Portal de Periódicos da CAPES

Triple Junction Polymer Solar Cells for Photoelectrochemical Water Splitting

2013; Volume: 25; Issue: 21 Linguagem: Inglês

10.1002/adma.201300439

1521-4095

Serkan Esiner, Harm van Eersel, Martijn M. Wienk, René A. J. Janssen,

Quantum Dots Synthesis And Properties

A triple junction polymer solar cell in a novel 1 + 2 type configuration provides photoelectrochemical water splitting in its maximum power point at V ≈ 1.70 V with an estimated solar to hydrogen energy conversion efficiency of 3.1%. The triple junction cell consists of a wide bandgap front cell and two identical small bandgap middle and back cells. Large-scale photovoltaic energy conversion is one of the most promising technologies to meet the future demand in renewable energy, but ultimately efficient conversion of solar-to- chemical energy is required to sustain our global energy demand as solar electricity supply and demand are intermittent. Because capturing solar energy in chemical bonds of molecular fuels is most effective in terms of energy density, solar fuels are attracting considerable attention recently. Successful construction of a direct artificial system for efficient solar fuel generation is possibly the most important research challenge in coming years,1 and the first firm ideas are emerging on how this can be achieved.2 Solar energy driven water splitting has initially been advanced by Turner et al.3 Subsequently, Licht et al.4 have shown that it is possible to perform photoelectrochemical water splitting at a solar-to-fuel conversion efficiency of 18.3% with a AlGaAs/Si tandem solar cell using RuO2 as oxygen evolving catalyst and Pt-black as hydrogen evolving catalyst. Despite high efficiencies, large-scale terrestrial application of such devices is hampered by the prohibitive high cost of the single-crystal semiconductors. More recently, Nocera et al. accomplished solar water splitting using silicon-based semiconductors triple junction cells and earth-abundant hydrogen and oxygen evolving catalysts achieving a solar-to-fuel efficiency of 2.5% with a wireless device configuration, and 4.7% with a wired configuration.5 Polymer solar cells are one of the candidates to contribute to large scale photovoltaic energy conversion due to their potentially cheap, high volume, solution-based manufacturing. The power conversion efficiencies (η) of polymer solar cells are rapidly approaching the 10% threshold.6 In recent years tandem configurations have been explored to further increase the efficiency of polymer solar cells. In a tandem configuration thermalization losses are reduced by the absorption of high-energy photons in a wide bandgap photoactive layer and transmission losses are lowered by the absorption of low-energy photons in a small bandgap photoactive layer. Tandem polymer solar cells that reach power conversion efficiencies of 9.5% have recently been published.7, 8 We considered it of interest to investigate the possibility of using semiconducting polymer:fullerene heterojunctions as the absorber in a photoelectrochemical water splitting device. For constructing such a device it is important to consider that the standard potential for water splitting (H2O → H2+ O2) is E0H2O = 1.23 V. In practice, however, water splitting occurs at a potential (VH2O) that is higher than E0H2O due to overpotential losses occurring at the electrodes. Depending on the type of the electrodes, the electrolyte, and the current density, the overpotentials vary for both H2 and O2 evolution electrodes and the electrolysis potential lies typically in the range VH2O = 1.4–1.9 V.9, 10 Even the best polymer tandem solar cells do not reach this potential in their maximum power point7, 8 and hence would be unable to provide significant current densities in photoelectrochemical water splitting. To overcome this problem it is possible to use a triple junction configuration. Examples of vacuum deposited and solution processed triple junction organic solar cells have been published,11-16 but so far these use three active layers that are identical or span the same spectral regions and, hence, are not exploiting the true advantage of a multiple junctions in reaching higher efficiency. On the other hand, presently a balanced set of three complementary