Correlation of Device Performance and Fermi Level Shift in the

Research Article www.acsami.org

Cite This: ACS Appl. Mater. Interfaces 2018, 10, 8877−8884

Correlation of Device Performance and Fermi Level Shift in the Emitting Layer of Organic Light-Emitting Diodes with Amine-Based Electron Injection Layers Sebastian Stolz,†,‡ Uli Lemmer,†,§ Gerardo Hernandez-Sosa,†,‡ and Eric Mankel*,‡,∥ †

Light Technology Institute, Karlsruhe Institute of Technology, Engesserstr. 13, 76131 Karlsruhe, Germany InnovationLab, Speyerer Str. 4, 69115 Heidelberg, Germany § Institute of Microstructure Technology, Karlsruhe Institute of Technology, Hermann-von-Helmholtz-Platz 1, 76344 Eggenstein-Leopoldshafen, Germany ∥ Materials Science Department, Surface Science Division, Technische Universität Darmstadt, Otto-Berndt-Straße 3, 64287 Darmstadt, Germany

ACS Appl. Mater. Interfaces 2018.10:8877-8884. Downloaded from pubs.acs.org by UNIV OF SUNDERLAND on 09/25/18. For personal use only.

‡

S Supporting Information *

ABSTRACT: We investigate three amine-based polymers, polyethylenimine and two amino-functionalized polyfluorenes, as electron injection layers (EILs) in organic light-emitting diodes (OLEDs) and find correlations between the molecular structure of the polymers, the electronic alignment at the emitter/EIL interface, and the resulting device performance. X-ray photoelectron spectroscopy measurements of the emitter/EIL interface indicate that all three EIL polymers induce an upward shift of the Fermi level in the emitting layer close to the interface similar to n-type doping. The absolute value of this Fermi level shift, which can be explained by an electron transfer from the EIL polymers into the emitting layer, correlates with the number of nitrogencontaining groups in the side chains of the polymers. Whereas polyethylenimine (PEI) and one of the investigated polyfluorenes (PFCON-C) have six such groups per monomer unit, the second investigated polyfluorene (PFN) only possesses two. Consequently, we measure Fermi level shifts of 0.5−0.7 eV for PEI and PFCON-C and only 0.2 eV for PFN. As a result of these Fermi level shifts, the energetic barrier for electron injection is significantly lowered and OLEDs which comprise PEI or PFCONC as an EIL exhibit a more than twofold higher luminous efficacy than OLEDs with PFN. KEYWORDS: OLEDs, polyethylenimine, amino-functionalized polyfluorenes, electron injection layers, solution processing

■

INTRODUCTION

be of conjugated as well as nonconjugated type, have been widely investigated as solution-processable electron injection layers (EILs) in OLEDs.7−15 In particular, polyethylenimine (PEI) received a lot of attention as its use as an EIL leads to a comparable OLED performance to state-of-the-art evaporated EILs. However, because of their nonconjugated backbone, the thickness of PEI films needs to be below 10 nm for a good OLED performance.8,9,11 This can be problematic with respect to the fabrication of such films by printing techniques as it has

An efficient operation of optoelectronic devices such as organic solar cells, organic field-effect transistors, or organic lightemitting diodes (OLEDs) requires a multilayer device architecture with a good electronic alignment between the electrodes and the transport levels of the adjacent organic layers.1,2 Typically, interlayers are applied between the electrodes and the functional layers to guarantee a good charge carrier injection/extraction.3−5 To enable the prospective lowcost fabrication of optoelectronic devices by high-throughput printing techniques, such interlayers do not only need to fulfill certain electronic requirements but also need to be solution processable.6 In recent years, amine-based polymers, which can © 2018 American Chemical Society

Received: October 27, 2017 Accepted: February 20, 2018 Published: February 20, 2018 8877

DOI: 10.1021/acsami.7b16352 ACS Appl. Mater. Interfaces 2018, 10, 8877−8884

Research Article

ACS Applied Materials & Interfaces

Figure 1. OLED stack and molecular structure of investigated EIL polymers (a), light-current-voltage (LIV) characteristics (b), current efficiency (c), and luminous efficacy (d) of prepared devices.

understanding the working mechanism of amine-based EIL materials.

