Plastic Solar Cells

Plastic Solar Cells

(Parte 1 de 3)

Plastic Solar Cells** By Christoph J. Brabec, N. Serdar Sariciftci,* and Jan C. Hummelen

1. Prologue

It is intriguing to think of photovoltaic (PV) elements based on thin plastic films. The flexibility offered through the chemical tailoring of desired properties, as well as the cheap technology already well developed for all kinds of plastic thin film applications would make such an approach a sure hit. The mechanical flexibility of plastic materials is welcome for all PVapplications onto curved surfaces for architectural integration. By casting semi-transparent plastic PV thin films between insulating window glass, large unused areas (the windows) can be employed for power generation in addition to the limited roof areas of crowded cities. Even the color of such PVelements can be varied by sacrificing some parts of the visible solar spectrum.

An encouraging breakthrough in realizing higher efficiencies has been achieved by mixing electron–donor-type polymers with suitable electron acceptors. The photophysics of conjugated polymer/fullerene solid composites has been particularly well investigated over the last eight years. An understanding of the photophysics in detail has allowed the realization of prototype PV devices with solar power conversion efficiencies of around 3%, and this has in turn triggered enhanced emphasis from several groups worldwide, pursuing this research with increasing supportfrom industry as well as publicfundingagencies.

On the other hand there is a common problem for all the applications of conjugated polymers: stability. Even though the expectation on the lifetimes of electronic devices is shrinking due to very short life/fashion cycles of such applications and even though industry may be interested in the cost of an item rather than in a very long durability of it, a shelf lifetime of several years as well as an operational lifetime of tens of thousands of hours are requested for all durable applications. Conjugated polymers have to be protected from air and humidity to achieve such lifetimes. These protection methods are being developed for light emitting diodes (LEDs) as well as PV elements. Their effectiveness is going to be crucial for the success of any applications using these materials. Recent reports from LED research indicate that the stability problem has been sufficiently overcome in order to enter into large scale applications, which in turn is a good sign for plastic solar cells.

It is the purpose of this review to give a state of the art report on plastic solar cells based on conjugated polymers as well as to give a basic introduction to the underlying photophysics.

2. Device Architectures 2.1. Single Layer Diodes

The simplest and most widely used organic semiconductor deviceisametal–insulator–metal(MIM)tunneldiodewithmetal electrodes ofasymmetricalworkfunction(Fig. 1).Under forward bias, holes from the high workfunction metal and electrons from the low workfunction metal are injected into a thin film of a single-component organic semiconductor. Because of the asymmetry of the workfunction of the cathode and the anode, forward bias currents for a single carrier type material are orders of magnitude larger than reverse bias currents at low voltages. The rectifying diode characteristics can be accompanied by radiative

Linzer Institut für organische Solarzellen (LIOS) Physikalische Chemie, Johannes Kepler Universität Linz Altenbergerstr. 69, A-4040 Linz (Austria) E-mail:

Prof. J. C. Hummelen Organic and Molecular Inorganic Chemistry, University of Groningen Nijenborgh 4, NL-9747 AG Groningen (The Netherlands)

[**] The authors gratefully acknowledge the co-workers at the University of

Linz and at the University of Groningen. as well as our numerous international collaboration partners. Financial supports of both the Universities, European Commission (DGXII, JOULE I), Quantum Solar Energy Linz GesmbH (Austria), Christian Doppler Society, the Magistrat Linz, the Land Oberösterreich (ETP), Austrian Foundation for Scientific Research (FWF P-12680CHE) as well as Netherlands Agency for Energy and Environment (NOVEM) and EET (The Dutch R&D Programme on Energy, Environment and Technology) are gratefully acknowledged.

Recent developments in conjugated-polymer-based photovoltaic elements are reviewed. The photophysics of such photoactive devices is based on the photo-induced charge transfer from donor-type semiconducting conjugated polymers to acceptor-type conjugated polymers or acceptor molecules such as Buckminsterfullerene, C60. This photo-induced charge transfer is reversible, ultrafast (within 100 fs) with a quantum efficiency approaching unity, and the charge-separated state is metastable (up to milliseconds at 80 K). Being similar to the first steps in natural photosynthesis, this photo-induced electron transfer leads to a number of potentially interesting applications, which include sensitization of the photoconductivity and photovoltaic phenomena. Examples of photovoltaic architectures are presented and their potential in terrestrial solar energy conversion discussed. Recent progress in the realization of improved photovoltaic elements with 3% power conversion efficiency is reported.