organic semiconductors absorbers for a true triple junction cell is not yet available. In this paper we present a fully solution processed triple junction polymer solar cell in a novel 1 + 2 type configuration (Figure 1) with an optimized efficiency of 5.3% and an open-circuit voltage of Voc = 2.33 V and demonstrate that the cell can be used for photoelectrochemical water splitting close to its maximum power point at V ≈ 1.70 V. The triple junction cell employs one wide bandgap bulk heterojunction front cell and two identical small bandgap bulk heterojunctions in the middle and back cells. The rationale of this design is that in polymer tandem cells the current generating capacity of the small bandgap cell is often not fully exploited, because the wide bandgap cell that absorbs a small part of the spectrum limits the current.17 In such a case, it can be advantageous to combine the wide bandgap active layer with two identical small bandgap sub cells. Because the two small bandgap cells compete for the same photons, the generated current is reduced and all three layers of the triple junction solar cell will have more balanced current generation than in the tandem device. Importantly, due to the additional junction, the voltage and efficiency of such a 1 + 2 type triple junction solar cell will be higher than that of the corresponding tandem device. Structures of the wide bandgap (PF10TBT, Eg = 1.95 eV) and small bandgap (PDPPTPT, Eg = 1.53 eV) polymers (top) and layouts of the single, tandem, and 1 + 2 type triple junction solar cells (bottom). For our 1 + 2 triple junction we chose donor materials that provide relatively high Voc with fullerene acceptors. The front cell is a blend of PF10TBT (poly[2,7-(9,9-didecylfluorene)-alt-5,5-(4′,7′-di-2-thienyl-2′,1′,3′-benzothiadiazole)], Figure 1)17 as donor with PCBM ([6,6]-phenyl-C61 butyric acid methyl ester) as acceptor, while the middle and back cells consist of a blend of PDPPTPT (poly[{2,5-bis(2-hexyldecyl)-2,3,5,6-tetrahydro-3,6-dioxopyrrolo[3,4-c]pyrrole-1,4-diyl}-alt-{[2,2′-(1,4-phenylene)bisthiophene]-5,5′-diyl}], Figure 1)18 and PCBM. The extinction coefficients of the PF10TBT:PCBM and PDPPTPT:PCBM blends (Figure 2) demonstrate the complementary absorption spectra of the two active layers with optical bandgaps at Eg = 1.95 and 1.53 eV. In optimized single junction cells they provide efficiencies of η = 3.7% and η = 4.5%, respectively with relatively high open circuit voltage of Voc = 1.04 and 0.78 V (Table 1). However, the short circuit current densities, Jsc = 6.1 mA cm−2 and 9.3 mA cm−2 (Table 1), of the optimized devices differ significantly. This large difference gives rise to unmatched currents of sub cells in a PF10TBT:PCBM/PDPPTPT:PCBM tandem cell, even in the optimized conditions. The wide bandgap sub cell is current limiting and the current generating potential of the back cell cannot be effectively used. Thus, the approach of combining the wide bandgap front cell with two identical narrow bandgap cells is expected to provide an improved device performance when compared to the corresponding tandem solar cell. Extinction coefficient and refractive index of the wide and small bandgap active layers. Towards an efficient triple junction solar cell, we first developed an intermediate contact that is compatible with the stack of layers inside the device and that is easy to process. ZnO nanoparticles and PEDOT:PSS are commonly used as electron transport (ET) and hole transport (HT) layers in solution processed polymer solar cells.12 To prevent deteriorating the underlying ZnO layer by the commonly used acidic PEDOT:PSS dispersion (pH ∼ 1.8), a pH-neutral PEDOT:PSS must be used.12 However, pH-neutralization of PEDOT:PSS lowers the work function from 5.05 to 4.65 eV (pH neutral PEDOT),19 and introduces voltage losses when used as an HT layer in a polymer solar cell with a polymer of deep (more negative) HOMO level. In the envisioned tandem and triple junction configurations, the pH-neutral PEDOT layers connect to the PDPPTPT: PCBM layer for hole collection and the effect of pH-neutral PEDOT on the high Voc = 0.78 V of the single junction PDPPTPT: PCBM device must be addressed. Figure 3 shows that by replacing the acidic PEDOT layer with pH-neutral PEDOT in single