Efficient, Air‐Stable Bulk Heterojunction Polymer Solar Cells Using MoOx as the Anode Interfacial Layer

The use of molybdenum oxide as the anode interfacial layer in conventional bulk heterojunction polymer solar cells leads to an improved power conversion efficiency and also dramatically increases the device stability. This indicates that the engineering of improved anode interface materials is an important method by which to fabricate efficient and stable polymer solar cells. The development of bulk heterojunction (BHJ) polymer solar cells continues to show progress.1-6 Power conversion efficiencies (PCE) in the range of 6–8% have been recently reported.7-12 However, for large-scale commercialization, further improvements in PCE and stability are needed. A BHJ polymer solar cell generally consists of a transparent anode (typically indium tin oxide, ITO), a BHJ active layer (a mixture of donor polymer and fullerene acceptor), and a cathode (e.g., Al or Ca). Anode/cathode interfacial layers are used as charge selective contacts between the BHJ active layer and the electrodes.13, 14 Experiments have shown that the lifetime of BHJ solar cells can be extended by inserting a cathode interfacial layer such as titanium oxide (TiOx) or lithium fluoride (LiF) between the BHJ layer and the cathode.15-17 Moreover, cathode interfacial layers (e.g., TiOx or zinc oxide) also function to redistribute the light intensity inside the BHJ active layer, thus enhancing the harvesting of photons.18-20 Studies of the effect of the anode interfacial layers are relatively few because poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS) is typically used. PEDOT:PSS is, however, both hygroscopic and acidic with an associated reduction in device stability. We report here that the use of molybdenum oxide (MoO3) as the anode interfacial material improves the PCE and significantly increases the device stability. Transition metal oxides, such as nickel oxide (NiO), MoO3, vanadium oxide (V2O5), and tungsten oxide (WO3) have been successfully demonstrated as replacements for PEDOT:PSS.21-24 With NiO as the anode interfacial layer (AIL) in poly(3-hexylthiophene):phenyl-C61-butyric acid methyl ester (P3HT:PCBM) solar cells, a PCE of 5.2% was reported.21 Herein, we selected MoO3 for in-depth studies because of its high transmittance (a wide bandgap ≈3.0 eV) and ease of processing (low evaporation temperature, melting point ≈795 °C). BHJ solar cells fabricated with polyN-9”-hepta-decanyl-2,7-carbazole-alt-5,5-(4’,7’-di-2-thienyl-2’,1’,3’-benzothiadiazole):6,6-phenyl C70-butyric acid methyl ester (PCDTBT:PC70BM) as the photoactive layer and MoO3 as the anode interfacial layer show performance that is superior to cells fabricated with PEDOT:PSS. First, a higher PCE is achieved at active layer thickness up to 200 nm. Second, the use of MoOx leads to improved air stability: PCEs of unencapsulated PCDTBT:PC70BM solar cells remain above 50% of the original value, even after storage in air for 720 h, while the cells fabricated with PEDOT:PSS show rapid degradation. The molecular structures of PCDTBT and PC70BM are shown in Figure 1a. Thermally evaporated MoOx films with thickness of 6 nm, 9 nm, 12 nm, and 22 nm were used as anode interfacial layers. Figure 1b shows the ultraviolet photoelectron spectroscopy (UPS) spectra of ITO and MoOx films on ITO. The valence band maximum (VBM) of MoOx films is 7.07 ± 0.04 eV and the work function is 5.44 ± 0.02 eV, which is close to the energy of the highest occupied molecular orbital (HOMO) of PCDTBT (Figure 1c). The stoichiometry of MoOx films deposited on an ITO substrate was examined by X-ray photoelectron spectroscopy (XPS) (Supporting Information, Figure S2). Mo 3d XPS spectra are plotted for MoOx films with different thickness. From the spin-orbit splitting, the Mo 3d core levels exhibit Mo 3d3/2 (235.8 eV) and Mo 3d5/2 (232.7 eV) with an intensity ratio of 2:3, which is consistent with literature values of MoO3.25 In addition, MoOx films deposited on ITO are composed of mainly Mo6+ (the