Architectures For Organic Photovoltaics Engineering Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

This first chapter provides a general introduction to semiconducting polymers, the field of OPVs, characterization and components of devices, the market potential and future opportunities. The research chapters delineate the construction of OPVs, starting with spray-coated transparent electrodes (Chapter 2), followed by self-assembled interfacial buffer layers (Chapter 3), and finally, investigation of an approach to control the morphology of photoactive layers (Chapter 4).

Semiconducting Polymers

A critical discovery enabling organic photovoltaics is semiconducting polymers. The 2000 Nobel Prize in Chemistry was awarded to Alan J. Heeger, Alan G. MacDiarmid and Hideki Shirakawa "for the discovery and development of conductive polymers".[1-4] These can conduct electricity while maintaining the mechanical properties and ease of processing of typical insulating polymers.[3]

The most structurally simple conducting polymer is polyacetylene (PA).[5] PA, a first generation semiconducting polymer, consisting of a carbon backbone with alternating single and double bonds as seen in Error: Reference source not foundA. Along the carbon chain σ-bonds hold the polymer together while carbon-carbon π-bonds enable semiconducting behaviour.[6] In trans-polyacetylene there are 4 bands in the band structure as seen in Error: Reference source not foundB. There are σ and σ* bands from the carbon-carbon bond as well as a π and π* bands originating from the overlap of adjacent carbon pz orbitals. The σ and π bands are occupied, while the σ* and π* bands are unoccupied. Thus, a band gap exists between the π (highest occupied molecular orbital [HOMO]) and π* (lowest unoccupied molecular orbital [LUMO]) bands, leading to semiconducting behaviour.[6] PA has limited solubility in organic solvents, limiting its use in solution processed devices.

A second generation semiconducting polymer is regioregular poly(3-hexylthiophene) (P3HT). P3HT consists of a thiophene polymer backbone, with hexyl groups added in improve solubility (Error: Reference source not found). It can have high carrier mobilities in the heat-to-tail arrangement. In films, P3HT forms 2-dimensional lamellae, which are typically oriented normal to the substrate.[7, 8] Alkyl chains interdigitate with one another forming a (100) crystalline plane, while π-π stacking of thiophene rings forms a (010) plane, and facilitate intrachain charge transport. P3HT has a band gap of ~2.0 eV.[9] The absorption of P3HT is dependent on a number of factors including: regioregularity, molecular weight, and processing conditions.[8-10] Well-established syntheses and ease of processing have enabled P3HT to become widely used as a standardized donor material for OPVs.[11]

Third generation semiconducting polymers consist of more complex structures with enhanced control over properties.[6] One class includes donor-acceptor copolymers, which has enhanced control of the band gap, and improved stability.[6] This strategy uses electron-rich and electron-deficient moieties as push-pull copolymers.[12] One example, poly[N-9″-hepta-decanyl-2,7-carbazole-alt-5,5-(4′,7′-di-2-thienyl-2′,1′,3′-benzothiadiazole)] (PCDTBT), consists of electron-rich carbazole and electron-deficient dithienyl-benzothiadiazole (Error: Reference source not found). PCDTBT exhibits remarkable air and thermal stability, the first polymer to combine both of these features.[13] The thermal stability is attributed to a deeper HOMO (-5.5 eV below vacuum), making it resistant to oxidation.[13] PCDTBT has also been used as a donor material in OPV devices with internal quantum efficiency (IQE) approaching 100%. This means that all photons absorbed, result in charges, which are extracted by the electrodes.[14] Continued efforts in designing third generation semiconducting polymers aims to realize high mobility and stability combined with ease of processing.

The typical one-dimensional chemical structure of semiconducting polymers leads to anisotropic carrier mobilities.[15] Charges have higher mobility along the backbone of polymers, compared to interchain charge hopping (carriers transported from one polymer chain to another).[16] The extent of π-π intermolecular interactions influences the interchain carrier transport. Thus, controlling the morphology of polymer films in order to maximize π-π stacking, improves charge carrier mobility. The charge mobilities in polymers can range from 10-6 cm2 V-1 s-1 for amorphous films to > 1 cm2 V-1 s-1 for crystalline morphologies.[17]

Semiconducting polymers can absorb photons, exciting an electron to the π* band and forming a hole in the π band (if the energy of the photons are greater than the band gap of the polymer). However, the electron and hole are not free charges, as is the case of bulk inorganic semiconductors. The electron and hole form a bound exciton. In semiconducting polymers, the exciton binding energy is a few tenths of an electron volt, which is much higher than the millielectron volts for bulk inorganic semiconductors.[17-19] Exciton dissociation can occur when the exciton migrates to an interface with lower chemical potential energy, overcoming the exciton binding energy.[18] This facilitates charge transfer from a semiconducting polymer donor to an acceptor-type material (such as buckminsterfullerene, C60).