been shown that the preparation of ultrathin homogeneous layers is usually very challenging because of hydrodynamic instabilities during the printing process or drying of the films.16,17 In that regard, amino-functionalized polyfluorenes are a very interesting class of solution-processable EIL polymers.12−14,18 As a result of their semiconducting conjugated backbone, the layer thickness of these materials is less critical for the performance of OLEDs. By adapting the amino side chains of a literature-known polymer (PFN), we recently synthesized an amidoamino-functionalized polyfluorene (PFCON-C) that exhibited the same device performance as PEI but a much wider processing window.18 In the literature, the working mechanism of such amine-based EILs has been widely explained by the formation of an interface dipole between the electrode and the EIL and hence a work function reduction of the electrode.8,9,11,19−22 In this work, we discuss an additional working mechanism for amine-based EILs besides the work function reduction of the electrode. After analyzing the performance of solutionprocessed OLEDs which comprise PEI, PFN, or PFCON-C as EILs, we investigate the electronic alignment at the emitter/ EIL interface by photoelectron spectroscopy. We show that all three polymers induce a shift of the Fermi level in the emitting layer close to the interface toward the lowest unoccupied molecular orbital (LUMO), similar to n-type doping. We find that the absolute value of this shift, which leads to a reduction of the energetic barrier for electron injection, correlates with the number of nitrogen-containing groups in the side chains of the polymers. For all three EILs, energy diagrams of the emitter/EIL interface are derived which are consistent with the observed OLED performance. Although an n-type doping by PEI-ethoxylated (PEIE), which has a molecular structure very similar to PEI, was already mentioned by Zhou et al.,8 following reports of PEI/PEIE mainly investigated the interface dipole induced by the polymer. In our opinion, however, the induced Fermi level shift in the emitting layer is crucial for fully

■

RESULTS AND DISCUSSION The OLED stack we investigate in this work and the molecular structure of the EIL polymers PEI, PFCON-C, and PFN can be seen in Figure 1a. The OLEDs comprise a poly(p-phenylene vinylene) derivative commonly known as super yellow (SY) as an emitting layer, and the EIL polymers are used in combination with Ag as the top electrode. Figure 1b presents the J−V−L characteristics of these devices. In Figure 1c,d, the current efficiency and luminous efficacy are plotted versus the luminance. As a reference, OLEDs without any EIL with only a plain Ag electrode are shown additionally. Because we discussed in detail the same device stack in our previous work,18 the OLED performance will only be briefly summarized here. Compared to the reference OLED with a pristine Ag cathode, all EIL polymers lead to a significantly improved device performance. However, in contrast to the OLEDs comprising PEI and PFCON-C, the OLED with PFN exhibits a much larger driving voltage and as a result a significantly lower luminous efficacy. The use of PEI and PFCON-C as EILs results in a very similar device performance with only marginal differences in luminous efficacy. To understand the differences in the device performance between PEI and PFCON-C on the one hand and PFN on the other hand, we investigate the interface SY/EIL by photoelectron spectroscopy. A usual way to investigate electronic semiconductor heterojunction properties such as electronic alignment and electrostatic potential drops is the stepwise adsorbate deposition on a substrate layer (e.g., via evaporation in ultrahigh vacuum) and subsequent XP spectra recording.23 As a thickness-controlled adsorbate layer growth from the liquid phase deposition is experimentally hard to handle, we change the adsorbate layer thickness by varying the polymer concentration in the solution. In Figure 2, XP spectra of SY 8878


Research Article


Figure 2. XP spectra of the interface SY/PEI. Besides a pristine SY sample, three SY samples covered by PEI films of different thicknesses were characterized. With the increasing PEI coverage, all spectra show an energetic shift to larger binding energies.

samples covered by PEI films are presented. To get insights on the electronic alignment at the interface, different PEI thicknesses/coverages are investigated by varying the PEI solid content in the spin-coating solution between 0.125 and 0.5 g L−1. The maximum concentration corresponds to the one used in the PEI OLED in Figure 1. PEI does not form homogeneous films on the top of SY;18 thus, we could not accurately determine the thickness of the PEI layers; however, with the increasing concentration, the average PEI thickness should increase. The molecular structure of SY, as it is known from the literature, consists of hydrogen, carbon, and oxygen.24 PEI on the other hand contains hydrogen, carbon, and nitrogen (see Figure 1). Therefore, the O 1s emissions in Figure 2 can be attributed to SY, and the N 1s emissions can be attributed to PEI. In contrast, the C 1s emissions contain contributions from both the polymers. As Figure 2b shows, the C 1s spectrum of SY can be fitted by two components at binding energies of about 284.7 and 286.2 eV with an intensity ratio of approximately 2.58:1. According to the molecular structure of SY, these components can be explained by the two different types of carbon atoms which are present in the polymer. Because of the inductive effect, the C 1s emission found at 286.6 eV is attributed to the carbon atoms with a C−O bond, whereas the one at a binding energy of 284.7 eV is attributed to the carbon atoms with only C−C bonds.25 When PEI is applied onto SY, the C 1s spectra tend to shift to larger binding energies. However, a detailed investigation of these spectra is not easily possible as they show a superposition of the signals from SY and from PEI. To separate both contributions, we calculated difference spectra. Therefore, we subtracted the C 1s spectrum of a PEI film on an Ag substrate (see Figure S1) from the C 1s spectra shown in Figure 2b. When calculating these difference spectra, we made sure that