small signal from Mo5+ is barely detectable). Thus, the XPS results imply that the chemical composition of the MoOx films is close to the stoichiometric MoO3. Moreover, in order to quantify the Mo:O ratio, Mo 3d and O 1s XPS spectra for MoOx films with different thickness were analyzed. The atomic ratio of Mo to O in MoOx films is approximately 1:2. a) Molecular structures of PCDTBT and PC70BM. b) UPS spectra of ITO and MoOx films on ITO. c) Energy level diagram of the component materials in PCDTBT:PC70BM solar cells fabricated with MoOx as the anode interfacial layer. Current density–voltage (J–V) characteristics under AM 1.5 G irradiation (100 mW cm−2) of PCDTBT:PC70BM solar cells fabricated with MoOx as anode interfacial layers are shown in Figure 2a. The device performance data obtained with a BHJ layer thickness of 110 nm are summarized in Table 1. The optimal thickness of MoOx was found to be 9 nm. This device shows a PCE = 6.50%, a short-circuit current Jsc = 10.88 mA cm−2, an open-circuit voltage Voc = 0.89 V, and a fill factor FF = 0.67. In contrast, the control device (fabricated with PEDOT:PSS) shows a PCE = 5.95%. Figure 2b shows the incident photon conversion efficiency (IPCE) spectra of the devices. An increase of IPCE at wavelengths between 350 nm and 500 nm was observed for devices with the MoOx layer in comparison to the control device. The maximum IPCE is over 70%, indicative of efficient photon-to-electon conversion. a) J–V characteristics of PCDTBT:PC70BM solar cells. The control device was fabricated with PEDOT:PSS. b) IPCE spectra of PCDTBT:PC70BM solar cells fabricated with PEDOT:PSS and MoOx. Atomic force microscopy (AFM) measurements of the MoOx films and PEDOT:PSS were carried out (Supporting Infor­mation, Figure S3). MoOx films consist of small islands that grow in size as the film thickness is increased. This “island” morphology extends the contact area between PCDTBT:PC70BM and MoOx, which may allow for efficient hole collection. However, it makes devices with thin-active layers susceptible to shorts. The optimal 9-nm-thick MoOx film exhibits a much rougher surface than PEDOT:PSS; however, AFM images of PCDTBT:PC70BM cast on PEDOT:PSS and MoOx exhibit no significant differences with root mean square (RMS) roughness values of 0.5 nm and 0.6 nm, respectively. Figure 3 compares the thickness dependence of Jsc, Voc, FF, and PCEs of PCDTBT:PC70BM solar cells based on PEDOT:PSS (40 nm) and MoOx (9 nm). The Voc and FF remain relatively constant as the thickness of the active layer increased up to 130 nm for both anode interfacial layers. This indicates that morphology and recombination do not severely limit device performance as the thickness of the active layer increases. Figure 3a shows the variation of Jsc with active layer thickness. For devices based on PEDOT:PSS, a significant decrease in the short circuit current is seen for active layers between 70 nm and 170 nm. MoOx-based devices also show a decrease in Jsc but it not as severe. At the active layer thickness of 130 nm, the Jsc of the device based on PEDOT:PSS is reduced to 9.6 mA cm−2 and the device based on MoOx is 10.5 mA cm−2. Jsc (a), Voc (b), FF (c), and PCE (d) of PCDTBT:PC70BM solar cells fabricated with PEDOT:PSS (40 nm) and MoOx (9 nm) as a function of the active layer thickness under AM 1.5 G irradiation with an irradiation intensity of 100 mW cm−2. The increased Jsc of the MoOx-based cells is attributed primarily to favorable redistribution of the light intensity within the active layer. The measured index of refraction of PCPDTBT:PC70BM active layer (≈2.1) nearly matches the MoOx (≈2.0) but this differs substantially from PEDOT:PSS (≈1.4). The electric field distribution within the device was simulated using the transfer matrix method for a variety of active layer thicknesses from 70 nm to 210 nm.26-28 Because of the refractive index matching between the MoOx and PCDTBT:PC70BM layers, the optical simulation indicates that the electric field intensity is redistributed inside the PCDTBT:PC70BM layer