Organic photovoltaics

Abstract

Organic photovoltaics, the technology to convert sun light into electricity by employing thin films of organic semiconductors, has been the subject of active research over the past 20 years and has received increased interest in recent years by the industrial sector. This technology has the potential to spawn a new generation of low-cost, solar-powered products with thin and flexible form factors. Here, we introduce the energy and environmental science community to the basic concepts of organic photovoltaics and discuss some recent science and engineering results and future challenges.

---

Bernard Kippelen

Bernard Kippelen received his PhD in solid-state physics in 1990 from the University Louis Pasteur, Strasbourg, France. He was Chargé de Recherches at the CNRS. In 1994 he joined the Optical Sciences Center at The University of Arizona where he became an Assistant Professor in 1998 and an Associate Professor in 2001. Since 2003, he has been Professor of Electrical and Computer Engineering at the Georgia Institute of Technology, Atlanta, USA. He is a Fellow of the Optical Society of America, a Fellow of SPIE, and a Senior Member of IEEE.

Jean-Luc Brédas

Jean-Luc Brédas received his PhD in Chemistry from the University of Namur, Belgium, in 1979. In 1988, he was appointed Professor at the University of Mons-Hainaut, Belgium. He joined the University of Arizona in 1999 before moving in 2003 to the Georgia Institute of Technology where he is a Professor of Chemistry and Biochemistry. He is the recipient of the 1997 Francqui Prize, the 2000 Quinquennial Prize of the Belgian National Science Foundation, the 2001 Italgas Prize (shared with Richard Friend), and a member of the team laureate of the 2003 Descartes Prize of the European Union.

---

Broader context

Organic photovoltaics, the technology to convert sun light into electricity by employing thin films of organic semiconductors, has been the subject of active research over the past 20 years and has received increased interest in recent years by the industrial sector. This technology has the potential to spawn a new generation of low-cost, solar-powered products with thin and flexible form factors. This review introduces the energy and environmental science community to the basic concepts of organic photovoltaics and discusses some recent science and engineering results and future challenges.

The photovoltaic effect—the conversion of light into electrical power—can be traced back to Becquerel's 1839 pioneering studies in liquid electrolytes¹ and has since been studied in a wide range of materials. In the modern era, the tipping point that transformed photovoltaics into a technology to convert sun light into electricity was the 1954 report by Chapin et al.² of a silicon-based single p–n junction device with a solar power conversion efficiency of 6%. Recognized initially as a sustainable power source for geostationary communications satellites, photovoltaic cells have now gained in efficiency and found many applications in the consumer market.^3–5 More importantly, in recent years, they are emerging as a clean and sustainable source of energy and are expected to play a major role in meeting the global energy challenge.⁶

Solar technologies are currently dominated by wafer-size single-junction solar cells based on crystalline silicon that are assembled into large area modules. However, other semiconductor materials and devices are under active investigation in order to further reduce the cost of produced electricity by: increasing the power conversion efficiency, reducing the amount of absorbing material needed, and lowering the assembly cost of modules. Thin-film photovoltaic technologies,⁷ referred to as second-generation photovoltaics, are based on inorganic semiconductor materials that are more absorbing than crystalline silicon and can be processed directly onto large area substrates. Such semiconductors include amorphous silicon, II–VI semiconductors such as CdS or CdTe, and chalcogenides such as CuInSe₂ (CIS) or CuInGaSe₂ (CIGS).⁸ Despite the laboratory demonstration of cells with high efficiencies (19% for CIGS⁹ and 16% for CdTe¹⁰), the controlled manufacturing of second-generation cells remains a challenge and their commercial use is growing but not as widespread yet.

Simultaneously, over the past two decades, the science and engineering of organic semiconducting materials have advanced very rapidly, leading to the demonstration and optimization of a range of organics-based solid-state devices, including organic light-emitting diodes (OLEDs),¹¹ field-effect transistors,^12,13 photodiodes,¹⁴ and photovoltaic cells. Seeded in the 1960s by fundamental studies on the optical and electronic properties of model organic molecules such as acenes¹⁵—molecules based on up to five fused benzene rings—this area of research gained significant momentum in the late 1970s and in the 1980s when high-purity small organic molecules with tailored structure and properties were synthesized and processed at room temperature into thin films using physical vapor deposition techniques. Building on such advances, Tang¹⁶ developed single heterojunction organic photovoltaic cells and reported in 1986 a power conversion efficiency of about 1%. This result represented a major milestone and a significant improvement in efficiency over the first report of a device with similar geometry by Kearns and Calvin¹⁷ in 1958. The advent in the 1990's of high-purity conjugated polymers allowed the fabrication of organic photovoltaic cells with materials simply processed from solution.^18,19

The low-temperature processing of either organic small molecules from the vapor phase or polymers from solution confers organic semiconductors with a critical advantage over their inorganic counterparts, as the high-temperature processing requirements of the latter limit the range of substrates on which they can be deposited. Particularly attractive for organic semiconductors are flexible plastic substrates that can lead to applications and consumer products with lower cost, highly flexible form factors, and light weight. Furthermore, low-temperature processing cuts on energy use during manufacturing, further reducing the energy payback time which is defined as the operating life of a power-generating device needed to produce the amount of energy invested during manufacturing, installation and maintenance. These attributes, combined with the ability to tune the physical properties of organic (macro)molecules by fine tuning their chemical structure, constitute the main drivers boosting research and industrial interest in organic photovoltaics.

The organics-based approaches and those that do not rely on conventional single p–n junctions are often referred to as third-generation technologies. They include: (i) the dye-sensitized solar cells pioneered by Grätzel,^20,21 which are electrochemical cells that require an electrolyte; (ii) multijunction cells fabricated from group IV and III–V semiconductors;²² (iii) hybrid approaches in which inorganic quantum dots^23–25 are doped into a semiconducting polymer matrix or by combining nanostructured inorganic semiconductors such as TiO₂ with organic materials;^26–30 (iv) and all-organic solid-state approaches. In this article we will focus mainly on the latter.

While the physics of conventional p–n junctions is reasonably well understood and solar cell properties can be derived from materials parameters and the nature of the electrical contacts with electrodes, the understanding of the underlying science of organic solar cells is far less advanced and remains an intense subject of research. For instance, a key difference in the physics of organic semiconductors compared to their inorganic counterparts is the nature of the optically excited states. The absorption of a photon in organic materials leads to the formation of an exciton, i.e., a bound electron–hole pair. The exciton binding energy is typically large, on the order of or larger than 500 meV (such binding energies represent twenty times or more the thermal energy at room temperature, kT(300 K) = 26 meV,^31,32 to be compared with a few meV in the case of inorganic semiconductors. Consequently, optical absorption in organic materials does not lead directly to free electron and hole carriers that could readily generate an electrical current. Instead, to generate a current, the excitons must first dissociate. The excitonic character of their optical properties is a signature feature of organic semiconductors, which has impacted the design and geometry of organic photovoltaic devices for the past decades.