the two components in the resulting C 1s spectra, which only contain the C 1s contributions of SY, exhibit the exact same intensity ratio that we observed in the C 1s spectrum of the pristine SY film. As the top right panel of Figure 2c shows, with the increasing PEI coverage, the C 1s substrate emission of the underlying SY film moves to larger binding energies. Starting from a binding energy of 284.7 eV, the main component of the C 1s emission is found at a binding energy of about 285.2 eV in the case of a PEI concentration of 0.5 g L−1. A similar energetic shift is observed in the case of the O 1s substrate emissions which are shown in Figure 2d. Whereas in the case of the pristine SY film, the O 1s emission is located at a binding energy of 533.3 eV, a binding energy of almost 533.9 eV is found for the SY/PEI sample with a PEI concentration of 0.5 g L−1. Although comparably small, with the increasing PEI coverage, an energetic shift to larger binding energies is also observed in the case of the N 1s adsorbate (PEI) emissions (Figure 2e). Finally, the bottom right panel of Figure 2f presents the secondary electron cutoff region of the XP spectra. With the increasing PEI coverage, the energetic position of the secondary electron cutoff also moves to larger binding energies. As the work function corresponds to the difference between the incoming photon energy and the position of the secondary electron cutoff, this translates into a decrease of the work function of the samples with the increasing PEI coverage. In Figure 3a, the determined energetic shifts of the core level emissions and of the work function are plotted versus the PEI concentration. As there is obviously no value for the N 1s emission in the case of a pristine SY film, the N 1s adsorbate data were plotted in such a way that they match the O 1s and C 1s substrate data at a PEI concentration of 0.125 g L−1. As can be seen, all energy levels shift approximately parallel with the increasing PEI concentration. Such a behavior cannot be explained by an interface dipole between SY and PEI which 8879


Research Article


sample and the average energetic shifts at a PEI concentration of 0.5 g L−1, the energy diagram shown in Figure 3b was derived. It was not possible to determine the highest occupied molecular orbital (HOMO) position of SY by ultraviolet photoelectron spectroscopy as it is commonly done because the sample got electrically charged during the measurement as a result of its rather low electric conductivity (see Figure S2). Therefore, in Figure 3b, the HOMO and LUMO energy levels of SY were taken from cyclic voltammetry (CV) measurements from the literature.26 The fact that energy levels, which were measured by different techniques, are used and compared in Figure 3b makes a detailed quantitative discussion of the derived energy diagram difficult. However, two important conclusions can be qualitatively drawn. First, because of the shift of the Fermi energy induced by PEI (i.e., the band bending in the SY layer), the energy difference between the LUMO and the Fermi energy at the SY/PEI interface is significantly reduced compared to that in a pristine SY film. As a result, the use of PEI as an EIL is expected to significantly reduce the energy barrier for electron injection. In the literature, the working mechanism of PEI as an EIL has been widely explained by the formation of an interface dipole and thus a reduction of the work function of the electrode.8,27 In our opinion, the observed Fermi level shift induced by PEI is an additional mechanism why PEI performs so well as EIL. Second, a space charge region formation is usually caused by a long-range integer charge transfer. In this case, the observed Fermi level shift can be explained by an electron transfer from PEI into SY. Such an electron transfer is consistent with the previous work from the literature where it was shown that PEI can be used as a reducing agent, that is, an n-type dopant.28−30 In that context, the question arises from which part of the PEI polymer these electrons are transferred. One likely possibility would be the amine groups of PEI. Because of their lone electron pair, amine groups are of basic character and can thus be expected to easily transfer electrons to other materials.

Figure 3. Energetic shifts from Figure 2 plotted vs the used PEI concentration (a). The fact that all energy levels shift in parallel indicates that PEI induces a shift of the Fermi level in SY toward the LUMO. By interpreting this shift as a band bending in SY, an energy diagram of the SY/PEI interface can be derived (b).

would result in a work function change at constant core level energies. Instead, the observed energy shifts indicate an energetic shift of the Fermi level in SY toward the LUMO. As a result, the core level binding energies (i.e., the energetic difference between the Fermi energy and the core levels) are affected in the same way as the work function (i.e., the energetic difference between the vacuum energy and the Fermi energy) of the samples. As X-ray photoelectron spectroscopy (XPS) is a very surface-sensitive technique with an information depth in the low nanometer range, we do not have any information on the position of the Fermi energy in the SY layer far from the SY/PEI interface. However, by assuming that the situation far from this interface corresponds to the one in the pristine SY film, the observed Fermi level shift can be interpreted as an electrostatic potential drop and the formation of a space charge region in the SY layer. It has to be pointed out, however, that we have no information about the spatial extension of this space charge region. By taking into account the determined values for the work function of the pristine SY

Figure 4. XP spectra of the interfaces SY/PFN and SY/PFCON-C. As in the case of PEI, all spectra move to larger binding energies with increasing thickness of the polyfluorene films. The observed shifts are significantly larger for PFCON-C than that for PFN. 8880


Research Article


Figure 5. Energetic shifts of PFN (open symbols) and PFCON-C (closed symbols) plotted vs the thickness of the polyfluorene films (a). As in the case of PEI, the observed energetic shifts can be explained by a shift of the Fermi level in SY close to the SY/EIL interface. By interpreting these shifts as a band bending in SY, energy diagrams of the interfaces SY/PFN and SY/PFCON-C were derived (b).