junction PDPPTPT:PCBM devices the Voc decreases by 0.20 V to 0.58 V. This loss, however, can largely be recovered when a solution of Nafion is spin-coated onto the pH-neutral PEDOT layer before applying the photoactive blend.20, 21 With a thin Nafion layer, Voc rises to 0.73 V without affecting the fill factor or the current (Figure 3a) leading to a minor voltage loss of 0.05 V compared to acidic PEDOT. We note, however, that the performance improvement obtained with Nafion varied somewhat in different experiments and that typically losses were in the range of 0.05–0.10 V. (a) Effect of pH-neutral PEDOT and Nafion on the J-V curves of single junction ITO/(acidic or pH neutral)PEDOT/(none or Nafion)/PDPPTPT:PCBM/LiF/Al solar cells. (b) Tandem solar cells using two photoactive layers (PALs) consisting of PF10TBT:PCBM or PDPPTPT:PCBM with an ITO/acidic PEDOT/PAL/ZnO/pH neutral PEDOT/Nafion/PAL/LiF/Al cell stack. (c) External quantum efficiency of the PF10TBT:PCBM/PDPPTPT:PCBM tandem cell shown in panel (b). As a next step, tandem cells were made with a Nafion modified pH-neutral PEDOT intermediate contact. We have previously shown that it is possible to predict the behavior of multiple junction devices under solar illumination using optical modeling in a transfer matrix approach based on extinction coefficient and refractive index as function of wavelength for all active layers and using the experimental J-V curves of the single junction devices at various thicknesses.17 The J-V curve of the optimized tandem cell with a wide bandgap PF10TBT:PCBM front cell and small bandgap PDPPTPT:PCBM back cell reaches η = 4.6% and shows a Voc of 1.72 V and (Figure 3b, Table 1), which represents a loss of 0.10 V compared to the sum of the optimized single junctions (Voc = 1.04 + 0.78 = 1.82 V). Figure 3c shows external quantum efficiency (EQE) measurements of the individual sub cells inside the PF10TBT:PCBM/PDPPTPT:PCBM tandem device, measured using appropriate optical and electrical bias.22 By integrating the EQE with AM1.5G solar spectrum, the current density generating capacities of the two sub cells can be determined. In the optimized tandem cell the back cell can generate a short circuit current density of Jsc = 7.30 mA cm−2, which is significantly higher than the corresponding value of Jsc = 4.68 mA cm−2 that can be generated by the front cell. This difference supports the need for the 1 + 2 type triple junction solar cell structure. To test the possibility of making such 1 + 2 type triple junction, we also made PDPPTPT:PCBM/PDPPTPT:PCBM tandem solar cells, i.e. using two identical small bandgap active layers (Figure 3b). Using a Nafion modified pH-neutral PEDOT intermediate contact we find η = 4.2% and Voc = 1.51 V, close to the value of 1.56 V that would be expected based on the optimized single junction. The efficiency of the tandem of η = 4.2% is slightly less than for a single junction PDPPTPT:PCBM cell (η = 4.5%), mainly due to losses in FF and Voc. In both tandem cells the potential in the maximum power point (V = 1.33 V and V = 1.16 V) is not sufficient for photoelectrochemical water splitting (E0H2O = 1.23 V), especially when considering overpotential losses. To raise the potential in the maximum power point, triple junction devices with PF10TBT:PCBM as the front cell and two PDPPTPT:PCBM layers as the middle and back cells were made with the optimized active layer thicknesses obtained from the simulations. The triple junction cell shown in Figure 4 gave η = 5.3%, which is significantly higher than the two tandem devices (η = 4.6 and 4.2%) and the two optimized single junction devices (η = 3.7% and 4.5%, respectively). Compared to Jsc = 4.84–4.90 mA cm−2 of the two tandem cells, the Jsc = 4.42 mA cm−2 of the triple junction cell is only slightly lower, demonstrating that for this material combination the 1 + 2 configuration is really beneficial. With Voc = 2.33 V for this triple junction cell, there is a moderate loss compared to the maximum possible value of Voc = 2.60 V (= 1.04 + 2 ×0.78). This is due to the potential loss (2 × ∼0.10 V) at the intermediate contacts and the reduced number of photons that can be absorbed in each of the layers.23 As mentioned, the voltage loss at the intermediate contact varied in different