Our objective in this review is to introduce organic photovoltaics to the general energy and environmental science community. We discuss the science and engineering of solid-state organic photovoltaic materials and devices with an emphasis on the physical processes involved in the operation of these devices and the comparison with traditional inorganic p–n devices. Thus, we do not intend to provide a comprehensive survey of the wide range of materials and device geometries that have been developed; to get more familiar with the latter topics, we refer the reader to excellent recent reviews.^14,33–37

Materials and device structures

Organic molecular and polymeric semiconductors can form films with complex morphologies and varying degrees of order and packing modes through the interplay of a variety of non-covalent interactions. Their molecular structure consistently presents a backbone along which the carbon (or nitrogen, oxygen, sulfur) atoms are sp²-hybridized and thus possess a π-atomic orbital. The conjugation (overlap) of these π orbitals along the backbone results in the formation of delocalized π molecular orbitals, which define the frontier (HOMO and LUMO) electronic levels and determine the optical and electrical properties of the (macro)molecules. The overlap of the frontier π molecular orbitals between adjacent molecules or polymer chains characterizes the strength of the intermolecular electronic couplings (also called transfer integrals or tunneling matrix elements), which represent the key parameter governing charge carrier mobilities. In crystalline inorganic semiconductors, the three-dimensional character and rigidity of the lattice ensure wide valence and conduction bands and large charge carrier mobilities (typically on the order of several 10² to 10³ cm²V⁻¹ s⁻¹). In contrast, in organic semiconductors, the weakness of the electronic couplings (due to their intermolecular character), the large electron–vibration couplings (leading to pronounced geometry relaxations), and the disorder effects all conspire to produce more modest carrier mobilities due to charge-carrier localization and formation of polarons; transport then relies on polarons hopping from molecule to molecule (here and in the remainder of the text, “molecule” should also be taken as meaning “polymer chain segment” where appropriate).³⁸ As a result, the charge carrier mobilities strongly depend on the morphology and can vary over several orders of magnitude when going from highly disordered amorphous films (typically, 10⁻⁶ to 10⁻³ cm²V⁻¹ s⁻¹) to highly ordered crystalline materials (>1 cm²V⁻¹ s⁻¹).³⁸ To describe the operation of organic solar cells, energy diagrams can be drawn in which the valence and conduction band energies are replaced by the HOMO and LUMO (polaron level) energies, respectively.³⁹

The general structure of an organic photovoltaic device similar to that reported by Tang¹⁶ in 1986 is shown in Fig. 1. It consists of a transparent electrode, typically a conducting oxide such as indium-tin oxide (ITO), two organic light-absorbing layers, and a second electrode. The two organic layers are made of different organic semiconductors, one with an electron-donor character and the other with an electron-acceptor character. Electron-donor molecules (D) exhibit a low ionization potential (and thus a high-lying HOMO energy), while electron-acceptor molecules (A) possess a high electron affinity (and thus a low-lying LUMO energy) (see Fig. 1). The D and A layers must provide for efficient hole and electron transport, respectively.

Fig. 1 Cross-section of a bilayer organic solar cell. Chemical structure of examples of donor and acceptor molecules used in the seminal work of Tang. Donor: copper-phthalocyanine; acceptor: perylene tetracarboxylic derivative.

To compare the operation of an organic solar cell with conventional cells and to highlight their fundamental differences, it is useful to review the operation of inorganic solar cells based on p–n junctions. When an abrupt p–n junction is formed from doped inorganic semiconductors, the carrier distributions and concentrations in the materials under equilibrium conditions can be modeled and derived from the semiconductors parameters and the properties of their interfaces with the electrodes.^40,41 The condition of invariance of the equilibrium Fermi level leads to the conventional band energy diagram used to describe a p–n junction in which the energies of the valence band and conduction band are position dependent near the junction (see Fig. 2a). The charge redistribution that takes place near the junction (in the depletion region) leads to a built-in potential and to an electric field that is confined to the vicinity of the junction. Assuming constant doping on each side of the junction, the electric field outside the depletion region vanishes. Illumination of the materials leads to photogenerated electron/hole pairs, for which excitonic effects can be ignored at room temperature. The excess of photogenerated electrons in the n-doped semiconductor and holes in the p-doped side of the junction are negligible compared to the densities of majority carriers on both sides at thermal equilibrium. In contrast, electron and hole densities created in the p-doped and n-doped semiconductors, respectively, are significant relative to the densities of minority carriers at equilibrium in the dark. Carrier transport generally occurs inside the semiconductors mainly through a combination of drift and diffusion. At the junction, the minority carriers are swept away through a drift process due to the junction potential. Away from the junction, the transport of the minority carriers is mainly governed by diffusion. In such conditions, the photocurrent density of a conventional inorganic solar cell can be derived from the minority carrier diffusion equations obtained by combining the continuity and current equations. Solutions are found by considering the boundary conditions in which the photogenerated minority carriers at the junction are zero and where the interfaces of the semiconductors with the front and back electrodes are characterized by surface recombination velocities.^40,41

Fig. 2 Comparison of the energy level diagrams for inorganic and organic solar cells. (a) Energy levels in an inorganic p–n junction under illumination. ε_Fe and ε_Fh denote the quasi-Fermi levels in the n-type and p-type semiconductors. The difference between the quasi-Fermi level energies determines the maximum open-circuit voltage (V_OC) under illumination. Absorption of photons with an average photon energy larger than the band gap on either side of the junction in the n-type and p-type semiconductors (step 1) is followed by thermalization of the holes and electrons near the top of the valence and conduction bands, respectively (step 2). Minority carriers (electrons in the p-type semiconductor; holes in the n-type semiconductor) diffuse to the junction where they are swept away and accumulate on the other side of the junction where they become majority carriers (step 3). For the sake of simplicity, these three steps have been drawn for electrons, but the same applies to holes. (b) Energy level diagram of an organic heterojunction under illumination. IP(D) and EA(A) denote the ionization potential (HOMO level energy) of the donor molecular layer and the electron affinity (LUMO level energy) of the acceptor molecular layer, respectively. Absorption of photons with an average photon energy larger than the optical band gap on either side of the heterojunction (step 1) is followed by thermalization and the formation of excitons (step 2). Excitons diffuse to the heterojunction (step 3) where they dissociate and transfer an electron [hole] into the acceptor [donor] layer (step 4). The difference between IP(D) and EA(A) determines the maximum open circuit voltage (V_OC) under illumination. The Δ arrows denote the energy offsets between the ionization potential values (HOMO energies) and electron affinities (LUMO energies).

For the organic solar cell shown in Fig. 1, a parallel can be drawn between the two organic light-absorbing layers and the n- and p-doped semiconductors in an inorganic solar cell. However, unlike their inorganic counterparts, organic semiconductors in the device structure shown in Fig. 1 are in most cases essentially intrinsic (although some amounts of uncontrolled impurities are likely to be present). The interface between the two layers—the D/A heterojunction—is responsible for efficient exciton dissociation. Therefore, it plays a role for excitons similar to that played by the junction for minority carriers in inorganic cells. The electrons and holes created at the interface can be transported through the A and D layers, respectively, and collected at the two electrodes, thereby contributing to an electrical current in the external circuit. Organic semiconductors have large extinction coefficients compared to crystalline Si leading to efficient light harvesting in layers that are relatively thin with thicknesses in the range of 100–200 nm. Since the organic layers are sandwiched between electrodes with different work functions, a built-in potential appears and results in an electric field that assists the transport of charges.⁴² Thus, the electron and hole currents in the device under illumination after exciton dissociation are governed by an interplay of drift and diffusion.⁴³