recombination channels of the injected electrons and holes within the molecular solid; the result is an LED.[1–5] If photo-induced free charge carrier generation is allowed at the same time, the device emits light under forward bias and exhibits a significant photocurrent under a reverse bias field (dual function).[6]

In order to use the devices for photodetection under reverse bias, the potential difference between the electrodes has to be high enough to overcome the Coulomb attraction for the photogenerated excitons. Otherwise the absorbed photons will mainly create excitons which decay geminately (either radiatively, with photoluminescence, or non-radiatively). Thus the photocurrent efficiency of such devices will be limited. In the PV mode, where no external voltage is applied and short circuit conditions exist, the potential difference available in the MIM device is caused by the difference between the workfunctions of the metal electrodes. In most of the cases (e.g., indium tin oxide, ITO, and Al) the potential difference due to this workfunction difference is not high enough to give efficient photo-induced charge generation, limiting the operation of the

Niyazi Serdar Sariciftci received his masters degree in Experimental Physics and a doctorate degree in Semiconductor Physics under the mentorship of Prof. H. Kuzmany at the University of Vienna, Austria, in 1986 and 1989, respectively. After a post doctoral period at Stuttgart University, Germany, with Prof. M. Mehring he joined the Institute for Polymers and Organic Solids at the University of California, Santa Barbara, with Prof. Alan Heeger. He was appointed Chair and Professor in Physical Chemistry at the Johannes Kepler University of Linz, Austria, in 1996. He is the Founding Director of the Christian Doppler Laboratory for Plastic Solar Cells and of the newly established Linz Institute for Organic Solar Cells (LIOS). His main research activities are on organic semiconductor physics and chemistry.

Christoph Brabec received his masters degree in Theoretical Physics in 1992 followed by a Ph.D. in Physical Chemistry under the mentorship of Prof. Hermann Janeschitz-Kriegl at the Johannes Kepler University of Linz, Austria, in 1995. After a post-doctoral stay at the Institute for Polymers & Organic Solids at the University of California, Santa Barbara, with Prof. Alan Heeger he joined the Christian Doppler Laboratory for Plastic Solar Cells at the Johannes Kepler University of Linz in 1998. His main research activities are in the photophysics and photochemistry of conjugated polymeric semiconductors and the development of plastic solar cells.

Jan C. (Kees) Hummelen received his masters degree in Chemistry and a doctorate degree in Science under the mentorship of Prof. H. Wynberg at the University of Groningen, The Netherlands, in 1979 and 1985, respectively. He continued working on the subject of his Ph.D. research until 1989. After four years of playing jazz (piano) and doing art video production in The Netherlands, he spent two years as a post-doctoral fellow with Prof. Fred Wudl at the Institute for Polymers and Organic Solids at University of California, Santa Barbara. After spending six months with Prof E. W. Meijer at the Eindhoven Technical University, he returned to the University of Groningen where he was appointed Universitair Hoofdocent in 1998, and full Professor in Chemistry in 2000. Over the last seven years, his main research activities have been in fullerene chemistry and the development of plastic photovoltaic technology.

Fig. 1. Charge generation process in a single layer conjugated polymer device under short circuit conditions in the MIM model. VB valence band, CB conduc- tion band, Eg bandgap, P ,P positive and negative polarons.

C. J. Brabec et al./Plastic Solar Cells

PV cells. Some improvement was reported for photodiodes utilizing a Schottky-type junction formed between the conjugated polymer and one of the metal electrodes (Fig. 2); however, the problem of inefficient charge generation in conjugated polymers was not overcome by this approach.[7–9]

To overcome this limitation of the photo-induced charge carrier generation, a donor/acceptor (dual molecule) approach has been suggested.[10–13] In general, in such devices, the photocarrier generation is enhanced by using a second, charge generation sensitizing component. For example, for a device consisting of a composite thinfilmwitha conjugated polymer/fullerenemixture, the efficiency of photogeneration of charges is near 100%. In such a single composite photoactive film a “bulk heterojunction” isformed between the electron donorsand acceptors(Fig. 3). An extensive discussion of this concept will be in Section 3 below.

