Results and discussion

The I-V measurements revealed that the Schottky properties of the Pt/70nm-PEDOT:PSS/Zn-ZnO structure as shown in Fig. 2.3(a). The Schottky barrier height (SBH; ΦB) and ideality factor were determined from I-V measurement by the thermionic emission theory in excess of several kT/q; I = I0exp[q(V – IRs)/nkT – 1] and I0 = AA**T²exp(–qΦB/kT),¹⁷ where I0 is the saturation current, Rs is the series resistance, k is Boltzmann’s constant, T is the absolute temperature, n is the ideality factor, A is the contact area, and q is the electronic charge. A** is the effective Richardson constant.

The theoretical value of A** = 32 A cm^–2 K^–2 was used in this study. The SBH is 0.71

25

eV (ideality factor: n = 1.28) for the PEDOT:PSS contact.

Figure 2.3(b) shows the photoelectron yield spectra of the PEDOT:PSS films on glass substrates. The ionization energy of the polymer film is defined as the energy difference between the vacuum level and the extrapolated edge of the highest occupied molecular orbital (HOMO).¹⁸ Therefore, the work function can be estimated by linearly extrapolating the edge of the square-root photoelectron yield to the baseline. The work

Figure 2.3. (a) I-V characteristics of PEDOT:PSS on a Zn-ZnO substrate. (b) The photoelectron yield spectra of the PEDOT:PSS films on glass substrates.

26

function of the as-coated PEDOT:PSS film was 5.16 eV, which is comparable to previously reported values (4.9 ~ 5.2 eV).^9,19 These electrical properties were consistent with the reported electrical properties.

Figure 2.4 shows the valence band spectra of the PEDOT:PSS/ZnO structure and ZnO substrate. The spectrum of PEDOT:PSS/ZnO structure consists of superimposed spectra of PEDOT:PSS and ZnO. The spectrum of the ZnO was dominated by the O 2p valence band, which had an onset energy, as defined by linear extrapolation of the valence band maximum (VBM), at a binding energy of 3.3 ± 0.1 eV below the Fermi level. For the PEDOT:PSS/ZnO structure, the state in the binding energy region from VBM of ZnO to the Fermi level was filled with electron, meaning that the PEDOT:PSS acted as an electrode with the work function of 5.16 eV.

Figure 2.4. Valence band photoemission spectra for a Zn-polar ZnO substrate and a PEDOT:PSS/ZnO structure at a TOA of 88°.

27

As regards the electron depletion of semiconductor layer, a clear difference was observed between the PEDOT:PSS and a metal electrode. Figure 2.5(a) shows Zn 2p3/2

spectra of the Zn-polar ZnO substrate, PEDOT:PSS/ZnO interface, and the Pt/ZnO interface. Since the core levels have a fixed binding energy difference from the conduction- and valence band edges, they can be used to trace shifts of the band edges with respect to the Fermi level. Taking account of that HX-PES proved at 20 nm below the surface, the spectrum of Zn-polar ZnO substrate indicates the bulk property of ZnO.

Therefore, with the Pt/ZnO interface, the shift of Zn 2p3/2 to the lower binding energy means that the Fermi level moved to middle of the band gap of ZnO at the interface by Pt electrode formation.

Figure 2.5. (a) Zn 2p3/2 core spectra of a Zn-polar ZnO substrate, a PEDOT:PSS/ZnO interface, and a Pt/ZnO interface at a TOA of 88°. (b) TOA dependence of the binding energy of the Zn 2p3/2 core spectra.

The circles and squares show the PEDOT:PSS/ZnO interface and the Pt/ZnO interface, respectively. The Zn 2p3/2 core spectrum of a Zn-polar ZnO substrate at a TOA of 88° is also indicated as a dashed line.

28

Figure 2.5(b) shows TOA dependence of the binding energy of the Zn 2p3/2

core spectra. The circles and squares show the PEDOT:PSS/ZnO interface and the Pt/ZnO interface, respectively. The Zn 2p3/2 core spectrum of a Zn-polar ZnO substrate at a TOA of 88° is also indicated as a dashed line. With the Pt/ZnO interface, the TOA dependence of the Zn 2p3/2 position revealed upward band bending of ZnO, suggesting the formation of depletion region at the ZnO surface. These results are consistent with a conventional Schottky contact formation of metal/semiconductor interface. On the other hand, with the PEDOT:PSS/ZnO interface, the position of Zn 2p3/2 core level sited at almost the same binding energy of the ZnO substrate.