While a critical step in inorganic solar cells is to collect minority carriers before they recombine by having them migrate to the opposite side of the junction, the challenge in organic cells is to dissociate the excitons before they decay to the ground state; this requires that they rapidly diffuse to the D/A heterojunctions, which is the only location where dissociation is efficient in pure materials (see Fig. 2b). Hence, the thickness of the organic layers has to be comparable to the exciton diffusion length L (L = (Dτ)^1/2 where D is the diffusion coefficient and τ the lifetime of the exciton). An optimal compromise regarding the thickness of the organic layers has to be found between allowing for efficient exciton diffusion to the heterojunction and efficient sun light absorption. The latter requires that the total absorbance of the film be preferably in the range 2–3 (corresponding to 86–95% absorption efficiency). However, if the layers of organic semiconductors have to be too thin as a result of small L values (e.g., d = 10–20 nm), the incident light does not get absorbed efficiently. For an organic semiconductor such as the copper-phthalocyanine compound (see structure in Fig. 1) used by Tang, the peak extinction coefficient at 620 nm is k = 0.74, which leads to an absorption coefficient (α = 4πk/λ where λ is the wavelength) of α = 0.015 nm⁻¹; for a 10 nm-thick film, this translates into an absorbance αd [absorption efficiency] of only 0.15 [14%] for a single pass and 0.3 [26%] for a double pass assuming the transmitted light gets reflected by the back electrode. Although devices based on pentacene have been shown recently to exhibit larger exciton diffusion lengths (70 nm),⁴⁴ this issue remains a limitation—often referred to as the “exciton bottleneck”—and has triggered the design of devices that are based on bulk heterojunctions in which the A and D components are mixed together and form an interpenetrating, phase-separated network D:A with a nanoscale morphology (see Fig. 3). This layer is then sandwiched between hole and electron transport (undoped or doped) layers. Bulk heterojunction devices can be fabricated by codeposition of two molecular materials^45–51 as pioneered by Yokoyama and co-workers⁵² (see Fig. 3a) or by blending two materials in solution as pioneered by the Heeger¹⁸ and Friend groups¹⁹ (see Fig. 3b). In such devices, the D/A heterojunction is distributed throughout the bulk of the composite film, enabling efficient exciton dissociation and leaving holes in the D component and electrons in the A component of the phase-separated network; control of the morphology^53–57 is critical to extract the carriers efficiently.

Fig. 3 Cross-sectional representations of organic photovoltaic cell geometries. (a) Geometry of a multilayer structure in which a mixed interlayer D:A is sandwiched between D and A layers. (b) Geometry of a bulk heterojunction. The chemical structures are examples of commonly used donor-like polymers (regio-regular poly(3-hexylthiophene): P3HT) and acceptor-like soluble molecules (methano-fullerene [6,6]-phenyl C₆₁-butyric acid methyl ester: PCBM). (c) Energy level diagram of a bulk heterojunction device. The HOMO and LUMO levels of the donor and acceptor materials are shown by solid and dot–dashed lines, respectively.

Device operation

We now turn to a more specific discussion of each of the processes taking place during device operation and highlight again what is particular to π-conjugated organic materials with respect to conventional inorganic materials.

Optical absorption

Conjugated oligomers and polymers display absorption bands that are usually: (i) very intense, as a result of the large wavefunction overlap between the ground state and the lowest excited states; and (ii) broad, due to the significant geometry relaxations that take place in the excited states (the width of the absorption bands can reach over 1 eV). In π-conjugated systems, the bond lengths are primarily modulated by the π-electron densities on the bonds; any change in electronic state, due to excitation or ionization, modifies the π-electron bond densities and thus the bond lengths.⁵⁸ This strong connection between electronic structure and geometric structure is referred to as strong electron–vibration (phonon) coupling.

The widths of the absorption bands allow for good matching with a sizeable portion of the solar spectrum. Once promoted to an excited state, the π-conjugated system relaxes down to the bottom of the potential energy surface of the lowest excited state, the excited state reaches its equilibrium geometry, and an exciton forms (we note that this relaxation/thermalization process constitutes a first source of power loss). We recall that the exciton is a bound electron-hole pair with a typical binding energy on the order of a few tenths of an eV. This high value is a reflection not only of the rather low dielectric constant of π-conjugated organic materials but also of the significant electron correlation and geometry relaxation effects present in these materials (because of the geometry relaxation associated to its formation, the exciton is sometimes referred to as a polaron-exciton). Thus, in contrast to inorganic semiconductors, the absorption of a photon at room temperature in conjugated materials does not lead to free charge carriers but to neutral, bound electron-hole pairs. This is the reason why two components, an electron donor and an electron acceptor, are required to promote the generation of charge carriers.

In general, the ground state of the π-conjugated system is singlet (total spin multiplicity is zero) and is denoted S₀; the lowest singlet excited state, S₁, is usually one-photon allowed.⁵⁸ In pure hydrocarbons with a coplanar conformation (such as pentacene), the spin–orbit coupling to triplet states (with total spin multiplicity of one) is vanishingly small and intersystem crossing between the singlet and triplet manifolds can be neglected; in systems with heavy atoms or far from planarity (for instance, metal phthalocyanines or fullerenes), this is no longer true and intersystem crossing to triplet excitons can be efficient. The lowest-energy triplet exciton, T₁, often lies a few tenths of an eV below S₁.

Since organic solar cells are composed of several thin layers of materials with different optical properties, mismatch of the complex refractive index at multiple interfaces leads to multiple reflections that produce optical interference effects. As a result, the light distribution inside the solar cell is highly inhomogeneous and determined by a complicated interplay of the relative optical constants of the materials and their thickness.^59,60 Knowledge of the light distribution in the solar cell is important to model the steady-state exciton distribution and consequently the photocurrent produced in the solar cell.

Exciton diffusion

In order to generate separated negative and positive charges, the excitons need to diffuse to the donor–acceptor interface where they can dissociate. Since excitons are neutral species, their motion is not influenced by any electric field and they diffuse via random hops; importantly, they need to reach the interface prior to their decay back to the ground state.

The hopping of singlet excitons is usually described via a generalized Förster mechanism, which involves the long-range electrostatic coupling between the excitation transition dipoles at the initial and final sites (we note that the traditional point-dipole Förster model is totally inappropriate here⁶¹ as it is based on resonance energy transfer between distant molecules); in the case of triplet excitons, hops are restricted to adjacent sites, as they depend on a short-range exchange (Dexter-type) mechanism. Thus, singlet excitons can move more quickly than triplets but decay more quickly as well (on a ns scale vs. µs or ms scale—which represents the difference between the fluoresecence and phosphorescence lifetimes). As a result, the efficiency with which singlets and triplets reach the interface is very much system dependent.

Exciton dissociation at the donor–acceptor interface

At the D/A interface, excitons can dissociate provided their energy is higher than that of charge-transfer or charge-separated states; here, we refer to charge-separated (CS) states as states where the electron and the hole have been completely freed from one another, while in charge-transfer (CT) states the electron and hole are still somewhat bound to one another. Interestingly, at the present time, no clear picture has emerged to describe the exciton dissociation process. Some of the most relevant electronic states are depicted in Fig. 4.

Fig. 4 Electronic state diagram. S₀ denotes the singlet ground state of the donor or the acceptor and S₁, the first singlet excited state reached after optical excitation (exciton state); at the D/A interface, intermolecular charge transfer leads to charge-transfer (CT) states where the hole is on donor molecule(s) and the electron on acceptor molecule(s): CT₁ is the lowest energy charge-transfer state; the CT_n states represent higher-energy, more diffuse charge-transfer states; the final state relevant for photovoltaics operation is a charge-separated state (CS) where the hole in the donor layer and the electron in the acceptor layer have become free from one another; the k_i terms indicate competing transfer rates between various electronic states.

