Poly(3,4-ethylenedioxythiophene)−polystyrene sulfonate (PEDOT: PSS) is a promising polymeric material for electrodes in flexible, stretchable electronic devices owing to its maintained conductivity under mechanical bending [1–3]. However, its relatively low conductivity compared with that of conventional transparent conducting oxides hinders its commercial application. The addition of dimethyl sulfoxide (DMSO) is an efficient method to increase the conductivity of PEDOT: PSS. In PEDOT:PSS, the insulating PSS surrounds the conductive PEDOT. DMSO induces a separation of phases between PEDOT and PSS, forming an elongated PEDOT-rich network [4–7]. This significantly increases the conductivity of PEDOT:PSS, making it suitable for use as an electrode in organic light-emitting diodes, organic solar cells, and organic field-effect transistors.

However, the DMSO additive also modifies the electronic structure and changes the work function [8,9]. Consequently, the energy-level alignment between PEDOT:PSS and adjacent organic semiconductors may be altered. Achieving an energy-level alignment favorable for charge transport is crucial for improving device performance in organic electronics [10–13]. Therefore, strategies to enhance device performance must consider both efficient energy-level alignment and increased electrode conductivity. However, studies on the energy-level alignment at the modified PEDOT:PSS/organic semiconductor interface are insufficient, and further research is needed.

In this context, we investigated the influence of DMSO on the energy-level alignment of the PEDOT:PSS/C₆₀ interface. The interfacial electronic structures were determined via in-situ ultraviolet photoelectron spectroscopy (UPS) and X-ray photoelectron spectroscopy (XPS) methods. The in-situ analysis, performed without breaking the vacuum, provides accurate information about the interface formation, excluding surface contamination. C₆₀ was selected because it is an n-type organic semiconductor that can be thermally evaporated. We obtained the energy-level diagrams of the PEDOT:PSS/C₆₀ and DMSO-modified PEDOT:PSS (DMSO-PEDOT:PSS)/C₆₀ interfaces and compared the electron and hole-injection barriers.

An indium tin oxide glass substrate (AMG, South Korea) was wetcleaned via ultrasonication in deionized (DI) water, detergent, acetone, methanol, and DI water for 10 min each. The substrate was then dried under a flow of N₂ gas and treated with ultraviolet (UV)–O₃ using a PSDP-UV4T cleaner (Novascan Technologies, Inc., United States) at 100 ∘C for 15 min. A PEDOT:PSS solution (CleviosTM PH 1000, Heraeus GmbH & Co. KG, Germany) was ultrasonicated for 5 min to eliminate aggregation. PEDOT:PSS was then deposited on the substrate by spin-coating at a spin rate of 1,500 rpm for 150 s. Subsequently, the sample was annealed at 150 ∘C for 10 min using a hot plate. For comparison, DMSO (purity: >99.5 %, Sigma Aldrich, Inc., United States) was added to the PEDOT:PSS solution at 5 vol%, which is known to be the best performance [14]. DMSO-PEDOT:PSS was deposited using the same procedure. All the solution processing was conducted under ambient conditions.

The sample was loaded into the entry chamber of an in-situ UPS and XPS analysis system. A detailed description of this system has been provided previously [15]. The energy distribution of the photo-electrons was analyzed using a PHOIBOS 150 electron analyzer (SPECS GmbH, Germany) equipped with a He I_α (hν = 21.22 eV) discharge lamp and an Al K_α (hν = 1,486.7 eV) X-ray source. C₆₀ (purity: >99.9 %; Luminescence Technology Co., Taiwan) was deposited on PEDOT:-PSS in the preparation chamber using thermal evaporation at a rate of 0.01 nm s⁻¹. The sample was then transferred to the analysis chamber under vacuum. UPS and XPS spectra were collected at each deposition step, and this procedure was repeated until the spectral changes reached saturation. The deposited film thickness was monitored using a quartz crystal microbalance calibrated using the intensity attenuation of the substrate components [16]. The base pressures of the preparation and analysis chambers were 2 × 10⁻⁸ and 3 × 10⁻¹⁰ Torr, respectively. A sample bias of −5 V was applied to measure the secondary electron cutoff (SEC).