Semiconducting polymers can be used for similar applications as their inorganic counterparts. A major application area is photonics, which include: photovoltaics,[20] lasers,[21] light emitting diodes,[22] and photodiodes.[4] Other areas of application include thin film transistors,[23, 24] supercapacitors,[4, 25] chemical and biological sensing.[26] The mobilities of semiconducting polymers are typically inferior to most inorganics (102 - 103 cm2 V-1 s-1).[17] However, polymers allow for solution processing, reducing fabrication costs, and can be used in flexible device applications, enabling flexible electronics such as televisions, cellular phones, and electronic paper.[27]

Introduction to Organic Photovoltaics


The advent of semiconducting polymers has allowed the development of plastic electrical power generation commonly termed organic photovoltaics. Solar energy represents a large, and renewable source of energy. In fact, there is sufficient solar energy hitting the earth's surface in one hour to provide enough energy to the world for an entire year.[28] With increasing world population, and increasing economic wealth of developing nations, there is projected to be an unprecedented demand for energy in the future. Conservative models project that energy consumption will double by 2050, and triple by the end of the century.[28] In order to meet future energy demands, large-scale carbon-neutral energy production solutions are required. Given earth's huge solar resource, technologies that can efficiently and inexpensively produce and/or store energy from the sun are of great interest.[28] Organic photovoltaics offer a potential solution to contribute to the energy demands of the future.

Compared to traditional silicon photovoltaics, OPVs have been projected to achieve lower costs of electricity production, due to highly scalable low-temperature solution processing, amendable to mass manufacturing via roll-to-roll printing on flexible substrates.[29-31] However, a number of significant challenges remain (low efficiencies and lifetimes), which need to be addressed before OPVs will be able to gain significant market share for utility scale energy production.[32, 33]

Development of Organic Photovoltaics

The photovoltaic effect was first observed by A.E. Becquerel in 1839.[34] From there, the photovoltaic effect has been studied in a number of inorganic semiconductors, resulting in a Si p-n junction photovoltaics (first generation),[35] and a number of second generation photovoltaic technologies including: amorphous silicon,[36] CdTe,[37] and CuInGaSe2 (CIGS).[17, 38] C.W. Tang developed an OPV based on a photoactive bilayer structure of copper phthalocyanine (donor) and perylene tetracarboxylic derivative (acceptor), obtaining a power conversion efficiency (PCE) of 1%.[39] Since Tang's milestone single junction OPV device, a number of advancements have led to increasing PCEs to 10% obtained by Mitsubishi Chemical in 2011.[40] These include: synthesis of high molecular weight and purity semiconducting polymers,[41, 42] influence of photoactive layer blend morphology,[43, 44] low band gap polymers,[9, 45] increased stability,[13, 46, 47] and application of roll-to-roll printing.[48-52] OPVs also include small molecule donors, which have achieved efficiencies of 6.7%.[53] However, this thesis will restrict the discussion to polymer-based OPVs.

The field of OPVs has attracted significant research attention over the past decade. Analysis of the Thomson Reuters Web of Science database searching for topics of "organic photovoltaic*" or "polymer solar cell*" or "plastic photovoltaic*" reveals a growing number of publications in the field as seen in Error: Reference source not found. Research intensified about a decade ago and in 2011 there were over 1000 publications in the area. The top five countries in terms of papers published are USA, China, South Korea, Germany, and Japan, with over 100 countries producing a least one publication. This data represents a truly global investment and growing effort in the field of OPVs.

Photovoltaic Device Characterization

Photovoltaic devices are tested under simulated solar radiation using a solar simulator. OPVs are typically characterized under air mass 1.5 global (AM1.5G) conditions.[54, 55] This represents light travelling through 1.5x air mass at a solar zenith angle of 48.2°. This represents the yearly average at mid-latitudes of the earth, and corresponds to an integrated power of 100 mW cm-2.[56] Global solar radiation includes both direct and scattered sunlight. Photovoltaic devices are electrically characterized in dark and under simulated solar light conditions. A source-meter sweeps voltage across the electrodes, while measuring the current. Using this data, a plot similar to Error: Reference source not found can be constructed.