In most instances, the dissociation process is described as involving a transition from the exciton state down to the lowest CT state, which corresponds to the situation where the hole sits on the HOMO level of a D molecule and the electron on the LUMO level of an adjacent A molecule (see Fig. 4). However, in such a case, since they remain in close proximity, the electron and the hole are still rather strongly Coulombically bound, which is precisely the reason why that CT state is lowest in energy. Several scenarios have been proposed to explain the eventual separation of the charges from the lowest CT state, for instance, the presence of disorder or dipoles at the interface or the assistance of phonons,^62,63 which would make the k_CS1 charge-separation rate larger than the k_CR charge-recombination rate.

Another proposition has been recently advanced.⁶⁴ It involves the efficient coupling of the exciton arriving at the interface to higher-lying CT_n states. By definition, such states are more diffuse than the lowest CT state and could be delocalized over a few sites; as a result, the electron and the hole could become more distant and more easily screened from one another, which would lead to easier charge separation. For that process to be relevant, the k_CTn and k_CSn rates have to be larger than those bringing the system down to CT₁.

Note that many factors can complicate the description of the D/A interfacial processes. Suppose that the exciton reaching the interface has formed in the donor. First, instead of direct electron transfer from D to A, there could occur energy transfer leading to the formation of an exciton on A, followed by hole transfer from A to D; this process has been demonstrated in the case of oligophenylene-fullerene dyads.⁶⁵ The final state is the same as for the direct electron-transfer process; however, the rates involved in the energy-transfer and hole-transfer processes can be markedly different. Secondly, even when singlet excitons are exclusively formed in D, triplet excitons can appear at the interface. For instance in bis-dicyanovinyl-oligothiophenes/fullerene blends, it has been observed⁶⁶ that, for certain thiophene oligomer lengths, excitons can efficiently transfer to C₆₀ where the large intersystem crossing leads to the formation of triplet excitons, which then hop back to the donor; such processes do not result in charge separation and constitute a loss mechanism.

Thus, the situation is much more complex than what the simple HOMO-LUMO diagrams often found in the literature (and illustrated in Fig. 3c) would lead one to believe. In addition, charge separation can also be influenced by concentration and morphology gradients near the heterojunction that take place during the formation of the organic films, or can be assisted by local electric fields.⁶⁷ New theoretical methodologies are being developed to provide better descriptions and understanding of all these competing mechanisms.

Charge carrier mobility and collection at electrodes

Once the charges have separated, they can drift and diffuse towards their respective electrodes with efficiency depending on their mobilities. Because of the large electron-vibration coupling in π-conjugated materials and of disorder effects, each charge is associated to a local geometry relaxation and constitutes a polaron (radical-ion in chemical terminology) which hops from molecule to molecule.⁶⁸ The corresponding polaronic electronic state has an energy defined by the (adiabatic) ionization potential (IP) of the donor or electron affinity (EA) of the acceptor (note that in the simple HOMO–LUMO diagrams, E_final would be crudely approximated by the difference between the energies of the donor HOMO and acceptor LUMO). The sum of the energies of the polaron states for the donor and acceptor, E_final = IP(D) + EA(A), represents the energy of the final state of the system, see Fig. 4. To a large extent, E_final defines the upper limit for the open-circuit voltage of the solar cell, as discussed below.

The nature of the electrode/organic layer interfaces is complex. The efficiency of the charge collection process cannot be simply determined from the difference between the workfunction of the isolated electrode and the donor IP or acceptor EA. The deposition of organic layers on electrodes (or vice versa) lead to interfacial charge-density redistributions and/or geometry modifications that strongly affect the alignment of the organic frontier electronic levels vs. the electrode Fermi level.³⁸ Much remains to be done to understand the intricate details of these interfaces. Surface modification of the electrodesvia deposition of self-assembled monolayers⁶⁹ is an efficient way to enhance the quality of the electrical contact as well as, in particular when dealing with conducting oxide electrodes, to promote compatibilization between the (hydrophilic) oxide surface and (hydrophobic) organic layer.

Device performance and modeling

Here, we discuss the electrical characteristics of organic solar cells and their performance. In the dark the solar cell works as a diode. As for conventional p–n solar cells, an organic photovoltaic device can be approximated by an equivalent circuit, see Fig. 5d, comprised of: (i) a diode with reverse saturation current density J₀ (current density in the dark at reverse bias) and ideality factor n; (ii) a current source (J_ph), which corresponds to the photocurrent upon illumination; (iii) a series resistance (R_S), which has to be minimized and takes account of the finite conductivity of the semiconducting material, the contact resistance between the semiconductors and the adjacent electrodes, and the resistance associated with electrodes and interconnections; and (iv) a shunt (R_P) resistance, which needs to be maximized and takes into account the loss of carriers via possible leakage paths; the latter include structural defects such as pinholes in the film, or recombination centers introduced by impurities. Solving for this simple circuit provides the following analytical expression for the current–voltage characteristics, referred to as the Shockley equation:^3–5where e denotes the elementary charge, kT the thermal energy, and A the area of the cell. Analysis of eqn (1) in various regimes of photocurrent shows that the series resistance is the critical factor, especially in regimes of high photocurrents, J_ph. From eqn (1), equations for the open-circuit voltage V_OC and the short-circuit current density J_SC can be derived:

Fig. 5 Optical and electrical properties of solar cells. (a) Spectral photon flux density in the standardized AM 1.5 G illumination conditions and corresponding integrated current that would be produced if each photon contributes to current with unity efficiency. (b) Current–density voltage characteristics of a solar cell in the dark and under illumination. (c) Semi-logarithmic plot of the same electrical characteristics, illustrating the effects of the parasitic resistances R_S and R_P in forward and reverse bias. (d) Equivalent circuit used to model solar cells. The notations are defined in the text.

Eqns (1)–(3) usually need to be solved numerically except for cases where R_S is very small and/or R_P sufficiently large so that the effect of R_S or R_P can be ignored; in such instances, the approximate expressions on the right hand side of eqns (2)–(3) apply. When the device is under illumination, two quantities can be easily determined experimentally: the intersects of the electrical characteristics with the vertical and horizontal axes, which correspond to J_SC and V_OC, respectively (see Fig. 5b). At any point on the electrical characteristic in the fourth quadrant (J_SC negative and V_OC positive), the solar cell produces an electrical power density given by the product of voltage and current density. This product is maximized at a point that corresponds to voltage V_max and current density J_max. The power conversion efficiency η, which represents the most important metric for a photovoltaic cell, is then defined as:

where P_inc is the incident power density and FF denotes the fill factor. These parameters are illustrated in Fig. 5b. The effects of the parasitic resistances on the shape of the current density/voltage characteristics are illustrated in Fig. 5c. A finite value of the series resistance R_S limits the current density in forward bias, while a finite shunt resistance R_P is responsible for a dark current increase in reverse bias. To characterize quantitatively the performance of solar cells for terrestrial applications, standardized illumination conditions are used in which the spectrum of the source simulates the solar spectrum (AM 1.5 G, see Fig. 5a) and has an intensity on the order of 100 mW cm⁻²; this corresponds to the average intensity of sun light with an angle of incidence θ = 48° relative to the normal to the earth's surface (AM denotes the air mass = 1/cosθ; G stands for global and refers to a small contribution of diffuse light to the direct incident light). Importantly, according to eqn (4), the power conversion efficiency of a solar cell is determined by three parameters: the fill factor, the short-circuit current density, and the open-circuit voltage. We now examine the limits of these three parameters in organic solar cells and discuss how they relate to materials and contacts properties.