Figure 1 shows the UPS spectra representing (a) the SEC and (b) the highest occupied molecular orbital (HOMO) regions of the PEDOT: PSS/C₆₀ interface. The spectra of the SEC region were plotted on a kinetic energy scale for a direct representation of the work function and were normalized for clarity of comparison. In the spectrum of the HOMO region, the He I_β features were excluded to accurately determine the HOMO onset [17]. From Fig. 1(a), the work function of PEDOT:PSS was measured to be 4.64 eV. As the C₆₀ layer was deposited, the work function gradually increased. For C₆₀ thicknesses of 0.3, 1.2, 2.1, and 3.7 nm, the work functions were 4.85, 4.93, 4.94, and 4.94 eV, respectively. In Fig. 1(b), the HOMO peak of PEDOT:PSS was observed around 3.0 eV, with states continuing to the Fermi level due to its metallic properties. After the C₆₀ layer was covered, the C₆₀ HOMO peak appeared at approximately 2.0 eV. The characteristic peaks of C₆₀ were also observed at 3.4, 5.2, and 5.6 eV. The peak intensities increased with increasing C₆₀ thickness. As the thickness of the C₆₀ layer was increased, no significant shifts of the peaks were observed. The HOMO onset was determined from the 1.2 nm thickness, which shows a clear feature and is thicker than the molecular size. For C₆₀ thicknesses of 1.2, 2.1, and 3.7 nm, HOMO onsets of 1.51, 1.51, and 1.53 eV were observed.

Figure 1. In-situ UPS spectra of the (a) SEC and (b) HOMO regions of C₆₀ (0, 0.3, 1.2, 2.1, and 3.7 nm) on PEDOT:PSS. E_F denotes the Fermi level.

Figure 2 shows the XPS spectra of the (a) S 2p, (b) O 1s, and (c) C 1s regions of the PEDOT:PSS/C₆₀ interface. In Fig. 2(a), the S 2p components were observed at 166.2−171.5 and 162.2−165.9 eV, corresponding to PSS and PEDOT, respectively [18]. Each component exhibited two peaks owing to the spin-orbit splitting of 1.2 eV. The S−O−H and S−O−Na bonds of the PSS peaks and the asymmetric nature of the PEDOT peaks owing to the delocalized charge were considered in the spectral fitting [19,20]. For PEDOT:PSS, the area ratio of the S 2p spectrum in PSS/PEDOT was 2.2, which was slightly lower than the reported value [20]. The exact origin of this low ratio in the current study is unclear; however, it may be related to the slightly lower work function. In Fig. 2(b), the PEDOT:PSS spectrum shows two O 1s peaks at 532.7 and 531.4 eV, corresponding to C−O and S−O bonds, respectively. In both the S 2p and O 1s spectra, with the deposition of the C₆₀ layer, only the peak intensities decreased, and no other spectral changes were observed. In Fig. 2(c), the PEDOT:PSS spectrum shows two C 1s peaks at 286.3 and 284.4 eV, corresponding to the C−O/C−S and C−C bonds, respectively. With the deposition of the C₆₀ layer, the intensity of these features decreased, while a new peak at 284.7 eV, originating from C₆₀, became dominant. These changes in the XPS spectra indicate the successful deposition of the C₆₀ layer. However, no significant chemical interactions between PEDOT:PSS and C₆₀ were observed, which is consistent with the UPS results.

Figure 2. In-situ XPS spectra of the (a) S 2p, (b) O 1s, and (c) C 1s regions of C₆₀ (0, 0.3, 1.2, 2.1, and 3.7 nm) on PEDOT:PSS.

Figure 3 shows the UPS spectra of the (a) SEC and (b) HOMO regions of the DMSO-PEDOT:PSS/C₆₀ interface. As shown in Fig. 3(a), the work function of DMSO-PEDOT:PSS was measured to be 4.79 eV, which is 0.15 eV higher than that of pristine PEDOT:PSS [Fig. 1(a)]. This result is somewhat different from those reported in the literature, suggesting that DMSO decreases the work function of PEDOT:PSS [7]. A possible explanation is that at an already low PSS/PEDOT ratio, DMSO does not decrease the work function but may slightly increase it. The exact origin of this phenomenon should be investigated in future studies. As the C₆₀ layer was deposited, the work function gradually increased. For C₆₀ thicknesses of 0.3, 1.2, 2.1, and 3.7 nm, the work functions were 4.88, 4.96, 5.00, and 5.01 eV, respectively. As shown in Fig. 3(b), the spectral shape of DMSO-PEDOT:PSS was almost the same as that of pristine PEDOT:PSS [Fig. 1(b)]. With the deposition of the C₆₀ layer, the C₆₀ HOMO and other characteristic peaks appeared, similar to those of pristine PEDOT:PSS. The HOMO onsets of 1.43, 1.43, and 1.44 eV were located at C₆₀ thicknesses of 1.2, 2.1, and 3.7 nm, respectively. Therefore, no significant peak shift was observed at the DMSO-PEDOT:PSS/C₆₀ interface.