experimental runs, resulting in maximum Voc = 2.50 V in favorable cases. Importantly however, and in contrast to the single junction or the tandem cells presented in this work, the triple junction cell provides V = 1.70 V in its maximum power point, sufficient for water splitting. Device structure of the triple junction (left). J–V curves of the optimized single, tandem, and triple junction solar cells (right). Using two platinum electrodes and a 1 M KOH electrolyte, electrochemical water splitting requires a potential of about 1.70 V. Figure 5 shows the I-V characteristics of a 1 + 2 triple junction polymer solar cell under white light illumination connected to two thin platinum electrodes in a 1 M KOH electrolyte. Under these conditions the photoelectrochemical evolution of hydrogen and oxygen can easily be observed (see inset). The part of the I-V curve of the triple junction cell between open-circuit and lowest potential where water splitting occurs can be measured by adjusting the area of the platinum electrodes that are in contact with the electrolyte and recording the potential and current. By lowering the area of the electrodes it is possible to effectively increase the resistance of the electrochemical cell, which enables measuring part of the I-V curve. Figure 5 shows that these electrochemical I–V measurements exactly match those of the same cell measured with a source-measurement unit. This experiment demonstrates the use of a triple junction polymer solar cell for water splitting and hydrogen generation. In a separate electrochemical experiment, we determined that the current to hydrogen conversion is better than 80%. This allows determining a lower estimate for the solar-to-hydrogen energy conversion efficiency as η = ηtriple × (E0H2O/VH2O) × 80% = 5.3% × (1.23/1.70) × 80% = 3.1%. Comparison of the I–V curves of the triple junction cell measured using a water electrolysis cell with different sized contacts and using a source-measurement unit. This particular triple junction cell had Voc = 2.50 V when measured under white light conditions close to AM1.5G. The inset shows the evolution of H2 and O2 during the photoelectrochemical experiment. In conclusion, a triple junction solar cell with one wide bandgap and two identical small bandgap active layers has been developed. A 5.3% efficient triple junction solar cell has been made with PF10TBT:PCBM as a front cell and PDPPTPT:PCBM as middle and back cells. The triple junction device showed a better efficiency than the single junction devices and the tandem device which also showed a high efficiency of 4.6%. In addition, a basic example of solar to fuel conversion has been illustrated exploiting the high voltage of the triple junction solar cell. A next step is combining better hydrogen and oxygen evolving catalysts to lower the water splitting voltage and making an integrated design of the photoelectrochemical cell. Device preparation: Solar cells were prepared by first spin-casting PEDOT:PSS (Clevios P VP AI 4083, H. C. Starck) onto pre-cleaned glass substrates with indium tin oxide (ITO) patterns (Naranjo Substrates). Then, the active layer was spin-cast. Single junction devices were completed by thermal evaporating a back contact of 1 nm LiF and 100 nm Al at 3 × 10−7 mbar. For tandem and triple junction devices, the intermediate contact consisting of ZnO nanoparticles (∼30 nm), pH-neutral PEDOT (∼15 nm), and Nafion (∼4 nm) was deposited sequentially by spin-casting. For optimal performance, ZnO nanoparticles were deposited inside a glove box with nitrogen environment while all other layers were spin-cast in air. No heat treatment was applied to any of the layers. The front cell was spin-cast from a warm solution of PF10TBT and PCBM (1:4 w/w) in chloroform containing 70 mg mL−1 o-DCB at 5 mg mL−1 polymer concentration. The middle and the back cells were spin-cast from a solution of PDPPTPT and PCBM (1:2 w/w) in chloroform containing 70 mg mL−1 o-DCB at 5 mg mL−1 polymer concentration. ZnO nanoparticles24, 25 of ∼5 nm diameter were spin-cast from a solution of 14.25 mg mL−1 ZnO in isopropanol (IPA). pH-neutral PEDOT was prepared by diluting 'Orgacon Neutral pH PEDOT' from Agfa with ultra-pure water at a 1:1 volume ratio. The solution was then filtered with a 5.0 μm Whatman Puradisc FP30 syringe filter. Nafion (Aldrich chemistry, perfluorinated ion exchange resin 5 wt.