Error: Reference source not found shows typical current-density-voltage (J-V) curves of a photovoltaic cell under dark and simulated solar radiation. Key photovoltaic performance parameters can be extracted from the light J-V curves. The open-circuit voltage (VOC) is the potential across the electrodes under zero current or open-circuit conditions. The short-circuit current density (JSC) is the current density at zero voltage. The maximum power (PMAX) point represents the maximum power point along the J-V curve, whereby P=IV. The current density and potential at the PMAX are referred to the JMAX and VMAX, respectively. The fill factor (FF) is a ratio of the PMAX in the device compared to the theoretical power at the JSC and VOC, and is represented by equation (1.1):


The power conversion efficiency (PCE, η) represent the efficiency of light to current conversion, within the photovoltaic device, as summarized in equation (1.2):


The PLIGHT is 100 mW cm-2 under standard AM1.5G conditions. The FF is controlled by both the series and shunt resistances (RS, RSH). The RS and RSH is the inverse slope of the J-V curve at the VOC and JSC, respectively. Ideal photovoltaics would have an RSH of infinity and an RS of zero, which would confer a FF of 1. These photovoltaic parameters will be discussed in OPV device characterization throughout this thesis.

Components of Organic Photovoltaics

General Device Architecture & Photocurrent Generation

Organic photovoltaics devices consist of a layered structure, as seen in Error: Reference source not found. Devices consist of two electrodes: a transparent electrode, such as indium tin oxide (ITO), and a reflective Al, Ag, or Au electrode. The photoactive layer consists of donor (p-type) and acceptor (n-type) materials. The donor material forms a type-II heterojunction (staggered energy levels) with the acceptor as seen in Error: Reference source not found.

The donor material absorbs photons of light depending on its band gap. The absorption of a photon excites an electron from the HOMO to the LUMO energy level (Ï€-Ï€* transition), leaving a hole at the HOMO level. This forms a bound exciton with a relatively large binding energy of ~0.2-0.4 eV.[57, 58] The exciton can be dissociated at a donor-acceptor interface as seen in Step (2) of Error: Reference source not found. However, excitons in semiconducting polymers typically have short exciton diffusion lengths of ~10 nm.[19] This significantly limits the distance an exciton can travel before recombining, and returning back to the ground state.[17]

After exciton dissociation, the electron and hole can drift and diffuse in the presence of an electric field (from the difference in work function of the two electrodes), depending on the charge carrier mobility of the material towards the interfacial buffer layers.[17, 18, 33] There are usually two interfacial buffer layers, modifying both electrodes, as seen in Error: Reference source not found. Electrons are transported by the cathodic buffer layer, and holes are transported by the anodic interfacial layer [Step (3) in Error: Reference source not found]. Electrons are then collected by the cathode, and holes at the anode, leading to current generation. Each layer of the OPV device will be discussed in more detail below.


OPVs can be fabricated on a range of substrates from glass to plastics, paper, and textiles.[17, 18, 59, 60] Solution processing of subsequent layers allows great versatility in substrates selection. Substrates for OPVs in the device architecture depicted in Error: Reference source not found, must be highly transparent across the solar spectrum, allowing photons to pass to the photoactive layer. Back-illumination OPVs have also been reported, where light would enter the top of the device in Error: Reference source not found.[51, 61, 62] In this case the substrate is not required to be transparent. The transparent electrode is coated on top of the devices, which usually consists of poly(3,4-ethylenedioxythiophene):poly(4-styrenesulfonate) (PEDOT:PSS) and a metal grid.[51, 61, 62] Back-illuminated OPVs are beneficial in module fabrication, as only the last-deposited layer (transparent electrode) is required to be patterned.[51]

To fabricate OPVs on a variety of materials, the surface chemistry of the substrate can be tuned, enhancing the adhesion of subsequent films.[63, 64] One example is the use of polydopamine to form a hydrophillic surface on polydimethylsiloxane (PDMS) substrates, as seen in Error: Reference source not found.[64] This enables great adhesion of spray-deposited Ag nanowires (NWs) forming reversibly stretchable transparent electrodes.[64]