At a binding energy of about 532 eV, a new component develops and its intensity increases. According to the changes in the intensity, the component at higher binding energy can be attributed to SY and the one at lower binding energy to PFCON-C. In the case of the higher energetic O 1s component (SY), a maximum shift of about 0.5 eV is observed at a concentration of 2 g L−1 which corresponds to a PFCON-C thickness of about 7 nm. At a thickness of 11 nm (3 g L−1), the O 1s emission is again found at a slightly lower binding energy and finally at a concentration of 6 g L−1, which corresponds to a layer thickness of about 23 nm, the intensity of this component is very low and no accurate binding energy can be determined. In the case of both polyfluorenes, the N 1s adsorbate emissions move to larger binding energies when the adsorbate thickness is increased. For SY/PFN (Figure 4b), a maximum shift of about 150 meV is observed compared to a value of about 200 meV for SY/PFCON-C (Figure 4e). In both cases, these maximum values are reached at the respective maximum concentrations/ layer thicknesses. With the increasing adsorbate thickness, the position of the secondary electron cutoff also moves to larger binding energies, which corresponds to a reduction in the work function. As shown in Figure 4c for the interface SY/PFN, a maximum work function reduction of about 130 meV is observed at a PFN thickness of 3 nm (0.5 g L−1) and a further increase in the PFN thickness does not have any effect on the work function anymore. In contrast, in the case of SY/PFCON-C (Figure 4f), the work function steadily decreases with the increasing EIL thickness and a maximum reduction of about 800 meV is observed. In Figure 5a, the determined energy shifts are plotted versus the thickness of the polyfluorene films. As can be seen, similar to the interface SY/PEI, all energy levels shift approximately parallel with the increasing polyfluorene thickness which is a strong indication that both polymers induce a space charge region formation and, therefore, a Fermi level shift in SY toward the LUMO analogous to PEI. The O 1s emissions at thicknesses of about 11 nm (PFCON-C) and 12 nm (PFN) seem to deviate slightly from the other data points, however. In our opinion, this can be explained by the fact that the samples were in contact with ambient air for a few seconds when they were transferred into the photoelectron spectrometer (see Experimental Section details). Because of that, we expect a small amount of adsorbed oxygen on their surface. At the mentioned polyfluorene thicknesses, the O 1s substrate (SY) signal is already very low and the determined binding energy is

Furthermore, density functional theory calculations carried out on ethylamine predicted an electron transfer from that molecule onto surfaces of ZnO and gold.8 If it is true that PEI’s amine groups are involved in the observed electron transfer, a similar Fermi level shift at the SY/EIL interface should also happen in the cases of PFCON-C and PFN because both polyfluorenes contain amine groups as well. In that respect, it has to be noted that Zhong et al. observed a similar Fermi level shift in the fullerene C60 caused by a PFN-covered indium tin oxide (ITO) electrode.31 To investigate the interfaces SY/PFN and SY/PFCON-C, we prepared SY samples covered by films of both polymers with various thicknesses. For both polyfluorenes, we used a minimal concentration of 0.5 g L−1 and a maximum concentration of 6 g L−1 (PFCON-C) or 4 g L−1 (PFN). Because both PFN and PFCON-C form homogeneous layers on top of SY,18 corresponding film thicknesses could be determined for all concentrations. In the case of both polyfluorenes, the maximum concentration corresponds to a thickness of 20−25 nm and matches the one used for the OLEDs in Figure 1. In Figure 4, the O 1s and N 1s core level emissions as well as secondary electron cutoff spectra of SY/PFN (top row) and SY/PFCONC (bottom row) are presented. As can be seen, similar to the interface SY/PEI, all spectra tend to move to larger binding energies with increasing polyfluorene thickness. The C 1s core level emissions of the samples are shown in Figure S3. A detailed investigation of these spectra will be skipped as both SY and the polyfluorenes contain carbon which makes an interpretation of the spectral characteristics difficult. The O 1s emissions of the SY/PFN samples shown in Figure 4a can be attributed to SY as PFN does not contain any oxygen. With the increasing PFN adsorbate thickness, an energetic shift to slightly larger binding energies is observed. In comparison to the interface SY/PEI, this shift is significantly smaller with a maximum value of only about 150 meV at a PFN concentration of 1 g L−1 which corresponds to a layer thickness of approximately 6 nm. When the PFN thickness is further increased to 12 nm (2 g L−1), the O 1s emission moves back to a slightly smaller binding energy and at a layer thicknesses of about 25 nm (4 g L−1) and the intensity of the O 1s emission is so low that its binding energy cannot be accurately determined. In contrast to PEI or PFN, PFCON-C contains oxygen. Therefore, the O 1s spectra of the SY/PFCON-C samples in Figure 4d contain contributions from both the polymers. With the increasing PFCON-C thickness, the following observations can be made. The initial O 1s component at a binding energy of 533.3 eV moves to a higher binding energy and gets damped. 8881