Fill factor

The maximum value for the fill factor is a function of the open-circuit voltage V_OC and the ideality factor of the diode n (optimally equal to 1). As for inorganic solar cells based on p–n junctions, its maximum value can be described by the empirical expression:³where V_OC is a normalized voltage defined as V_OC = eV_OC/nkT. Eqn (5) is a good approximation for V_OC values > 10. For organic solar cells with V_OC = 0.5–1 V, and ideality factors in the range of n = 1.5–2, this condition is satisfied. We note that in p–n junction-based cells, the various recombination mechanisms are taken into account by drawing equivalent circuits with two diodes, where the first diode (with n = 1) describes radiative band-to-band recombinations while the second (with n = 2) describes recombination via impurities with energy states within the band gap. In organic solar cells, little is known at this stage about the specific recombination processes of excitons and carriers. In view of the higher disorder and impurity levels present in amorphous organic semiconductors compared to the pure grades of silicon wafers that can be fabricated, ideality factors that deviate from unity are expected; this negatively impacts both the maximum fill factor and the maximum open-circuit voltage (for instance, for V_OC = 0.5 V, FF₀ drops from 0.80 to 0.69 when n goes from 1 to 2, to be compared with FF₀ = 0.85 in monocrystalline silicon solar cells). It is worth mentioning that the impact of the parasitic resistances R_S and R_P in reducing the fill factor varies with cell performance, cell area, and operation conditions. A rule of thumb is that the value of R_S must be small compared to the characteristic resistance defined as R_CH = V_OC/J_SCA (where A is the cell area) while R_P must be large compared to R_CH.

Short-circuit current density

The maximum J_SC is given by:Here, N_ph(λ) is the photon flux density in the incident AM 1.5 G spectrum (see Fig. 5a) at wavelength λ and a total intensity of 100 mW cm⁻² (integrated over the full spectrum). The external quantum efficiency η_EQE(λ) is defined by how efficiently an incident photon gives rise to an electron flowing in the external circuit. In organic cells, η_EQE can be broken down into the product of efficiencies associated with each of the steps discussed previously: absorption, exciton diffusion, exciton dissociation into free carriers, charge transport, and charge collection. Upper limits on short-circuit current density are obtained according to eqn (6) by integrating from the high photon energy side (short wavelength) of the spectrum to the wavelength corresponding to the optical band gap of the material (as shown in Fig. 5a). Hence, the smaller the optical band gap, the larger the maximum short-circuit current. In the case of silicon, the band gap is 1.1. eV (∼1130 nm), which yields a maximum value J_SC = 43.6 mA cm⁻² for AM 1.5 G. It is worth mentioning that it is the fraction of photons that are harvested in the AM 1.5 G spectrum that matters rather than the fraction of the intensity contained in that same portion of the spectrum. Since the high-energy side of the spectrum gets harvested, the average photon energy in the absorbed portion of the spectrum is larger than the band gap energy. In silicon, the average photon energy is 1.8 eV, to be compared with the 1.1 eV band gap. As mentioned earlier, this results in a significant loss mechanism since energy goes down during thermalization of the electron-hole pairs. These losses can be somewhat minimized by using tandem cell geometries²² in which materials with different optical band gaps are stacked on top of one another and absorb different parts of the spectrum. Tandem cell geometries have recently been demonstrated with organic cells.^70–77

Open-circuit voltage

As in the case of conventional solar cells, maximization of the short-circuit current density by using organic semiconductors with decreasing optical absorption gap is not overall an effective strategy since the maximum open-circuit voltage presents an opposite trend with optical absorption gap. It follows that the determination of the optimum light-harvesting conditions to maximize the efficiency of organic solar cells depends largely on the understanding of the origin of V_OC and its dependence on materials properties, in particular the relative energies of the relevant energy levels at the organic heterojunction.

Neglecting the effects of the parasitic resistances, according to eqn (2), V_OC is a logarithmic function of the ratio of the short-circuit current and the reverse saturation current. In the dark, in the absence of carriers, the device is in equilibrium and no photovoltage or V_OC is observed since the dark J–V characteristics cross the origin. Upon illumination, absorbed photons generate charge carriers, whose distributions can be described by non-equilibrium quasi-Fermi levels (see Fig. 2a). For illumination levels such that the produced photocurrent is larger than the reverse saturation current (which holds true under average illumination conditions), V_OC is observed experimentally to increase logarithmically with intensity. Since the maximum short-circuit current can be estimated using eqn (6), reaching the maximum V_OC value requires that the reverse saturation current density J₀ be kept at a minimum. Recent studies by Scharber and co-workers⁷⁸ on bulk heterojunction cells and Rand and co-workers⁷⁹ on multilayer solar cells, indicate that V_OC in organic solar cells depends on the energy difference between the ionization potential of the D component and the electron affinity of the A component forming the heterojunction. Studies of the temperature dependence of the reverse saturation current J₀ by Waldauf et al.⁸⁰ in bulk heterojunctions and Rand⁷⁹ in multilayer structures of small molecules, show that the reverse saturation current can be approximated by:

where B is a coefficient with a value in the range of 1000 A cm⁻² and E_final = IP(D) + EA(A). Eqn (7) shows that the reverse saturation current is thermally activated with a barrier height equal to E_final/n′ where n′ in an ideality factor that corrects for effects such as vaccum level misalignments at the heterojunction caused by energy level bending and interfaces dipoles and the formation of charge-transfer states. By combining eqns (2), (3), and (7), the maximum open-circuit voltage V_OC for an organic cell can then be written as:

From the energy diagram in Fig. 2b, E_final = IP(D) + EA(A) is increased as the energy offsets Δ between the D and A molecular states are decreased. However, too strong a reduction in Δ compromises the efficiency of exciton dissociation at the heterojunction and thus decreases the photocurrent; it is generally accepted that Δ has to be on the order of ∼0.3–0.5 eV to overcome the exciton binding energy. In this case, optimized conditions to maximize the power conversion efficiency in an organic single heterojunction cell are obtained for a band gap in the range E_G = 1.6–1.9 eV (775–650 nm).

It is worth mentioning that molecules with EA energies larger than 4.2 eV are less sensitive to oxidation in the presence of oxygen and moisture.⁸¹ Hence, optimized material combinations should not only satisfy conditions associated with relative state energies but also absolute energies referenced to the vacuum level.

Based on this analysis, a limit of the maximum power efficiency of an idealized single junction organic solar cell can be estimated. With an optimized optical band gap of E_G = 1.6 eV (775 nm) and a hypothetical average external quantum efficiency of η_EQE = 0.8, a maximum short-circuit current of J_SC,max = 20.2 mA cm⁻² is calculated from eqn (6) (see also Fig. 5a). A required energy offset of Δ = 0.5 eV would translate into E_final = IP(D) + EA(A) = E_G − Δ = 1.1 eV. From eqn (8) and assuming that the solar cell has an ideality factor of n = 1.5 and with n′ ≈ n, the maximum open-circuit voltage is estimated to be V_OC,max = 0.68 V. Such a V_OC,max leads to a normalized voltage V_OC = eV_OC/nkT= 17.4 V, from which, according to eqn (5), an ideal maximum fill factor FF₀ = 0.79 is obtained. Following eqn (4), these values translate into a power conversion efficiency η = 10.8%. Table 1 collects a summary of these estimates and provides examples of parameters measured in solar cells fabricated from different materials.