Transparent Electrodes

The next layer in typical OPV device architectures is the transparent electrode. Transparent electrodes have high transmissivity with low sheet resistances, which are often conflicting properties. The most common material used as transparent electrodes in OPVs is indium tin oxide (ITO). The properties of ITO are sensitive to processing conditions, but films typically have low resistivity (10-4 Ω cm), and high transmissivity in the visible region (>80%).[65, 66] ITO is a doped n-type semiconductor with an optical band gap of 3.6 eV.[66-68] ITO can be prepared by a number of vacuum deposition techniques, but is commonly deposited with magnetron reactive sputtering using a  In2O3 target with 10 wt.% SnO2.[69, 70] Electron conduction in ITO arises from oxygen vacancies in the lattice, as well as the substitution of Sn4+ for In3+ providing an electron for conduction.[71] ITO has a work function of 4.3-4.8 eV depending on atomic stoichiometry and surface cleaning treatments.[67, 68] When using ITO as a transparent electrode in optoelectronic devices, an oxygen plasma is used to remove carbon contamination. The plasma also improves hole-injection by increasing the work function and favourable wettability for coating subsequent layers.[67, 72, 73] However, the declining reserves of indium in the earth's crust introduce significant cost variability.[65] One of the unique properties of OPVs is the ability to create flexible modules.[74-76] ITO supported plastic films exhibit limited flexibility, and are prone to cracking with repeated flexing to small radii of curvature and low strain as seen in optical image in Error: Reference source not found.[77-81] The image shows cracks in ITO on polyethylene terephthalate (PET) substrates under 2.5% strain. These cracks are concomitant with a sharp increase in the ITO film resistance.[79]

There has been considerable interest in the literature evaluating alternatives to ITO transparent electrodes. Some of these include nanomaterials (Ag, and Cu nanowires[NWs]),[82, 83] conducting carbon allotropes (carbon nanotubes [CNTs], graphene)[84, 85], and conducting polymers.[86, 87]

Metals (Ag, Cu, Au) are known to be highly conductive, but are also reflective. To circumvent the high reflectivity, thin metal films or nanowire meshes have been used to form highly transparent and low resistance electrodes.[88-94] Ligands can solubilize metal nanowires, enabling scalable solution processing through spray-coating,[64, 95] Meyer rod coating,[82] and inexpensive roll-to-roll coating techniques.[96] Metal nanowire mesh films allow essentially all light to pass through 'holes' in the film, while forming a conductive network. The properties of these NW mesh films are highly dependent on the density of metal NWs, whereby higher density films are conductive but have reduced transmission. Careful tuning is required to form a percolation network, without sacrificing film transmission.[82, 97]

Yi Cui et al. used Ag NWs to coat PET substrates as seen in Error: Reference source not found.[82] Films were deposited with a Meyer rod and had sheet resistances of 20 Ω â-¡-1 with 80% transmissivity, which are in the same range as ITO.[82] Transmittance values are typically reported as specular. However, NW meshes can scatter light and have significantly higher diffuse transmittance. In OPVs diffuse light can be collected by the photoactive layer, making it a useful parameter.

The ligands on metal NWs enable solution processing. But, the organic coatings significantly increase resistance of the films. Ligands prevent conduction across junctions resulting in >109 increase in resistance compared to a single NW.[82] Annealing films at 200 °C is required to remove the ligand coating on glass substrates.[98] However, this temperature is not compatible to processing on plastic substrates. Several strategies have emerged to process Ag NW films on plastic substrates including: galvanic displacement forming gold-coated Ag NWs,[82] applying mechanical pressure to as-deposited films,[98, 99] and plasmonic welding.[100] Plasmonic welding was applied to Ag NW films, which selectively heats up and epitaxially joins NW junctions as seen in Error: Reference source not found.[100] The localized heating does not affect underlying plastic substrates, and is amendable to low-cost roll-to-roll processing.

Carbon nanotubes (CNTs) and graphene have been applied as transparent electrodes for OPV devices.[101-104] Carbon nanotubes have great potential as a transparent electrode, as a result of high mobilities > 105 cm2 V-1 s-1, and low transmittance with thin films.[105] However, it is difficult to obtain high purity, monodisperse CNTs, limiting the commercial applicability of transparent CNT films.[96] Another challenge is dispersing CNTs in solution. Several strategies have been developed including the use of surfactants to solubilize CNTs in water,[106] and the addition of chemical functional groups to improve solubility.[107] Upon formation of a dispersible CNT ink, films can be fabricated using similar methods as used for Ag NW meshes.[96] Marks et al. fabricated CNT transparent electrodes with sheet resistance of 150 Ω â-¡-1 and 80% transmission as seen in Error: Reference source not found.[108] CNT films have a relatively flat transmission spectra, making them appear light grey, with superior colour neutrality compared to ITO.[96] A major limiting factor for CNT transparent electrodes are high sheet resistances. Considering the above CNT film of 150 Ω â-¡-1, the sheet resistance would have to decrease an order of magnitude at 80% transmission to replace ITO. Theoretical conductivity of CNTs have been calculated to be 9 x 104 S cm-1.[109] If this level of conductivity can be achieved, CNT films will beat the performance metrics of ITO, and be an ideal transparent electrode for OPVs.