At the TOA = 30° (interface sensitive angle), asymmetric property of Zn 2p3/2

core level was pronounced at the lower binding energy as shown in Fig. 2.6(a). The inset shows the asymmetry factor peak of the spectrum at TOA = 30° compared to the spectrum at TOA = 88°. The asymmetry factor peak is denoted temporarily as Zn-O-X in Fig. 2.6(a). With the O 1s core level, the asymmetry factor is also observed at the lower binding energy as shown in Fig. 2.6(b). The asymmetry factor in the O 1s core level is small due to the strong signal from O1s of PEDOT:PSS, which can be speculated to be the band bending of the ZnO at the interface. However, there are two more possibilities for the asymmetry of Zn 2p3/2 core level. The one is formation of an interface layer based on Zn atoms. In the Zn 2p3/2 core level, the peak position of metal Zn is at a little lower binding energy (0.1 ~ 0.2 eV)^20,21 than that of the ZnO bonding state. The position of Zn-O-X is much lower than that of the Zn metal state. If the asymmetry is caused by the formation of an interface layer, Zn should be strongly ionized at the interface and affect other core levels. However, the other core levels indicated that there was no change according to the strong Zn ionization.

29

The other possibility is the chemical state of the Zn embedded in the PEDOT:PSS layer. This is because the ZnO was etched by the PEDOT:PSS solution due to its acid property, which was especially observed at the O-polar face.¹⁰ To verify the Zn state in the PEDOT:PSS layer, conventional XPS measurements and HX-PES were performed on a 500-nm-thick PEDOT:PSS/O-polar ZnO structure. The

Figure 2.6. (a) Zn 2p3/2 and (b) O 1s core spectra of the Pt/PEDOT:PSS/ZnO structure at TOAs of 88° and 30°. The solid lines show spectra and the open circles show sum-fitted curves. The dashed lines are fitted curves for each bond. The inset in (a) shows the asymmetry factor peak of the spectrum at TOAs of 30°

compared to that at TOAs of 88°.

30

500-nm-thick PEDOT:PSS is considered to be enough thick to avoid detecting Zn 2p photoelectrons from the ZnO substrate.

Figure 2.7 shows valence band region and the Zn 2p3/2 core level spectra for the 500-nm-thick PEDOT:PSS/O-polar ZnO substrate and 30-nm-PEDOT:PSS/Zn-ZnO substrate. The ZnO band structure was not observed in the valence band region for the 500-nm-thick PEDOT:PSS/O-polar ZnO substrate, meaning that no signal from ZnO

Figure 2.7. (a) Valence band photoemission and (b) Zn 2p3/2 core level spectra for a 30-nm-PEDOT:PSS/Zn-polar ZnO substrate and a 500-nm-PEDOT:PSS/O-polar ZnO structure.

31

substrate can be detected. In contrast, the Zn 2p3/2 core level was observed at around 1023 eV (Fig. 2.7(b)), which is the higher binding energy side of the semiconducting Zn-O bonding state. These results revealed that the etched Zn embedded in the PEDOT:PSS layer did not form ZnO or affect the valence band structure of the PEDOT:PSS layer. In the case of PEDOT:PSS/Zn-polar ZnO interface, the etched Zn embedded in the PEDOT:PSS layer was not confirmed. Thus, the asymmetric property originated in the ZnO band bending.

Figure 2.8(a) and 2.8(b) shows the fitting results and the simulated potential energy of the Zn 2p spectra, respectively. The ZnO surface was bent upward by approximately 0.8 eV, and the length was 4 nm.

Based on these results, the band offsets of PEDOT:PSS/ZnO interfaces are illustrated in Fig. 2.9. The band structure of PEDOT:PSS/ZnO heterojunction is obviously different from that of conventional Pt/ZnO Schottky contact. The

Figure 2.8. (a) The inset shows fitted curves at TOAs of 88, 50, and 30° (b) Simulated energy potential of the ZnO interface. A depth of 0 nm indicates the interface.

32

Figure 2.9. Schematic energy level diagrams for a PEDOT:PSS/ZnO structure. EF: Fermi level of ZnO (Fermi level of measurement system), Eg: energy band gap, ΦCIB: charge injection barrier of the PEDOT:PSS contact to the Zn-polar ZnO substrate.

PEDOT:PSS formed a charge injection barrier of approximately 0.8 eV as shown Fig.

2.9. The depletion length is very short (~4 nm). The estimated upward bending is slightly greater than the charge injection barrier based on I-V measurements. For the PEDOT:PSS side, no TOA dependence was observed in the S 1s core spectra. There is a possibility of the existence of a dipole layer.^9,22

33