compared to the device based on PEDOT:PSS. Optical simulations predict up to a maximum of 16% increase in light absorption for MoOx devices (Supporting Information, Figure S6). The simulations are in good agreement with the measured Jsc and IPCE. For example, at a BHJ layer thickness of 110 nm, the simulated result indicates a ≈15% improvement in Jsc; the improvement of the measured Jsc is ≈8%. As shown in Figure 3, BHJ solar cells with MoOx as the anode interfacial layer show higher Jsc, Voc, and FF at thicker active layer thickness than devices based on PEDOT:PSS. PCEs >6% were maintained for thicknesses as large as 200 nm, which is among the highest PCE values reported so far for polymer solar cells with relatively thick active layers. Generally, high-efficiency polymer solar cells have been reported with active layer thickness less than 100 nm,7-12 which is a challenge for the manufacture of polymer solar cells using full roll-to-roll processing. Here, we found that refractive index matching between the anode interfacial layer and the BHJ layer could increase the light absorption, resulting in a high PCE at a thicker active layer film, which may provide an approach to address the issue of the active layer thickness for polymer solar cells. By further optimizing the optical effects using a structured antireflection coating, a high PCE up to 7.2% was achieved with Jsc = 11.95 mA cm−2, Voc = 0.907 V, and FF = 0.664 (Supporting Information, Figure S8). The efficiency was certified by Newport Corporation. The air stability of PCDTBT:PC70BM solar cells fabricated with a 9-nm-thick MoOx film and PEDOT:PSS as a function of storage time in air under ambient conditions is shown in Figure 4. The unencapsulated BHJ solar cells (conventional architecture; not inverted) based on the MoOx film exhibit significantly better air stability. The PCE remains at approximately 50% of the original value even after storage in air for 720 h (30 days), while the PCE of control PEDOT:PSS device degraded by a factor of 2 after air exposure for only 16 h and fell to less than 10% of the original value after storage in air for 480 h. It is well known that PEDOT:PSS is a polymer mixture of two ionomers. PSS is present in excess compared to PEDOT. It is known that PSS can diffuse to other layers and react with other components.29 PSS is also hygroscopic and acidic and, therefore, a source of interface instability. The use of MoOx as the anode interfacial layer in BHJ solar cells yields a significant improvement in air stability. Normalized PCEs as a function of storage time for PCDTBT:PC70BM solar cells fabricated with PEDOT:PSS and MoOx in air under ambient conditions (no encapsulation). In summary, efficient and stable BHJ polymer solar cells in the conventional (non-inverted) architecture based on a PCDTBT:PC70BM active layer and MoOx as the anode interfacial layer were fabricated. The use of the MoOx improved light absorption within the active layer and thereby leads to a PCE >6% at BHJ layer thicknesses up to 200 nm. A further improvement in PCE up to 7.2% was demonstrated by using an antireflection coating. Additionally, BHJ solar cells based on MoOx exhibit superior long-term air stability when compared to the cells fabricated with PEDOT:PSS. Our results indicate that the engineering of improved anode interface materials is an important opportunity for achieving efficient and stable polymer solar cells. Materials: PCDTBT (molecular weight, Mw ≈ 100 000; polydispersity index, PDI ≈ 2.75) and PC70BM were supplied by Konarka Technologies. MoOx was purchased from Sigma-Aldrich (≈99.5%) and used as received. Fabrication of PSCs: PSCs were fabricated on ITO-coated glass substrates. The ITO-coated glass substrates were first cleaned with detergent, ultrasonicated in water, acetone, and isopropyl alcohol, and subsequently dried overnight in an oven. MoOx films were deposited onto ITO substrates by thermal evaporation in a vacuum of about 1 × 10−6 Torr. The evaporation rate was 0.1 Å s−1. For the control device, PEDOT:PSS (Baytron PH) was spin-cast from aqueous solution at 5000 rpm for 40 s to form a film of 40 nm thickness and dried at 140 °C for 10 min in air. A solution containing a mixture of PCDTBT:PC70BM (1:4) in a mixed solvent (dichlorobenzene:chlorobenzene = 3:1) with a concentration of 7 mg mL−1 was spin-cast on top of MoOx films with various thicknesses by controlling the spin-casting rate. The thickness of the active layer and MoOx films was measured with a profilometer. Then, the BHJ films were heated at 70 °C for 10 min. After transferring to air, a thin TiOx layer was spin-cast on top of active layer, followed by baking at 80 °C for 10 min in air. Finally, the cathode (Al, ≈100 nm) was deposited through a shadow mask by thermal evaporation in a vacuum of about 3 × 10−6 Torr. The active area of device was 19.60 mm2. During the measurement, an aperture with the area of 12.38 mm2 was used. Current density–voltage (J–V) characteristics were measured using a Keithley 236 Source Measure Unit. Solar cell performance used an Air Mass 1.5 Global (AM 1.5 G) solar simulator with an irradiation intensity of 100 mW cm−2. The spectral mismatch factor was calculated by comparison of the solar simulator spectrum and the AM 1.5 spectrum at room temperature. Thin Film Characterization: Optical constants of pristine materials were obtained via variable angle spectroscopic ellipsometry (JA Woollam). Multiple samples for MoOx, PEDOT:PSS, and PCDTBT:PC70BM were analyzed in order to minimize correlation between thickness and optical functions. Thickness was independently measured via Veeco Dektak surface profilometry, and ellipsometric measurement of the transmission spectra of multiple sample thicknesses for each material additionally decoupled thickness from optical properties. Data were collected over a wavelength range (350–1000 nm) that fully described the optical constants of the materials, and at multiple angles of analysis (50–80°, 5° step) to probe the extraordinary index of refraction via the p-polarization of the electric field. Transmission spectroscopy at normal incidence probed the optical constants in the plane of the substrate and determined the complex index of refraction and the extinction coefficient. Scattering of light at the rough MoOx surfaces was expected to be negligible due to the good index matching between the MoOx and active layer. The transmission measurements were recorded at room temperature with an Agilent 8453 spectrophotometer. AFM imaging was carried out in air using an Asylum Research MFP-3D AFM. The XPS and UPS measurements were performed in a Kratos Ultra Spectrometer (base pressure of 1 × 10−9 Torr) using monochromatized Al Kα X-ray photons (hv = 1486.6 eV for XPS) and a HeI (21.2 eV for UPS) discharge lamp. MoOx films were thermally deposited on top of ITO substrates. The pass energy and step size were 40 eV and 0.05 eV for XPS and 10 eV and 0.025 eV for UPS. XPS data, curve fitting, and linear background subtraction used CASA XPS (version 2.3) software. For UPS, a sample bias of –9 V was used in order to separate the sample and the secondary edge for the analyzer. All samples were kept inside a high-vacuum chamber overnight to remove the solvent. Supporting Information is available from the Wiley Online Library or from the author. This research was supported by the US Army General Technical Services (LLC/GTS-S-09–1-196) and by the Air Force Office of Scientific Research (AFOSR FA9550–08-1–024, Charles Lee Program Officer). S.C. is supported as part of the Center for Energy Efficient Materials, an Energy Frontier Research Center, funded by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences under Award No. DOE/DE-SC0001009. The authors thank Dr. David Waller (Konarka Technologies) for supplying the PCDTBT material. Detailed facts of importance to specialist readers are published as ”Supporting Information”. Such documents are peer-reviewed, but not copy-edited or typeset. They are made available as submitted by the authors. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.