Research Article


As this Fermi level shift in SY correlates with the number of amine/amide groups in the side chains of the investigated EIL polymers, it might be caused by an electron transfer from the lone electron pairs of amine/amide groups into the SY layer. In that context, it has to be pointed out that the basicity and hence the electron-donating properties of amine groups differ depending on their type. In solvents which cannot form hydrogen bridge bonds, ternary amines possess the highest basicity and primary amines the lowest basicity. Furthermore, the basicity of amides is lower than that of amines because of the inductive effect of oxygen.32 Therefore, in future, it should be investigated how the observed Fermi level shift in SY compares for EIL polymers which have the same number of amine or amide groups but of different types.

probably slightly affected by the O 1s signal of the adsorbed oxygen on the surface of the samples. The observed Fermi level shifts are vastly different for PFN and PFCON-C. Whereas in the case of PFN, a Fermi level shift (average of work function shift and N 1s shift) of approximately 220 meV is observed, the respective value equals more than 700 meV in the case of PFCON-C. These numbers correlate well with the number of nitrogen atoms with lone electron pairs in the side chains of both polymers. Whereas only two amine groups exist in the side chains of PFN, PFCON-C possesses two amide groups and four amine groups in its side chains. Therefore, the same amount of nitrogen atoms with lone electron pairs are found in the side chains of PFCON-C and PEI. Consequently, in the case of PFCON-C and PEI, the Fermi level shifts are significantly larger than that in the case of PFN. The fact that the maximum observed Fermi level shifts differ for PEI (∼0.5 eV) and PFCON-C (∼0.7 eV), despite both polymers possessing the same number of amine groups in their side chains, might be explained by the different thicknesses of the polymer layers. As mentioned before, we could not accurately measure the thickness of the PEI films18 but we expect them to be below 10 nm for all investigated concentrations (as it was widely discussed in the literature, thicker layers would not lead to a good OLED performance8,9,11). For a thickness of 10 nm, the observed Fermi level shift for PFCON-C is about 0.5 eV, a value which is very similar to the maximum shift that we observed for PEI. It is important to note again that such a Fermi level shift cannot be explained by the intrinsic molecular dipole moment associated with the amine groups of the EIL polymers. A common orientation of these molecular dipoles would lead to a short-range interface dipole and cannot explain the formation of a long-range space charge region, which is usually caused by an integer charge transfer. By again interpreting the observed Fermi level shifts as a band bending in SY, the energy diagrams of the interfaces SY/ PFN and SY/PFCON-C shown in Figure 5b were derived. As in the case of Figure 3b, the HOMO and LUMO positions of SY were taken from the literature.26 The HOMO positions of both polyfluorenes were measured by CV, and the LUMO positions were estimated by the optical band gap of the polymers. The respective measurements can be found in our previous work.18 As Figure 5b shows, the energetic difference between the Fermi level and the LUMO position in the two polyfluorene films differs significantly. In the case of PFN, a value of 1.73 eV is determined in contrast to a value of only 0.94 eV in the case of PFCON-C. According to these band diagrams, one therefore expects a significantly larger energy barrier for electron injection when PFN is used as an EIL compared to PFCON-C. This explains why OLEDs with PFN showed much larger driving voltages and as a result a significantly lower luminous efficacy than OLEDs with PFCON-C in Figure 1. Moreover, it should be noted that the determined energetic difference between the Fermi level and the LUMO in the first semiconducting layer next to the cathode is almost the same for the interface SY/PEI (0.93 eV) and SY/PFCON-C (0.94 eV). The OLED performance correlates very well with these values as the luminous efficacy of OLEDs with PEI and PFCON-C is almost identical (see Figure 1). This underlines that the observed Fermi level shifts are highly relevant for a deep understanding of the working mechanism of amine-based EIL materials.

■

CONCLUSIONS In this work, we investigated three amine-based EIL polymers, PEI and two amino-functionalized polyfluorenes (PFN and PFCON-C) and we found correlations between the molecular structure of the polymers, the device performance of the solution-processed OLEDs, and the electronic alignment at the emitter/EIL interface. We showed by XPS that the investigated EIL polymers induce an energetic shift of the Fermi level in the emitting layer close to this interface toward the LUMO and as a result, the energy barrier for electron injection is reduced. The observed Fermi level shifts correlated very well with the number of amine/amide groups in the side chains of the investigated EIL polymers. In the cases of PEI and PFCON-C, which both possess six such groups per monomer unit, we measured a Fermi level shift in the range of 0.5−0.7 eV in comparison to a value of only 0.2 eV for PFN, which only has two amine groups per monomer unit. Correspondingly, OLEDs which comprise PEI or PFCON-C as an EIL exhibited a twofold higher luminous efficacy than OLEDs with PFN. Our data suggest that the observed Fermi level shifts are caused by an electron transfer from the basic nitrogen-containing amine/ amide groups of the EIL materials into the emitting layer because such groups possess a strong electron-donating character. We therefore think that future work should investigate the performance of EILs with respect to the basicity of their functional groups. This could deliver valuable information for the design of high-performance interlayer materials to be used in solution-processable devices such as OLEDs or solar cells.