% in mixture of lower aliphatic alcohols and H2O, containing 45% water) was first diluted in ethanol with a 1:200 volume ratio. The resulting solution was spin-cast directly onto pH-neutral PEDOT for an optimized thickness of 2-5 nm. Characterization: A Keithley 2400 source-measurement unit was used to measure current density to voltage (J-V) characteristics of the devices. The illumination was carried out with ∼100 mW cm−2 white light from a tungsten-halogen lamp filtered by a Schott GG385 UV filter and a Hoya LB120 daylight filter. No mismatch correction was performed. The measurements were performed inside a glove box with a nitrogen atmosphere. The tandem and triple junction devices were exposed to UV illumination (with a Spectroline EN-160L/F 365 nm lamp from Spectronics Corporation) for about 10 min. to provide an ohmic contact between the ZnO and pH-neutral PEDOT layers before being measured. To prevent parasitical charge collection due to the high lateral conductivity of pH-neutral PEDOT, tandem and triple junction devices were measured with a mask of the same size with the device area, which is determined by the overlap of the ITO and Al electrodes. For single junction devices, more accurate short circuit currents under AM1.5G were calculated by convolution of the EQE measurements with the AM1.5G solar spectrum. EQE measurements were performed in a homebuilt set-up. Mechanically modulated (SR 540, Stanford Research) monochromatic (Oriel Cornerstone 130) light from a 50 W tungsten halogen lamp (Osram 64610) was used as a probe light together with continuous bias light from a solid state laser (B&W Tek Inc., λ = 532 nm, 30 mW and λ = 780 nm, 21 mW) through an aperture of 2 mm diameter. The intensity of the bias laser was adjusted using a variable neutral density filter. The response was recorded using a lock-in amplifier (Stanford Research Systems SR830), over a resistance of 50 Ω. For all the single junction devices and the PDPPTPT:PCBM sub cell, the measurement was carried out under representative illumination intensity (AM1.5G equivalent, provided by the 532 nm laser). For the PF10TBT:PCBM sub cell, the measured EQE was mathematically corrected for the intensity difference between the monochromatic light and AM1.5G. In order to maintain the short circuit conditions for the measured sub cell in a tandem device, the extra electrical bias was applied by the lock-in-amplifier during the EQE measurements. A calibrated silicon solar cell was used as reference. Devices were kept behind a quartz window in a nitrogen filled box during the measurements. Layer thicknesses were measured with 'Veeco Dektak 150' Surface Profiler. UV/vis absorption spectra of the active layers spin-cast on glass were measured with a Perkin-Elmer Lambda 900 spectrometer. Simulation: The details of the procedure have been described elsewhere.17 The optical modeling was done in SetFos version 2.1.439 (Fluxim AG, Switzerland) interfaced to a home built program using Python 2.6 scripts. The optical constants were mostly available from literature,17 and for PDPPTPT:PCBM determined from reflection/transmission measurements on active layers on quartz. Electrolysis experiments: Platinum electrodes used in the experiments were 5 cm long and 0.35 mm in diameter. The I–V curve of the solar cell was recorded by varying the depth of the electrodes inside the 1 M KOH electrolyte from 0 to 4 cm depth. For higher current generation, four triple junction solar cells with areas varying between 0.09 cm2 and 0.16 cm2 were connected in parallel. The solar cells were illuminated by an arbitrary white light source with a power density of less than 100 mW cm−2. We thank Jan Gilot for important contributions to initial phases of this work and Jorgen Sweelssen, Johan Bijleveld, and Koen Hendriks for providing the polymer semiconductors. This work is part of the research programme of the Foundation for Fundamental Research on Matter (FOM), which is part of the Netherlands Organisation for Scientific Research (NWO), the Solliance Organic Photovoltaics programme, and we acknowledge funding from the Ministry of Education, Culture, and Science (NWO Gravity program 024.001.035).