Conducting polymers have the potential to replace ITO transparent electrodes as a result of better mechanical properties for flexible and stretchable applications,[110, 111] and reduced fabrication costs driven by scalable solution processing.[51] However, conducting polymers are more resistive than ITO. This has motivated the field to look for polymer conductivity enhancements. Two conducting polymers have emerged as candidates for transparent electrodea: polyaniline (PANI) and poly(3,4-ethylenedioxythiophene) (PEDOT). PEDOT is often charge stabilized by poly(4-styrenesulfonate) (PSS) forming a PEDOT:PSS blend (Error: Reference source not found). The monomer (EDOT) can undergo oxidative polymerization in the presence of water soluble PSS. H.C. Stark commercialized PEDOT:PSS, and it is commercially available as a colloidal solution.[96]

PEDOT:PSS has a conductivity < 1 S cm-1 which is considerably too low for transparent electrode application.[112-114] However, a number of additives have been used to significantly increase the conductivity of PEDOT:PSS films including: solvents,[112, 113, 115-117] fluorosurfactants,[86, 118, 119] and ionic liquids.[120, 121] These additives can act as dopants or affect the morphology of PEDOT:PSS films (by removing excess PSS and extending the conjugation length in PEDOT:PSS).[113, 122] Using a dilute H2SO4 treatment led to an enhancement from 0.3 S cm-1 to 3065 S cm-1.[123] This is one of the best conductivities reported for PEDOT:PSS, to date and resulted in sheet resistances of 39 Ω â-¡-1 with 80% transmittance. OPV devices using the high conductivity PEDOT:PSS electrode achieved 87% of the PCE compared to the ITO-based devices. Although the properties of PEDOT:PSS are approaching that of ITO, some have identified stability issues when films of PEDOT:PSS are exposed to air, humidity and UV light.[90, 124-127]

None of the emerging transparent electrode technologies have yet to match the electrical and optical properties of ITO. However, considering the energy used to make the material, and fabricate films, ITO represents 74% of the embedded energy of OPV modules, as seen in Error: Reference source not foundA. This is in contrast to the 7%, 7%, and 10% embedded energy for Ag NWs, CNTs, and PEDOT:PSS, respectively.[128] Using any of these transparent electrodes would have shorter energy payback times. Most of the emerging transparent electrodes discussed herein are lower-cost than ITO, as seen in Error: Reference source not foundB. With the exception of single walled carbon nanotubes (SWCNTs), which represent 69% of the module cost, due to difficulties in obtaining high yield and purity material. PEDOT:PSS is the most inexpensive with a minimum 50x cost reduction per square meter of printed film.[128] Taking the embedded energy, cost, and flexibility into account, PEDOT:PSS and Ag NWs have the potential to be implemented as transparent electrodes in OPV modules. It will not be long before alternatives gain industrial and market acceptance.

Interfacial buffer layers modify OPV electrodes as seen in Error: Reference source not found. These layers have a number of functions, which affect both the photovoltaic performance and stability of OPVs. These functions include: tuning the interfacial energy level alignment between the photoactive layer and the electrode, electron or hole blocking, controlling the surface properties of the subsequent layer (photoactive layer),[124] improving stability,[129] and defining the polarity or charge selectivity of the device.[130]

OPVs can operate in both forward- and inverted-mode as seen in Error: Reference source not found. This defines the electrode where electrons and holes are extracted. In the forward-mode of operation, holes are extracted by the transparent anode and electrons by the reflective top cathode (Error: Reference source not found). The opposite occurs in inverted-mode; electrons are extracted by the transparent cathode and holes by the reflective top anode. The polarity of the device is defined by the work function of the interfacial buffer layers. In forward-mode, ITO is modified with a high work function buffer layer, reducing the hole extraction barrier, enabling preferential hole transport. In inverted-mode, ITO is modified with a low work function material, reducing the electron extraction barrier. By simply raising or lowering the work function of the buffer layer on ITO, the polarity of the device can be switched. The proper top buffer layer and electrode must also be selected to ensure extraction of the opposite charge carrier. For example as seen in Error: Reference source not found, in forward-mode, a lower work function top buffer layer and electrode are selected (LiF/Al), and in inverted-mode, a higher work function top buffer layer and electrode are chosen (V2O5/Ag). A primary advantage of the inverted-mode architecture is significantly improved stability and lifetimes of OPV modules. This is primarily explained by the selection of a higher work function metal (Ag, Au) as the top anode, which improves resistance to oxidation.[131, 132]