Table 1 Selected examples of device parameters for solar cells measured under the standard global spectrum (AM 1.5G, 100 mW cm⁻²)

Device type	J SC/mA cm^−2	V OC/V	FF	PCE^c (%)	Reference
a PERL: passivated emitter, rear locally diffused; fabricated from high quality fusion zone (FZ) monocrystalline silicon substrates. b Triple junction cell. c PCE: Power conversion efficiency.
Monocrystalline Si PERL^acell	42.2	0.706	0.828	24.7	82
Multicrystalline Si cell	35.6	0.631	0.808	18.2	83
Monocrystalline Si commercial module	—	—	—	16.9	SunPower
a-Si cell ^b	8.11	2.297	0.697	13.0	84
a-Si commercial module	—	—	—	6.3	United Solar
CdTe cell	26.08	0.840	0.731	16.0	10
CIGS cell	35.7	0.689	0.781	19.2	9
Grätzel cell	20.53	0.721	0.704	10.4	85
Nanocrystal hybrid cell	13.2	0.45	0.49	2.9	24
OPV: 1986 Tang cell	2.3	0.45	0.65	0.9	16
OPV: Small molecule cell	15.4	0.50	0.46	3.5	45
OPV: P3HT:PCBM single cell	9.5	0.63	0.68	5.1	56
OPV tandem cell	7.8	1.24	0.67	6.5	76
Ideal single heterojunction OPV cell	20.2	0.68	0.79	10.8	See text

---

Future prospects and challenges

The future of organic solar cells as a pervasive technology for portable power will largely rely on their economic potential.⁸⁶ This depends on a number of intricate factors such as efficiency, manufacturing cost, lifetime, form factor, weight, scalability, and sustainable manufacturing. At this point, two main manufacturing techniques can be foreseen, vacuum processing and wet processing. Vacuum processing presents the advantages of relatively easy fabrication of, on one hand, high-quality thin films from highly purified materials with well-controlled thickness and, on the other hand, devices with complex multilayer architectures. The limiting factors are the deposition rates and the tooling costs associated with vacuum techniques. Wet processing allows for high throughputs using various printing techniques and holds the long-term promise of lowest manufacturing costs. However, printing of organic semiconductors into pinhole-free and 100 nm-thick films over large areas remains a major challenge. Semiconductor inks generally have low viscosity which limits the range of adequate printing techniques. Furthermore, substrates and electrodes have surface energies that are significantly different from those of organic semiconductor inks, which can lead to wetting issues. Vertical segregation⁸⁷ and other rheological effects also make it difficult to control the nanoscale phase-separated morphology⁸⁸ necessary for good operation of bulk heterojunctions. Stability of the inks after formulation is another challenge that vacuum processing does not face.

A cost factor common to all material platforms is packaging. Coatings with small transmission rates to oxygen and moisture are necessary to protect the organic semiconductors from undergoing photo-oxidative reactions that limits their stability over time. Packaging techniques must be compatible with the substrate and active materials⁸⁹ and keep the overall manufacturing cost low. Since highly flexible form factors are possible, much work remains to be done to understand how flexing affects the operation of the devices and their packaging. Most laboratory cells with efficiencies in the range of 2–6% have been fabricated over small areas, and area scaling without loss of efficiency is required. A clear advantage of organics over crystalline silicon is the relative ease of monolithic integration,⁹⁰ which reduces the module assembly cost considerably.

If organic photovoltaic technologies mature beyond niche consumer market applications and become players in power generation, their composition must be based on materials available on large scales at low cost. Also, concerns about the limited supply of indium ($900 per kg) are currently fueling research efforts to replace ITO as the transparent electrode.^91–93 Promising results have been demonstrated, for instance, with polymers doped with carbon nanotubes.

Last but not least, organic solar cells must demonstrate lifetimes of several years. While organic cells might not show twenty years of operational lifetime like crystalline silicon cells, the recent demonstration of 100 000 h operational lifetime in organic light-emitting diodes is indicative that long lifetimes are within reach with organic semiconductors.

To summarize, organic photovoltaics provides an exciting playground at the frontiers of science, engineering, and technology. Advances in the near term are likely to lead to solar cells with efficiencies close to 10% in single heterojunction geometries and efficiencies up to 15% in tandem-cell geometries. If organic photovoltaics holds its promise, it can soon become an ubiquitous, clean and sustainable technology for portable power and potentially provide large-scale energy production for future generations.

Acknowledgements

This material is based upon work supported in part by the STC Program of the National Science Foundation under Agreement Number DMR-0120967, the Office of Naval Research, the Department of Energy, the Georgia Research Alliance, and the AtlanTICC Alliance.

