Operation of a reversed pentacene-fullerene discrete heterojunction photovoltaic device

D. M. Nanditha, M. Dissanayake, Ross A. Hatton, Richard J. Curry, S. R. P. Silva
2007 Applied Physics Letters  
The photoresponse of reversed bilayer organic photovoltaic device based on pentacene and C 60 is examined, and the mechanism of photocurrent generation is shown to be different to that in conventional heterojunction devices, with free charge carriers generated at the electrode-organic interfaces rather than the organic heterojunction. This hypothesis is tested with silver nanoclusters incorporated at the organic heterojunction to quench excitons and facilitate recombination of free charge
more » ... rs, which shows a predicted increase in J sc . The large V oc in this reversed cell structure is also rationalized in the context of the model proposed. The potential for low cost, flexible optoelectronic devices with large area capability has motivated significant research effort into the development of small conjugated organic molecule and polymer based light-emitting diodes ͑OLEDs͒ and photovoltaic devices ͑OPVs͒. 1 While the development of OLEDs has led to commercial products, OPVs have yet to show the performance required to enable their commercialization. OPVs are based on junctions between electron accepting and electron donating molecules, at which photoinduced excitons formed in both materials are dissociated to form free charge carriers. OPVs can be broadly categorized into two groups according to whether the photoactive component comprises two discrete layers of donor and acceptor molecules ͑bilayer͒ or a complex interpenetrating network ͑bulk heterojunction͒. To date, both bilayer and bulk-heterojunction OPVs have achieved power conversion efficiencies ͑ p ͒ of 4%-5% under 1 sun simulated solar illumination 2-5 which is considered to be half that required for market entry today. Furthermore, neither approach has a clear lead in the race to achieve commercial viability as yet. To improve the performance of OPVs, the open circuit voltage ͑V oc ͒, short circuit current density ͑J sc ͒, and fill factor must all be optimized. Hence, it is essential to understand how these parameters relate to the fundamental physical processes of light absorption, exciton dissociation, charge carrier transport, and extraction to the external circuit. While the factors affecting V oc are still the subject of debate, 6-9 it is generally accepted that the maximum attainable V oc in heterojunction OPV is determined by the difference in potential between electrons in the lowest unoccupied molecular orbital ͑LUMO͒ of the acceptor material and the highest occupied molecular orbital ͑HOMO͒ of the donor material. Furthermore, the maximum V oc is only achieved when photogenerated free charge carriers are extracted to the external circuit by the electrodes having Fermi levels aligned with these molecular orbital energies. Employing electrodes having a difference in work function greater than the donor HOMOacceptor LUMO offset in an attempt to increase the built-in electric field is counterproductive, since free carriers have to surmount barriers to be extracted to the external circuit. The complexity of many metal-organic interfaces makes achieving optimal energy level alignment and predicting V oc based on the work function of the pristine electrodes problematic. 10 This letter reports the photoresponse of reversed discrete heterojunction OPV based on pentacene/C 60 heterojunctions. To investigate the wider applicability of this study, reverse cell structures employing the soluble C 60 derivative ͓6,6͔phenyl-C61 butyric acid methyl ester ͑PCBM͒ in place of C 60 was also investigated and found to exhibit similar characteristics. Consequently, in the interests of conciseness, only the former is present herein with experimental results for the later included in the accompanying supplementary information. 11 Twice sublimed pentacene ͑99.99%͒ was purchased from H.W. Sands. C 60 ͑99.99%͒ and PCBM were purchased from American Dye Source and Solenne, respectively. All organic materials were used as received. Devices were fabricated on indium tin oxide ͑ITO͒ coated glass cleaned using a three stage ultrasonic bath treatment in toluene, an aqueous surfactant solution, and acetone. Immediately prior to use, the ITO substrates were suspended in the vapor of refluxing acetone followed by microwave oxygen plasma treatment. Discrete heterojunction OPVs were fabricated by subliming pentacene and C 60 under high vacuum at a deposition rate of ϳ0.1 nm s −1 onto precleaned ITO glass substrates. Metal counterelectrodes to a 50 nm thickness were deposited at ϳ0.5 nm s −1 using a shadow mask without breaking vacuum. Current-voltage ͑J-V͒ measurements were performed using 100 mW cm −2 AM1.5 simulated solar irradiation. The J-V characteristics of ITO/pentacene ͑45 nm͒ /C 60 ͑50 nm͒ /Al ͑normal͒ and the ITO/ C 60 ͑50 nm͒/pentacene ͑50 nm͒ /Al ͑reversed͒ devices are shown in Fig. 1. Figures 2͑a͒ and 2͑b͒ show the flatband energy level diagrams for both device configurations. The work functions of the clean ITO glass ͑⌽ ITO ͒, measured directly using ultraviolet photoelectron spectroscopy, and the aluminium counter electrode ͑⌽ Al ͒ were 4.4 ͑Ref. 12͒ and 4.3 eV, respectively. Notably, these schematic energy level diagrams do not account for the likely formation of abrupt vacuum level shifts, which often occur at metal-organic interfaces and which can drastically modify interfacial energy level alignment ͓Fig. 2͑e͔͒. 13 Remarkably, the J-V characteristics of both normal and reverse a͒ Electronic mail: m.dissanayake@surrey.ac.uk APPLIED PHYSICS LETTERS 90, 113505 ͑2007͒
doi:10.1063/1.2713345 fatcat:m7h4yw7syfdh3ne4zdmc3nneyi