■

EXPERIMENTAL SECTION

PFN and PFCON-C Synthesis. PFN and PFCON-C were synthesized as reported previously.18,33 Preparation of Solutions. PEI and polyfluorene solutions were prepared by dissolving the polymers in 1-propanol. In the case of PFN, 1 vol % acetic acid was added to increase the solubility. PFN and PFCON-C solutions were filtered with a 0.45 μm poly(vinylidene difluoride) (PVDF) filter prior to use. PDY-132 (SY, light-emitting polymer), acquired from Merck KGaA, was dissolved in toluene and used without filtering. Poly(3,4-ethylenedioxythiophene) polystyrene sulfonate (PEDOT:PSS, VPAi 4083, acquired from Heraeus) was filtered with a 0.45 μm PVDF filter prior to use. OLED Fabrication. Glass substrates covered by 180 nm of ITO (10 Ω sq−1 from Kintec) were subsequently cleaned in acetone and isopropanol under sonication for 15 min and treated by O2 plasma for 5 min. PEDOT:PSS was spin-cast and annealed (135 °C for 20 min) under ambient conditions. Spin-coating parameters were ω = 3800 rpm, a = 1000 rpm s−1, and t = 30 s such that layers of 25 nm were obtained. PDY-132 (5 g L−1 in toluene) was spin-cast in an inert 8882



■

atmosphere. Spin-coating parameters were ω = 2000 rpm, a = 1000 rpm s−1, and t = 60 s so that a layer of 65 nm was obtained. PEI (0.5 g L−1), PFN (4 g L−1), and PFCON-C (6 g L−1) were spin-cast and annealed (120 °C for 10 min) in inert atmosphere. Spin-coating parameters were ω = 5000 rpm, a = 1000 rpm s−1, and t = 60 s for PEI and ω = 3000 rpm, a = 1000 rpm s−1, and t = 60 s for PFN and PFCON-C. Ag layers were thermally evaporated in a vacuum system with a base pressure of 1 × 10−7 mbar at a rate of 0.2 nm s−1. XPS Sample Preparation. Glass substrates covered by 180 nm of ITO (10 Ω sq−1 from Kintec) were subsequently cleaned in acetone and isopropanol under sonication for 15 min and treated by O2 plasma for 5 min. PDY-132 (3 g L−1 in toluene, a layer thickness of 30 nm), PEI, PFN, and PFCON-C were spin-cast in an inert atmosphere. Spincoating and annealing parameters matched the ones used for the fabrication of OLEDs. During the transfer to the photoelectron spectrometer, samples were in contact with ambient atmosphere for a short amount of time. Device Characterization. Devices were characterized with a Botest LIV Functionality Test System. The driving voltage of the system can be varied between −20 and 40 V, and the current measurement resolution is as low as 5 nA. X-ray Photoelectron Spectroscopy. The photoelectron spectroscopy characterization was performed using a PHI VersaProbe II Scanning XPS Microprobe located at the InnovationLab in Heidelberg. The base pressure of the spectrometer is 1 × 10−9 mbar, and it is equipped with a monochromatized Al Kα X-ray source (1486.6 eV) and a concentric hemispherical analyzer. Detail spectra of the core level lines were recorded with a pass energy of 11.75 eV. The energetic resolution determined by the 2σ Gaussian broadening used to fit the Fermi edge of a freshly sputter-cleaned silver sample measured at room temperature is 0.35 eV. The spectra are referenced in binding energy with respect to the Fermi edge and the core level lines of in situ cleaned Au, Ag, and Cu metal foils. Secondary electron cutoff spectra were measured with a pass energy of 2.95 eV.