References

1. E. Becquerel, Mémoire sur les effets électriques produits sous l'influence des rayons solaires, C. R. Acad. Sci., 1839, 9, 561–567.
2. D. M. Chapin, C. S. Fuller and G. L. Pearson, A new silicon p-n junction photocell for converting solar radiation into electrical power, J. Appl. Phys., 1954, 25, 676.
3. M. A. Green, Solar Cells, The University of New South Wales, Kensington, Australia, 1998.
4. P. Wurfel, Physics of Solar Cells, Wiley VCH, Weinheim, Germany, 2005.
5. R. H. Bube, Photovoltaic Materials, Imperial College Press, London, UK, 1998.
6. For general information about solar energy, consult the International Energy Agency Photovoltaic Power Systems Program at http://www.iea-pvps.org.
7. S. Hegedus, Thin film solar modules: The low cost, high throughput and versatile alternative to Si wafers, Prog. Photovoltaics, 2006, 14, 393–411.
8. A. Slaoui and R. T. Collins, Advanced inorganic materials for photovoltaics, MRS Bull., 2007, 32, 211.
9. K. Ramanathan et al., Properties of 19.2% efficiency ZnO/CdS/CuInGaSe₂ thin-film solar cells, Prog. Photovoltaics, 2003, 11, 225.
10. T. Aramoto et al., 16.0% efficient thin-film CdS/CdTe solar cells, Jpn. J. Appl. Phys., Part 1, 1997, 36, 6304.
11. R. H. Friend et al., Electroluminescence in conjugated polymers, Nature, 1999, 397, 121.
12. F. Garnier, R. Hajlaoui, A. Yassar and P. Srivastava, All-polymer field-effect transistor realized by printing techniques, Science, 1994, 265, 1684.
13. C. D. Dimitrakopoulos and P. R. L. Malenfant, Organic thin film transistors for large area electronics, Adv. Mater., 2002, 14, 99.
14. P. Peumans, A. Yakimov and S. R. Forrest, Small molecular weight organic thin-film photodetectors and solar cells, J. Appl. Phys., 2003, 93, 3693.
15. M. Pope and C. E. Swenberg, Electronic Processes in Organic Crystals and Polymers, Oxford University Press, New York, 1999.
16. C. W. Tang, Two-layer organic photovoltaic cell, Appl. Phys. Lett., 1986, 48, 183.
17. D. Kearns and M. Calvin, Photovoltaic effect and photoconductivity in organic laminated systems, J. Chem. Phys., 1958, 29, 950–951.
18. G. Yu, J. Gao, J. C. Hummelen, F. Wudl and A. J. Heeger, Polymer photovoltaic cells – enhanced efficiencies via a network of internal donor–acceptor heterojunctions, Science, 1995, 270, 1789.
19. J. J. M. Halls et al., Efficient photodiodes from interpenetrating polymer networks, Nature, 1995, 376, 498.
20. M. Grätzel, Photoelectrochemical cells, Nature, 2001, 414, 338.
21. M. K. Nazeeruddin et al., Conversion of light to electricity by cis-x₂bis(2,2′-bipyridyl-4,4′-dicarboxylate)ruthenium(II) charge-transfer sensitizers (x = Cl⁻, Br⁻, I⁻, CN⁻, and SCN⁻) on nanocrystalline TiO₂ electrodes, J. Am. Chem. Soc., 1993, 115, 6382.
22. F. Dimroth and S. Kurtz, High-efficiency multijunction solar cells, MRS Bull., 2007, 32, 230.
23. W. U. Huynh, J. J. Dittmer and A. P. Alivisatos, Hybrid nanorod-polymer solar cells, Science, 2002, 295, 2425.
24. I. Gur, N. A. Fromer, M. L. Geier and A. P. Alivisatos, Air-stable all-inorganic nanocrystal solar cells processed from solution, Science, 2005, 310, 462.
25. D. J. Milliron, I. Gur and A. P. Alivisatos, Hybrid organic–nanocrystal solar cells, MRS Bull., 2005, 30, 41.
26. A. Petrella et al., Colloidal TiO₂ nanocrystals/MEH:PPV nanocomposites: Photo(electro)chemical study, J. Phys. Chem. B, 2005, 109, 1554.
27. K. Sayama et al., Efficient sensitization of nanocrystalline TiO₂ films with cyanine and merocyanine organic dyes, Sol. Energy Mater. Sol. Cells, 2003, 80, 47.
28. A. C. Arango et al., Efficient titanium oxide/conjugated polymer photovoltaics for solar energy conversion, Adv. Mater., 2000, 12, 1689.
29. K. M. Coakley, Y. X. Liu, C. Goh and M. D. McGehee, Ordered organic-inorganic bulk heterojunction photovoltaic cells, MRS Bull., 2005, 30, 37.
30. K. M. Coakley, Y. X. Liu, M. D. McGehee, K. L. Frindell and G. D. Stucky, Infiltrating semiconducting polymers into self-assembled mesoporous titania films for photovoltaic applications, Adv. Funct. Mater., 2003, 13, 301.
31. J. L. Brédas, J. Cornil and A. J. Heeger, The exciton binding energy in luminescent conjugated polymers, Adv. Mater., 1996, 8, 447.
32. S. F. Alvarado, P. F. Seidler, D. G. Lidzey and D. D. C. Bradley, Direct determination of the exciton binding energy of conjugated polymers using a scanning tunneling microscope, Phys. Rev. Lett., 1998, 81, 1082.
33. J. Nelson, Organic photovoltaic films, Curr. Opin. Solid State Mater. Sci., 2002, 6, 87.
34. Organic Photovoltaics: Mechanisms, Materials and Devices, ed. S. S. Sun and N. S. Sariciftci, Taylor and Francis, Boca Raton, FL, 2005.
35. S. E. Shaheen, D. S. Ginley and G. E. Jabbour, Organic-based photovoltaics. Toward lowm-cost power generation, MRS Bull., 2005, 30, 10.
36. H. Hoppe and N. S. Sariciftci, Organic solar cells: An overview, J. Mater. Res., 2004, 19, 1924.
37. H. Spanggaard and F. C. Krebs, A brief history of the development of organic and polymeric photovoltaics, Sol. Energy Mater. Sol. Cells, 2004, 83, 125.
38. V. Coropceanu et al., Charge transport in organic semiconductors, Chem. Rev., 2007, 107, 926.
39. Conjugated Polymers and Molecular Interfaces, ed. W. R. Salaneck, K. Seki, A. Kahn and J.-J. Pireaux, Marcel Dekker, New York, 2002.
40. R. F. Pierret Semiconductor Device Fundamentals, Addison Wesley, Reading, Massachusetts, 1996.
41. S. M. Sze and K. K. Ng, Physics of Semiconductor Devices, Wiley Interscience, Hoboken, New Jersey, 2007.
42. I. H. Campbell, T. W. Hagler, D. L. Smith and J. P. Ferraris, Direct measurement of conjugated polymer electronic excitation energies using metal/polymer/metal structures, Phys. Rev. Lett., 1996, 76, 1900.
43. P. W. M. Blom, V. D. Mihailetchi, L. J. A. Koster and D. E. Markov, Device physics of polymer: Fullerene bulk heterojunction solar cells, Adv. Mater., 2007, 19, 1551.
44. S. Yoo, B. Domercq and B. Kippelen, Efficient thin-film organic solar cells based on pentacene/C₆₀ heterojunctions, Appl. Phys. Lett., 2004, 85, 5427.
45. S. Uchida, J. G. Xue, B. P. Rand and S. R. Forrest, Organic small molecule solar cells with a homogeneously mixed copper phthalocyanine:C₆₀ active layer, Appl. Phys. Lett., 2004, 84, 4218.
46. P. Peumans, S. Uchida and S. R. Forrest, Efficient bulk heterojunction photovoltaic cells using small-molecular-weight organic thin films, Nature, 2003, 425, 158.
47. J. Drechsel et al., Efficient organic solar cells based on a double p-i-n architecture using doped wide-gap transport layers, Appl. Phys. Lett., 2005, 86, 244102.
48. F. Yang, M. Shtein and S. R. Forrest, Controlled growth of a molecular bulk heterojunction photovoltaic cell, Nat. Mater., 2005, 4, 37.
49. A. K. Pandey, S. Dabos-Seignon and J. M. Nunzi, Pentacene: PTCDI-C₁₃H₂₇ molecular blends efficiently harvest light for solar cell applications, Appl. Phys. Lett., 2006, 89.
50. K. Suemori, T. Miyata, M. Yokoyama and M. Hiramoto, Three-layered organic solar cells incorporating a nanostructure-optimized phthalocyanine:fullerene codeposited interlayer, Appl. Phys. Lett., 2005, 86.
51. B. Maennig et al., Organic p-i-n solar cells, Appl. Phys. A, 2004, 79, 1.
52. M. Hiramoto, H. Fujiwara and M. Yokoyama, 3-layered organic solar-cell with a photoactive interlayer of codeposited pigments, Appl. Phys. Lett., 1991, 58, 1062.