Figure 3. In-situ UPS spectra of the (a) SEC and (b) HOMO regions of C₆₀ (0, 0.3, 1.2, 2.1, and 3.7 nm) on DMSO-PEDOT:PSS. E_F denotes the Fermi level.

Figure 4 shows the XPS spectra of the (a) S 2p, (b) O 1s, and (c) C 1s regions of the DMSO-PEDOT:PSS/C₆₀ interface. As shown in Fig. 4(a), the S 2p spectrum of DMSO-PEDOT:PSS exhibits a shape similar to that of PEDOT:PSS with a slight peak shift toward lower binding energy (0.1 eV), which originates from the increased work function induced by DMSO. The area ratio of the S 2p spectrum in PSS/PEDOT was 2.2, which remained almost unchanged compared to that of PEDOT:PSS, indicating no significant effect on this ratio. In Fig. 4(b), the O 1s spectrum of DMSO-PEDOT:PSS was similar to that of PEDOT:PSS. Upon the deposition of the C₆₀ layer, the spectral intensities in the S 2p and O 1s regions decreased, similar to the PEDOT:PSS/C₆₀ case. Figure 4(c) shows the C 1s spectrum of DMSO-PEDOT: PSS, which mirrors that of PEDOT:PSS. With the deposition of the C₆₀ layer, the PEDOT:PSS signals attenuated, while a new peak at 284.6 eV from C₆₀ became prominent. Notably, no new signals from the DMSO residue were observed in any of the spectra. Overall, no significant peak shifts were observed during the C₆₀ layer deposition, suggesting that the chemical interactions between DMSO-PEDOT: PSS and C₆₀ remained marginal, similar to the case of PEDOT: PSS/C₆₀.

Figure 4. In-situ XPS spectra of the (a) S 2p, (b) O 1s, and (c) C 1s regions of C₆₀ (0, 0.3, 1.2, 2.1, and 3.7 nm) on DMSO-PEDOT:PSS.

Based on the UPS and XPS results, the energy-level diagrams of the PEDOT:PSS/C₆₀ and DMSO-PEDOT:PSS/C₆₀ interfaces are presented in Fig. 5. The ionization energy of C₆₀ was determined to be 6.45 and 6.47 eV, which agrees well with the previously reported value [21]. The transport gap of C₆₀ (2.25 eV) was obtained from literature results of UPS and inverse photoelectron spectroscopy [21]. Marginal HOMO level shifts (0.01−0.02 eV) within the experimental error were not considered in the analysis. The work function of PEDOT:PSS was measured to be 4.64 eV, and an interface dipole (0.30 eV) formed between PEDOT:PSS and C₆₀ for achieving thermal equilibrium [22]. The electron and hole-injection barriers at the PEDOT:PSS/C₆₀ interface were measured to be 0.72 and 1.53 eV, respectively. Therefore, electron injection from PEDOT:PSS to C₆₀ was more favorable than hole injection. In the case of DMSO-PEDOT:PSS/C₆₀, the work function of DMSO-PEDOT:PSS was measured to be 4.79 eV, and the interface dipole was 0.22 eV. Consequently, the electron and hole-injection barriers were measured to be 0.81 and 1.44 eV, respectively, due to the increased work function. Therefore, the addition of DMSO improves hole injection while deteriorating electron injection.

Figure 5. Energy-level diagrams of PEDOT:PSS/C₆₀ and DMSO-PEDOT:PSS/C₆₀. E_F, EVAC, Ψ, eD, IE, LUMO, and HOMO denote the Fermi level, vacuum level, work function, interface dipole, ionization energy, lowest unoccupied molecular orbital, and highest occupied molecular orbital, respectively.

We compared the energy-level alignments of the PEDOT:PSS/C₆₀ and DMSO-PEDOT:PSS/C₆₀ interfaces. The pristine PEDOT:PSS showed a work function of 4.64 eV, with the electron and hole-injection barriers to C₆₀ measured to be 0.72 and 1.53 eV, respectively. With the addition of DMSO, the work function of PEDOT:PSS increased to 4.79 eV, resulting in changed electron and hole-injection barriers of 0.81 and 1.44 eV, respectively. Therefore, the addition of DMSO is beneficial for junctions with p-type organic semiconductors but detrimental for those with n-type organic semiconductors. However, further studies on the interfaces of PEDOT:PSS modified with various additives are necessary to achieve ohmic contact with organic semiconductors.

This study was supported by the National Research Foundation of Korea (Grant No. NRF-2021R1A2C1009324 and 2018R1A6A1A03025-582).

The authors declare no conflicts of interest.