One of the most common anodic interfacial layers is poly(3,4-ethylenedioxythiophene):poly(4-styrenesulfonate) (PEDOT:PSS). The chemical structure is presented in Error: Reference source not found. PEDOT:PSS has been used as an anodic modifying layer in organic light emitting diodes (OLEDs) and OPVs. PEDOT:PSS has been shown to improve hole extraction, prevent electron leakage, provide a smoother electrode surface, and improve ohmic contact.[133-137] These attributes have made PEDOT:PSS an ubiquitous anodic buffer layer in both forward- and inverted- mode OPVs. However, PEDOT:PSS has been shown to be a major contributor to the degradation of OPVs.[138] Studies have shown that acidic (pH ~1) PEDOT:PSS solutions can etch ITO,[137, 139] excess poly(sodium 4-styrenesulfonate) (NaPSS) can migrate within the device,[140] and the hygroscopicity of PEDOT:PSS can accelerate the oxidation of the cathode.[141]

Another organic approach is the use of carboxylated polythiophenes as the anodic buffer layer. We developed a technique for the self-assembly of poly[3-(5-carboxypentyl) thiophene-2,5-diyl] (P3CPenT) into nanowires as shown in Error: Reference source not found.[124] The P3CPenT NWs increased the work function and provided great interfacial energy alignment, since the optoelectronic properties of P3HT and P3CPenT are near identical.[124] In addition, films of P3CPenT NWs decreased the interfacial surface energy, allowing a more favourably morphology of the subsequent film: the photoactive layer.[124] These attributes led to an increase in the power conversion efficiency of OPV devices.

Metal oxides have also been extensively investigated as anodic buffer layer for OPVs. Examples include NiO,[137, 142-144] WO3,[145, 146] MoO3,[147-150] IrOx,[151] V2O5,[152] Cu2O.[153] Metal oxides have been deposited from solution via sol-gel chemistry,[148] synthesis of dispersible nanoparticles,[153] acidified aqueous dispersions,[147] metal-organic precursor decomposition,[142, 154] and under vacuum by thermal evaporation.[150, 152] NiO shows particularly beneficial properties as an anodic buffer layer. NiO is highly transparent in thin films, allowing photons to pass to the photoactive layer for harvesting. As seen in Error: Reference source not found, p-type NiO shows great energy level alignment with P3HT to transport holes, while the conduction band is high enough to block electrons.[137] Metal oxides offer tunable electronic properties with doping and stoichiometry, stability towards oxidation, and improved electrical stability as anodic interfacial buffer layers.[149, 154]

Cathodic buffer layers decrease the electron extraction energy barrier, block holes prevent recombination, modify the surface chemistry affecting the next deposited layer, and can stabilize the performance of OPVs.[155] In the forward-mode of operation cathodic buffer layers modify the top reflective electrode (commonly Al), where in inverted-mode these layers modify the transparent electrode. When used in forward-mode of operation this layer also serves to protect the underlying photoactive layer from hot metal atoms deposited by thermal evaporation.[156-158]

A large variety of materials have been investigated as cathodic interfacial materials including: low work function metals (Ca, Ba, Mg),[155, 159] alkali metal compounds (LiF, Cs2CO3, CsF, CsCl),[160-166] metal oxides (TiO2, ZnO),[167-171] and polymers.[172-184] Low work function metals can make ohmic contact between the photoactive layer and the cathode, but they are sensitive to water, degrading OPV performance when stored under ambient conditions.[130, 155, 185] LiF has been extensively used as a cathodic interfacial material in OLEDs and OPVs. Typically a very thin ~0.9 nm layer is thermally evaporated onto devices. This layer forms ohmic contact between the cathode and photoactive, through the formation of a dipole, which lowers the energetic barrier for electron extraction.[164, 186, 187] Low work function metals, LiF, CsF, and CsCl have typically been used as anodic modifiers in the forward-mode of operation for OPVs.