53. E. Moons, Conjugated polymer blends: Linking film morphology to performance of light emitting diodes and photodiodes, J. Phys.: Condens. Matter, 2002, 14, 12235.
54. F. Padinger, R. S. Rittberger and N. S. Sariciftci, Effects of postproduction treatment on plastic solar cells, Adv. Funct. Mater., 2003, 13, 85.
55. J. Peet et al., Efficiency enhancement in low-bandgap polymer solar cells by processing with alkane dithiols, Nat. Mater., 2007, 6, 497.
56. W. L. Ma, C. Y. Yang, X. Gong, K. Lee and A. J. Heeger, Thermally stable, efficient polymer solar cells with nanoscale control of the interpenetrating network morphology, Adv. Funct. Mater., 2005, 15, 1617.
57. G. Li et al., High-efficiency solution processable polymer photovoltaic cells by self-organization of polymer blends, Nat. Mater., 2005, 4, 864.
58. L. Salem, The Molecular Orbital Theory of Conjugated Systems, Benjamin, New York, 1966.
59. L. A. A. Pettersson, L. S. Roman and O. Inganas, Modeling photocurrent action spectra of photovoltaic devices based on organic thin films, J. Appl. Phys., 1999, 86, 487.
60. J. Gilot, I. Barbu, M. M. Wienk and R. A. J. Janssen, The use of ZnO as optical spacer in polymer solar cells: Theoretical and experimental study, Appl. Phys. Lett., 2007, 91, 113520.
61. H. Wiesenhofer et al., Limitations of the Förster description of singlet exciton migration: The illustrative example of energy transfer to ketonic defects in ladder-type poly(para-phenylenes), Adv. Funct. Mater., 2005, 15, 155.
62. V. I. Arkhipov, P. Heremans and H. Bassler, Why is exciton dissociation so efficient at the interface between a conjugated polymer and an electron acceptor?, Appl. Phys. Lett., 2003, 82, 4605.
63. T. Offermans, S. C. J. Meskers and R. A. J. Janssen, Charge recombination in a poly(para-phenylene vinylene)-fullerene derivative composite film studied by transient, nonresonant, hole-burning spectroscopy, J. Chem. Phys., 2003, 119, 10924.
64. A. C. Morteani, P. Sreearunothai, L. M. Herz, R. H. Friend and C. Silva, Exciton regeneration at polymeric semiconductor heterojunctions, Phys. Rev. Lett., 2004, 92, 247402.
65. E. Peeters et al., Synthesis, photophysical properties, and photovoltaic devices of oligo(p-phenylene vinylene)-fullerene dyads, J. Phys. Chem. B, 2000, 104, 10174.
66. R. Schueppel et al., Enhanced photogeneration of triplet excitons in an oligothiophene fullerene blend, ChemPhysChem, 2007, 8, 1497.
67. M. Koeler, M. C. Santos and M. G. E. da Luz, Positional disorder enhancement of exciton dissociation at donor/acceptor interface, J. Appl. Phys., 2006, 99, 053702.
68. G. Heimel et al., Toward control of the metal-organic interfacial electronic structure in molecular electronics: A first-principles study on self-assembled monolayers of pi-conjugated molecules on noble metals, Nano Lett., 2007, 7, 932.
69. S. Khodabakhsh, B. M. Sanderson, J. Nelson and T. S. Jones, Using self-assembling dipole molecules to improve charge collection in molecular solar cells, Adv. Funct. Mater., 2006, 16, 95.
70. M. Hiramoto, M. Suezaki and M. Yokoyama, Effect of thin gold interstitial-layer on the photovoltaic properties of tandem organic solar-cells, Chem. Lett., 1990, 327.
71. B. P. Rand, P. Peumans and S. R. Forrest, Long-range absorption enhancement in organic tandem thin-film solar cells containing silver nanoclusters, J. Appl. Phys., 2004, 96, 7519.
72. J. G. Xue, S. Uchida, B. P. Rand and S. R. Forrest, Asymmetric tandem organic photovoltaic cells with hybrid planar-mixed molecular heterojunctions, Appl. Phys. Lett., 2004, 85, 5757.
73. A. Hadipour, B. de Boer and P. W. M. Blom, Solution-processed organic tandem solar cells with embedded optical spacers, J. Appl. Phys., 2007, 102 074506.
74. J. Gilot, M. M. Wienk and R. A. J. Janssen, Double and triple junction polymer solar cells processed from solution, Appl. Phys. Lett., 2007, 90, 143512.
75. A. G. F. Janssen, T. Riedl, S. Hamwi, H. H. Johannes and W. Kowalsky, Highly efficient organic tandem solar cells using an improved connecting architecture, Appl. Phys. Lett., 2007, 91 073519.
76. J. Y. Kim et al., Efficient tandem polymer solar cells fabricated by all-solution processing, Science, 2007, 317, 222.
77. K. Tvingstedt, V. Andersson, F. Zhang and O. Inganas, Folded reflective tandem polymer solar cell doubles efficiency, Appl. Phys. Lett., 2007, 91, 123514.
78. M. C. Scharber et al., Design rules for donors in bulk-heterojunction solar cells - towards 10 % energy-conversion efficiency, Adv. Mater., 2006, 18, 789.
79. B. P. Rand, D. P. Burk and S. R. Forrest, Offset energies at organic semiconductor heterojunctions and their influence on the open-circuit voltage of thin-film solar cells, Phys. Rev. B, 2007, 75, 115327.
80. C. Waldauf, M. C. Scharber, P. Schilinsky, J. A. Hauch and C. J. Brabec, Physics of organic bulk heterojunction devices for photovoltaic applications, J. Appl. Phys., 2006, 99, 104503.
81. D. M. De Leeuw, M. M. J. Simenon, A. R. Brown and R. E. F. Einerhand, Stability of n-type doped conducting polymers and consequences for polymeric microelectronic devices, Synth. Met., 1997, 87, 53.
82. J. H. Zhao, A. H. Wang and M. A. Green, 24.5% efficiency silicon PERT cells on MCZ substrates and 24.7% efficiency PERL cells on FZ substrates, Prog. Photovoltaics, 1999, 7, 471.
83. A. Rohatgi, S. Narasimha, S. Kamra and C. P. Khattak, Fabrication and analysis of record high 18.2% efficient solar cells on multicrystalline silicon material, IEEE Electron Device Lett., 1996, 17, 401.
84. J. Yang, A. Banerjee and S. Guha, Triple-junction amorphous silicon alloy solar cell with 14.6% initial and 13.0% stable conversion efficiencies, Appl. Phys. Lett., 1997, 70, 2975.
85. M. K. Nazeeruddin et al., Engineering of efficient panchromatic sensitizers for nanocrystalline TiO₂-based solar cells, J. Am. Chem. Soc., 2001, 123, 1613.
86. C. J. Brabec, J. A. Hauch, P. Schilinsky and C. Waldauf, Production aspects of organic photovoltaics and their impact on the commercialization of devices, MRS Bull., 2005, 30, 50.
87. A. C. Arias, Vertically segregated polymer blends: Their use in organic electronics, Polym. Rev., 2006, 46, 103.
88. J. S. Huang et al., Investigation of the effects of doping and post-deposition treatments on the conductivity, morphology, and work function of poly (3,4-ethylenedioxythiophene)/poly (styrene sulfonate) films, Adv. Funct. Mater., 2005, 15, 290.
89. W. J. Potscavage, S. Yoo, B. Domercq and B. Kippelen, Encapsulation of pentacene/C₆₀ organic solar cells with Al₂O₃ deposited by atomic layer deposition, Appl. Phys. Lett., 2007, 90, 253511.
90. S. Yoo et al., Integrated organic photovoltaic modules with a scalable voltage output, Appl. Phys. Lett., 2006, 89, 233516.
91. J. van de Lagemaat et al., Organic solar cells with carbon nanotubes replacing In₂O₃:Sn as the transparent electrode, Appl. Phys. Lett., 2006, 88, 233503.
92. M. W. Rowell et al., Organic solar cells with carbon nanotube network electrodes, Appl. Phys. Lett., 2006, 88, 233506.
93. A. D. Pasquier, H. E. Unalan, A. Kanwal, S. Miller and M. Chhowalla, Conducting and transparent single-wall carbon nanotube electrodes for polymer-fullerene solar cells, Appl. Phys. Lett., 2005, 87, 203511.

---