2.2. Heterojunction Diodes

Considering the energy band diagram of a bilayer in Figure 4, the heterojunction formed between (for example) a conjugated polymer and C60 should have rectifying current–voltage characteristics even using the same metal contact on both sides (analo- gous to a p–n junction). One bias direction of such a device (electron injection on the semiconducting polymer side or hole injection on C60) is energetically unfavorable. This polarity of the device results in very low current densities. On the other hand, electron injection onto C60 or hole injection into the semiconducting polymer is energetically favorable. This polarity of the device results in relatively high current densities. Thus, devices of organic semiconductors using two layers with different electronic band structures as illustrated in Figure 4 have rectifying diode characteristics. A photophysical interaction between the two molecular units (photo-induced electron transfer) happens at the interface and gives rise to photocurrents as well as a PVeffect. In that sense the essential difference between the linear heterojunc- tion of two organic thin films displayed in Figure 4 and the “bulk heterojunction” displayed in Figure 3 is the effective interaction area between the donor and acceptor components: in the linear heterjunction device it is the geometrical interface, in bulk heterojunction it is the entire volume of the composite layer. This results in an enhancement of short circuit photocurrent for several orders of magnitude making the bulk heterojunction approach quite attractive.[1,12]

3. Conjugated Polymers as Photoexcited Donors

Many conjugated polymers (see examples in Table 1) in their undoped, semiconducting state are electron donors upon photoexcitation (electrons promoted to the antibonding p* band). The idea of using this property in conjunction with a molecular electron acceptor to achieve long lived charge separation was based on the stability of the photo-induced nonlinear excitations (such as polarons) on the conjugated polymer backbone. Once the photoexcited electron is transferred to an acceptor unit, the resulting cation radical (positive polaron) species on the conjugated polymer backbone is known to be highly delocalized, mobile, and stable as shown in electrochemical and/or chemical oxidative doping studies. Analogous to the chemical doping process, we will describe the photo-induced electron transfer from the conjugated polymer donor onto an acceptor moiety as “photodoping”.

Independently, Sariciftci et al. and Yoshino et al. reported studies on the photophysics of mixtures of conjugated polymers with fullerenes.[10,14–2] The observations clearly evidenced an ultrafast, reversible, metastable photo-induced electron transfer from conjugated polymers onto Buckminsterfullerene in solid films. A schematic description of this phenomenon is displayed in Figure 3 and Figure 4. Using this molecular effect at the interface of bilayers consisting of a semiconducting polymer (poly((2- methoxy-5-(2¢-ethylhexoxy)-p-phenylene) vinylene), hereafter referred to as MEH–PPV), and C60 films, diodes with rectification ratios on the order of 104 and a PV effect[1,23] were made.

Significant improvement of the relatively low charge collection efficiency of the donor/acceptor (D/A) bilayer was achieved by using phase separated composite materials, processed through control of the morphology of the phase separation into an interpenetratingnetwork(“bulkheterojunction”).Thepowerconversion efficiency of solar cells made from MEH–PPV/methanofullerene composites was subsequently increased dramatically.[12] In parallel, the groups of Heeger in Santa Barbara and Friend in Cambridge developed an approach using acceptor-type conjugated polymers in an interpenetrating polymer–polymer composite with MEH–PPV, yielding polymeric PV devices with efficiencies comparable to fullerene mixed devices.[13,24]

3.1. Ultrafast Photo-induced Electron Transfer from Conjugated Polymers onto C60

The linear optical absorption spectrum of a MEH–PPV/C60 film is a simple superposition of the two components without any indication of states below the p–p* gap of the conducting

Fig. 2. Energy diagram of a metal1/semiconductor/metal2 Schottky barrier under open circuitconditions, whenthemetals have different workfunctions(u workfunc- tion,vs electron affinity; IP ionization potential; Eg bandgap, W depletion width).

C. J. Brabec et al./Plastic Solar Cells

polymer as might arise from interaction between the two materials in the ground state.

The strong luminescence of MEH–PPV was, however, quenched by a factor in excess of 103.[10] The luminescence decay time was reduced from so = 550 ps to srad << 60 ps (the instrumental resolution) indicating the existence of a rapid quenching process; e.g., sub-picosecond electron transfer.[14] The strong quenching of the luminescence of another conjugated polymer, P3OT, reported by Morita et al.[19] was also consistent with efficient photo-induced electron transfer. Thus, the quenching of luminescence was observed in a number of conjugated polymers in composites with fullerenes, indicating this to be a general phenomenon for the non-degenerate ground state conjugated polymers.[25] This confirmed that photo-induced electron transfer occurs in a timescale sufficiently fast to quench the radiative relaxation of the conjugated polymer excited state.

Photo-induced absorption detected magnetic resonance (PIADMR) experiments were performed in conjugated poly- mer/C60 composites giving evidence for a complete quenching of the MEH–PPV triplet–triplet absorption signal at 1.35 eV.

