PtO}_{x}$ . Binding energy shifts of the $Zn {2 p}_{3 / 2}$ core level and of the Zn LMM Auger line are indicated.
Reuse & Permissions
Figure 2
Zn LMM Auger emission lines of the bare substrates (dashed lines) and after deposition of 1 nm of Pt (a) or of 1 nm ${PtO}_{x}$ (b) (solid lines). The spectra have been normalized and shifted for better comparison and to calculate the difference (dots). The emissions at $495 eV$ and $492 eV$ in (a) indicate the formation of metallic Zn.
Reuse & Permissions
Figure 3
Development of binding energies of the $Zn {2 p}_{3 / 2}$ and the Zn LMM emissions during deposition of Pt (a) and ${PtO}_{x}$ (b). The evolution of the Fermi energy at the interface as calculated using Eq. (3), which takes final state effects into account, is also shown.
Reuse & Permissions
Figure 4
XP valence band spectra of (i) $c$ -axis oriented single crystal surfaces without high-temperature treatment after cleaning by heating at 400 $^{\circ} C$ , (ii) after high-temperature treatment without cleaning, (iii) after high-temperature treatment and subsequent deposition of a ZnO buffer layer at 400 $^{\circ} C$ , and (iv) after high-temperature treatment and subsequent deposition of a ZnO buffer layer at room temperature. The spectra were recorded with $10^{\circ}$ take-off angle. The different shapes and intensities of the valence bands are characteristic for polar Zn- and O-terminated ZnO surfaces [48, 49]. The difference vanishes for room temperature deposition of the buffer layer, indicating that no single polarity of the surface remains.
Reuse & Permissions
Figure 5
Evolution of the Fermi energy in the ZnO band gap for ${RuO}_{2}$ deposition onto (a) ZnO single crystals, (b) ZnO single crystals with room temperature ZnO buffer layers, (c) ZnO single crystals with ZnO buffer layers deposited at 400 $^{\circ} C$ . The Fermi level shifts are calculated according to Eq. (3). The Zn LMM Auger emissions in (a) were only recorded up to 3 nm.
Reuse & Permissions
Figure 6
Representation of changes in the electrostatic potential of a polar oxide with imperfect screening (top) and perfect screening (bottom) of the polarization charge as proposed by Stengel et al. [61]. An imperfect screening is equivalent to a spatial separation of the bound surface charge of the polar material and the screening charge in the metal. Imperfect screening results in a potential drop across the dead layer. XPS will probe the potential below the dead layer, as this extends only about an atomic bond length, which is much thinner than the information depth of XPS.
Reuse & Permissions

Physical Review Materials