In inverted-mode OPVs Cs2CO3, the metal oxides and polymers have been investigated to modify the transparent electrode (usually ITO). One of the first cathodic interfacial buffer layers for inverted-mode OPVs was Cs2CO3. Yang et al. reported a work function decrease from 4.5 eV to 3.1 eV by modifying ITO with a thin annealed film of Cs2CO3 (Error: Reference source not found), enabling ohmic contact with the photoactive layer.[166] Annealing at 170 °C decomposes Cs2CO3 to form n-doped Cs2O. Cs2O on ITO forms an interfacial dipole as depicted in Error: Reference source not foundB. The dipole moment is directed towards vacuum, and its magnitude is relative to the work function shift.[188] In addition, Cs2O reduces the interfacial surface energy, and leads to improved morphology of the subsequent photoactive layer.[189]

N-type metal oxides such as TiO2 and ZnO are commonly used as cathodic modifiers in both forward- and inverted- modes of OPVs. They are non-toxic and transparent across the visible and NIR spectrum and have suitable conduction band energies to match fullerene acceptors.[130] In addition, their deep valence bands are able to block holes, making them highly electron selective. These materials have also been used as optical spacers in OPVs, helping to redistribute light intensity.[155, 170, 190] One disadvantage of these metal oxides is the requirement for high temperature (>300 °C) annealing to produce crystalline, high mobility films.[130, 191, 192] However, sol-gel solution processing has been developed, decreasing the annealing temperatures to < 200 °C.[167, 192] TiO2 and ZnO can be modified by self-assembled monolayers (SAMs), which can further tune the work function via dipole formation, and can tune the interfacial surface energy. Carboxylic acid-modified fullerenes have been used to form SAMs on ZnO.[193] These SAMs were shown to passivated surface traps states in ZnO, improve electron extraction, and optimize the photoactive layer morphology.[193-195]

Semiconducting organic small molecules and polymers have also been used as cathodic interfacial modifiers, and will be discussed in Chapter 3.

A myriad of cathodic interfacial modifiers have been developed offering a range of tunability in optical and electronic properties, work functions, and interfacial surface energies to match the required properties of the photoactive layer. These layers play a major role in both the photovoltaic performance and stability of OPVs. However, much of the research to date has focused on modifying ITO transparent electrodes. With a number of emerging substitutes to ITO, cathodic interfacial buffer layers on alternative transparent electrodes should be examined.

Photoactive Layers

The photoactive layer is responsible for the absorption of photons, and the formation of charge carriers, which are then transported through the interfacial layers and extracted by the electrodes (Error: Reference source not found). The photoactive layer of OPVs consists of two semiconducting components: a p-type donor and an n-type acceptor. One of the most common donor/acceptor combinations is regioregular poly(3-hexylthiophene) and [6,6]-phenyl-C60-butyric acid methyl ester (PC60BM), respectively. The chemical structures of both are presented in Error: Reference source not found. The donor material is responsible for the majority of the absorption, creating excitons. Excitons can be dissociated at a donor-acceptor interface.[196] The n-type semiconductor accepts electrons, which are then transported to the cathode, while holes are transported through the donor material to the anode. Excitons have a diffusion length of ~10 nm.[19] Meaning that if an exciton does not reach a donor/acceptor interface within 10 nm, the exciton will recombine, and the charge carriers will be lost. This limits the domain sizes within the photoactive layers.

One of the first OPV devices consisted of a bilayer photoactive layer, as schematically depicted in Error: Reference source not foundA.[39] This architecture limits the donor material thickness to ~20 nm, in order to efficiently dissociate excitons. However, such thin films are not able to absorb a significant portion of photons. Thus in bilayer architectures, the exciton dissociation competes with photon absorption, leading to limited photovoltaic performance.[20] Heeger et al. and Friend et al. discovered that by blending the donor and acceptor materials together, it was possible to form a bicontinuous interpenetrating network of both components, termed a bulk heterojunction (BHJ) (Error: Reference source not foundB).[41, 42, 44] The controlled formation of a BHJ led to high interfacial area between donor and acceptor components with a domain size of ~20 nm. This enabled thicker photoactive layers on the order of 100-220 nm, significantly increasing photon absorption.[17]

The photoactive layer is typically deposited by spin-coating, but there are a number of scalable deposition methods being developed including: spray-coating,[95, 197] ink jet printing,[198, 199] and traditional roll-to-roll methods.[48, 50, 61, 200-203] Since solution processing is one of the major advantages of OPVs, the materials selected for the photoactive layer should be highly soluble, allowing scalable processing techniques.