Instead a strong spin = 1/2 signal dominated the PIADMR spectrum, indicating charged polarons as photoexcitations on the polymer donor.[18] This confirmed that the photo-induced

Fig. 3. Formation of a bulk heterojunction and subsequent photo-induced electron transfer inside such a composite formed from the interpenetrating donor/acceptor network plotted with the device structure for such a kind of junction (a). The diagrams with the energy levels of a

MDMO–PPV/C60 bulk heterojunction system (as an example) under flat band conditions (b) and under short circuit conditions (c) do not take into account possible interfacial layers at the metal/ semiconductor interface.

C. J. Brabec et al./Plastic Solar Cells

electron transfer occurs on a time scale sufficiently fast to quench the intersystem crossing to the triplet state.

Using sub-ps photo-induced absorption (PIA) studies[15,26,27] ultrafast (<1 picosecond) formation of the polarons was demonstrated.

Very recently the forward transfer of the photoexcited electrons from conjugated polymer donors onto 1-(3-methoxycar- bonyl)-propyl-1-1-phenyl-(6,6)C61 (PCBM) acceptors was resolved by pump-probe experiments with an unprecented time resolution of 10 fs.[28] In these studies, the relaxation of the photo-induced excitations on the polymer chain switched from radiative (intense stimulated emission in pristine polymer, DT > 0) to non-radiative (electron transfer in composites, PIA, DT < 0) upon mixing PCBM into the polymer matrix. This experimental setup allowed to time-resolve the photo-induced electron transfer time with ~40 fs after an initial relaxation (Kasha relaxation). As a consequence of the resonant excitation of the conjugated polymer by a sub-10 fs laser pulse, the phonon modes which are strongly coupled to the p-electron excitations were directly observed as coherent oscillations

Fig. 4. Schematic diagram of a bilayer and subsequent photo-induced electron transfer at the interface of the two layers with the device structure for such a kind of junction (a). The diagrams with the energy levels of a MDMO–PPV/C60 system (as an example) under flat band conditions

(b) and under short circuit conditions (c) do not take into account possible interfacial layers at the metal/semiconductor interface.

C. J. Brabec et al./Plastic Solar Cells within the pump probe experiment. By Fourier transforming the oscillatory decay of the time dependent signal the resonant/non-resonant Raman spectrum of the conjugated polymer was reproduced. Furthermore, by adding PCBM to the conjugated polymer matrix these vibrational oscillations were quenched nearly completely indicating a rapid depopulation of the excited state on the conjugated polymer due to the electron transfer. These results demonstrated a direct competition of photo-induced electron transfer with vibrational oscillationson the conjugated polymer donor indicating a photo-induced elec- tron transfer rate of >1013 s–1 (sct < 100 fs). As such this photo-induced electron transfer is indeed ultra- fast, resulting in a quantum yield of photo-induced charge generation of 100%. The time-resolved transient photocurrent (PC) of MEH–

PPV/C60 composites with different C60 content[29] shows an increase of initial photocurrent by an order of magnitude upon admixture of 1% of C60. This increase of the photocarrier generation efficiency was accompanied by the successive increase in lifetime of the photocarriers upon adding more and more C60. Thus, the ultrafast photo-induced electron transfer from the semiconducting polymer to C60 not only enhances the charge carrier generation in the host polymer but also serves to prevent recombination by separating the charges and stabilizing them.[29]

Definitive evidence of charge transfer and long-lived charge separation was obtained from light induced electron spin resonance (LESR) experiments.[10] Figure 5 shows the integrated ESR signal upon illuminating the MDMO–PPV/PCBM (where MDMOstandsforpoly(2-methoxy-5-(3,7-dimethyloctyloxy)-1,4-

phenylene vinylene)) composites with light of hm = Ep–p* where

Ep–p* is the energy gap of the conjugated polymer (donor). Two photo-inducedESR signalscan beresolved;one at g =2 .0026a nd the other at g = 1.9997. The higher g-value line is assigned to the conjugatedpolymercation(polaron)andthelowerg-valuelineto PCBM– anion.[30] The saturation behavior of the LESR of conjugatedpolymer cation signal indicates a different relaxationmechanism compared to the fullerene radical anion signal which is not saturating at all.[31] We can safelyconcludefrom thesestudies that the photo-induced radicalsinthese polymer/fullerene composites areindependentfromeachotheranddissociatedcompletely.

4. Towards Improved Efficiency Organic Photovoltaic Cells

(Parte 1 de 3)