■

REFERENCES

(1) Yin, Z.; Wei, J.; Zheng, Q. Interfacial Materials for Organic Solar Cells: Recent Advances and Perspectives. Adv. Sci. 2016, 3, 1500362. (2) Geffroy, B.; le Roy, P.; Prat, C. Organic Light-Emitting Diode (OLED) Technology: Materials, Devices and Display Technologies. Polym. Int. 2006, 55, 572−582. (3) Chiba, T.; Pu, Y.-J.; Kido, J. Solution-Processable Electron Injection Materials for Organic Light-Emitting Devices. J. Mater. Chem. C 2015, 3, 11567−11576. (4) Parker, I. D. Carrier Tunneling and Device Characteristics in Polymer Light-Emitting Diodes. J. Appl. Phys. 1994, 75, 1656. (5) Ishii, H.; Sugiyama, K.; Ito, E.; Seki, K. Energy Level Alignment and Interfacial Electronic Structures at Organic/Metal and Organic/ Organic Interfaces. Adv. Mater. 1999, 11, 605−625. (6) Zhong, C.; Duan, C.; Huang, F.; Wu, H.; Cao, Y. Materials and Devices toward Fully Solution Processable Organic Light-Emitting Diodes. Chem. Mater. 2011, 23, 326−340. (7) Xiong, T.; Wang, F.; Qiao, X.; Ma, D. A Soluble Nonionic Surfactant as Electron Injection Material for High-Efficiency Inverted Bottom-Emission Organic Light Emitting Diodes. Appl. Phys. Lett. 2008, 93, 123310. (8) Zhou, Y.; Fuentes-Hernandez, C.; Shim, J.; Meyer, J.; Giordano, A. J.; Li, H.; Winget, P.; Papadopoulos, T.; Cheun, H.; Kim, J.; Fenoll, M.; Dindar, A.; Haske, W.; Najafabadi, E.; Khan, T. M.; Sojoudi, H.; Barlow, S.; Graham, S.; Brédas, J.-L.; Marder, S. R.; Kahn, A.; Kippelen, B. A Universal Method to Produce Low-Work Function Electrodes for Organic Electronics. Science 2012, 336, 327−332. (9) Stolz, S.; Scherer, M.; Mankel, E.; Lovrinčić, R.; Schinke, J.; Kowalsky, W.; Jaegermann, W.; Lemmer, U.; Mechau, N.; HernandezSosa, G. Investigation of Solution-Processed Ultrathin Electron Injection Layers for Organic Light-Emitting Diodes. ACS Appl. Mater. Interfaces 2014, 6, 6616−6622. (10) Stolz, S.; Zhang, Y.; Lemmer, U.; Hernandez-Sosa, G.; Aziz, H. Degradation Mechanisms in Organic Light-Emitting Diodes with Polyethylenimine as a Solution-Processed Electron Injection Layer. ACS Appl. Mater. Interfaces 2017, 9, 2776−2785. (11) Kim, Y.-H.; Han, T.-H.; Cho, H.; Min, S.-Y.; Lee, C.-L.; Lee, T.W. Polyethylene Imine as an Ideal Interlayer for Highly Efficient Inverted Polymer Light-Emitting Diodes. Adv. Funct. Mater. 2014, 24, 3808−3814. (12) Wu, H.; Huang, F.; Mo, Y.; Yang, W.; Wang, D.; Peng, J.; Cao, Y. Efficient Electron Injection from a Bilayer Cathode Consisting of Aluminum and Alcohol-/Water-Soluble Conjugated Polymers. Adv. Mater. 2004, 16, 1826−1830. (13) Huang, F.; Wu, H.; Cao, Y. Water/alcohol Soluble Conjugated Polymers as Highly Efficient Electron Transporting/injection Layer in Optoelectronic Devices. Chem. Soc. Rev. 2010, 39, 2500−2521. (14) Hu, Z.; Zhang, K.; Huang, F.; Cao, Y. Water/alcohol Soluble Conjugated Polymers for the Interface Engineering of Highly Efficient Polymer Light-Emitting Diodes and Polymer Solar Cells. Chem. Commun. 2015, 51, 5572−5585. (15) Lee, B. R.; Lee, S.; Park, J. H.; Jung, E. D.; Yu, J. C.; Nam, Y. S.; Heo, J.; Kim, J.-Y.; Kim, B.-S.; Song, M. H. Amine-Based Interfacial Molecules for Inverted Polymer-Based Optoelectronic Devices. Adv. Mater. 2015, 27, 3553−3559. (16) Xie, R.; Karim, A.; Douglas, J. F.; Han, C. C.; Weiss, R. A. Spinodal Dewetting of Thin Polymer Films. Phys. Rev. Lett. 1998, 81, 1251−1254. (17) Bornemann, N.; Sauer, H. M.; Dörsam, E. Gravure Printed Ultrathin Layers of Small-Molecule Semiconductors on Glass. J. Imaging Sci. Technol. 2011, 55, 040201. (18) Stolz, S.; Petzoldt, M.; Dück, S.; Sendner, M.; Bunz, U. H. F.; Lemmer, U.; Hamburger, M.; Hernandez-Sosa, G. High-Performance Electron Injection Layers with a Wide Processing Window from an Amidoamine-Functionalized Polyfluorene. ACS Appl. Mater. Interfaces 2016, 8, 12959−12967. (19) Kang, H.; Hong, S.; Lee, J.; Lee, K. Electrostatically SelfAssembled Nonconjugated Polyelectrolytes as an Ideal Interfacial Layer for Inverted Polymer Solar Cells. Adv. Mater. 2012, 24, 3005.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsami.7b16352. C 1s core level spectrum of a PEI film on the top of an Ag substrate, C 1s core level spectra of a SY sample without and with additional UV illumination from a helium discharge lamp, and C 1s core level spectra of the interfaces SY/PFN and SY/PFCON-C (PDF)

■

Research Article

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Sebastian Stolz: 0000-0001-8804-0207 Gerardo Hernandez-Sosa: 0000-0002-2871-6401 Eric Mankel: 0000-0001-6566-157X Author Contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors thank M. Petzoldt and M. Hamburger for fruitful discussion. We acknowledge the support of the German Federal Ministry for Education and Research through the projects Poesie (FKZ 13N13691) and InterPhase (FKZ 13N13658). 8883


Research Article

ACS Applied Materials & Interfaces (20) Kyaw, A. K. K.; Wang, D. H.; Gupta, V.; Zhang, J.; Chand, S.; Bazan, G. C.; Heeger, A. J. Efficient Solution-Processed SmallMolecule Solar Cells with Inverted Structure. Adv. Mater. 2013, 25, 2397−2402. (21) Saracco, E.; Bouthinon, B.; Verilhac, J.-M.; Celle, C.; Chevalier, N.; Mariolle, D.; Dhez, O.; Simonato, J.-P. Work Function Tuning for High-Performance Solution-Processed Organic Photodetectors with Inverted Structure. Adv. Mater. 2013, 25, 6534−6538. (22) He, Z.; Zhong, C.; Su, S.; Xu, M.; Wu, H.; Cao, Y. Enhanced Power-Conversion Efficiency in Polymer Solar Cells Using an Inverted Device Structure. Nat. Photonics 2012, 6, 591−595. (23) Mankel, E.; Hein, C.; Kühn, M.; Mayer, T. Electric Potential Distributions in Space Charge Regions of Molecular Organic Adsorbates Using a Simplified Distributed States Model. Phys. Status Solidi A 2014, 211, 2040−2048. (24) Gambino, S.; Chen, S.; Samuel, I.; André, P. Effect of Coated Magnetic Nanocolloids on Hybrid Organic LEDs. SPIE Newsroom 2010, DOI: 10.1117/2.1201012.003314. (25) XPS Reference Pages. http://www.xpsfitting.com/search/label/ carbon (accessed Jan 11, 2017). (26) Sandström, A.; Matyba, P.; Edman, L. Yellow-Green LightEmitting Electrochemical Cells with Long Lifetime and High Efficiency. Appl. Phys. Lett. 2010, 96, 053303. (27) Höfle, S.; Schienle, A.; Bruns, M.; Lemmer, U.; Colsmann, A. Enhanced Electron Injection into Inverted Polymer Light-Emitting Diodes by Combined Solution-Processed Zinc Oxide/Polyethylenimine Interlayers. Adv. Mater. 2014, 26, 2750−2754. (28) Xuan, Y.; Sandberg, M.; Berggren, M.; Crispin, X. An AllPolymer-Air PEDOT Battery. Org. Electron. 2012, 13, 632−637. (29) Maiti, U. N.; Lee, W. J.; Lee, J. M.; Oh, Y.; Kim, J. Y.; Kim, J. E.; Shim, J.; Han, T. H.; Kim, S. O. 25th Anniversary Article: Chemically Modified/Doped Carbon Nanotubes & Graphene for Optimized Nanostructures & Nanodevices. Adv. Mater. 2014, 26, 40−67. (30) Fabiano, S.; Braun, S.; Liu, X.; Weverberghs, E.; Gerbaux, P.; Fahlman, M.; Berggren, M.; Crispin, X. Poly(ethylene Imine) Impurities Induce N-Doping Reaction in Organic (semi)Conductors. Adv. Mater. 2014, 26, 6000−6006. (31) Zhong, S.; Wang, R.; Ying Mao, H.; He, Z.; Wu, H.; Chen, W.; Cao, Y. Interface Investigation of the Alcohol-/water-Soluble Conjugated Polymer PFN as Cathode Interfacial Layer in Organic Solar Cells. J. Appl. Phys. 2013, 114, 113709. (32) Sykes, P. A Guidebook to Mechanism in Organic Chemistry, 6th ed.; Longman Group UK Ltd., 1986. (33) Stolz, S.; Petzoldt, M.; Kotadiya, N.; Rödlmeier, T.; Eckstein, R.; Freudenberg, J.; Bunz, U. H. F.; Lemmer, U.; Mankel, E.; Hamburger, M.; Hernandez-Sosa, G. One-Step Additive Crosslinking of Conjugated Polyelectrolyte Interlayers: Improved Lifetime and Performance of Solution-Processed OLEDs. J. Mater. Chem. C 2016, 4, 11150−11156.

8884


Correlation of Device Performance and Fermi Level Shift in the

Recommend Documents