Poly(3-hexylthiophene) is a donor semiconducting polymer, commonly used in the photoactive layer of OPVs.[11] Highly regioregular P3HT can be synthesized via both the McCollough and Rieke routes, as seen in Error: Reference source not found.[204, 205] The McCollough route involves a Kumada coupling to alkylate 3-bromothiophene and subsequent bromination in acetic acid. The one-pot polymerization involves metalation with LDA at the 5-position, conversion with MgBr2*EtO2, and polymerization using a NiCl2dppp catalyst.[205][55] Rieke's method uses activated zinc to form the reactive thiophene precursor, and polymerization using NiCl2dppp at -78 °C.[204]

P3HT has a band gap of ~ 2.0 eV, which defines the portion of the solar spectrum which photons can be absorbed. As seen in Error: Reference source not found the box representing the absorption range of P3HT is fairly limited. Lower energy photons in the NIR are not absorbed by P3HT, limiting the potential photocurrent. A strategy to circumvent this problem is to reduce the band gap of the donor polymer. Recently, the synthesis of low band gap polymers has been the subject of intensive research.[9, 12] Promising low band gap polymers are shown in Error: Reference source not found, and include: poly[N-9″-hepta-decanyl-2,7-carbazole-alt-5,5-(4′,7′-di-2-thienyl-2′,1′,3′-benzothiadiazole)] (PCDTBT),[13, 206] poly({4,8-di(2-ethylhexyloxyl)benzo[1,2-b:4,5-b']dithiophene}-2,6-diyl)-alt-({5-octylthieno[3,4-c]pyrrole-4,6-dione}-1,3-diyl) (PBDTTPD),[207-209] and poly[[4,8-bis[(2-ethylhexyl)oxy]benzo[1,2-b:4,5-b']dithiophene-2,6-diyl][3-fluoro-2-[(2-ethylhexyl)carbonyl]thieno[3,4-b]thiophenediyl]] (PTB7).[183, 210, 211] These polymers have lower band gaps than P3HT, and have all achieved power conversion efficiencies (PCEs) in excess of 7% in single junction OPVs (Error: Reference source not found).

Important factors when selecting a donor polymer for high efficiency OPVs include: band gap, HOMO and LUMO energy levels, good hole mobility, and high solvent solubility. As previously mentioned, the band gap defines the absorption range, which affects the short-circuit current density (JSC). The open-circuit voltage (VOC) is related to the energetic offset between the EDonorHOMO and EPCBMLUMO, as depicted in Error: Reference source not found. Larger energetic offsets will have increased open-circuit voltages as seen in equation (3):


where e represents the elementary charge, and a PCBM LUMO of -4.3 eV.[212] The donor:acceptor LUMO-LUMO energy difference must be greater than 0.3 eV to efficiently dissociate excitons.[212] The low band gap polymers all have larger EPCBMLUMO - EDonorHOMO values than P3HT, which increase their VOC in OPV devices.[212] A VOC of 0.6 V is typical for P3HT:PCBM OPVs, but can be improved to > 0.9 V when using PBDTTPD.[209, 213]

Fullerenes are the most common acceptor used in OPV devices. This is a result of high electron affinity and mobility, and the ability to accept multiple electrons. Fullerene was discovered in 1985, and the Nobel Prize in Chemistry was awarded to Robert Curl, Harold Kroto, and Richard Smalley in 1996 for its discovery.[214] Fullerene is made on a "large" scale by the arc vaporization of graphite, which was developed in 1990.[215]. This has allowed further synthetic modification and functionalization of fullerenes for a range of properties and applications.[216] Fullerene has limited solubility in organic solvents. Organic functional groups can be added to fullerenes to increase solubility, and tune their optoelectronic properties.[216] [6,6]-phenylC60-butyric acid methyl ester (PC60BM) is a functionalized fullerene, soluble in organic solvents, with widespread use in OPV devices. The chemical structure can be seen in Error: Reference source not foundB. More recently [6,6]-phenyl-C70-butyric acid methyl ester (PC70BM) has been used as an acceptors, because it has stronger visible absorption (Error: Reference source not foundA).[217, 218] Both PC60BM and PC70BM are synthesized following the same procedure as seen in Error: Reference source not found.[218, 219] The reaction forms a number of multiadduct products, which are separated with silica gel chromatography.

Although P3HT:PC60BM is the standard photoactive layer for OPVs, this combination exhibits poor band alignment, limiting the VOC.[11] Bisadduct fullerenes such as indene-C60 bisadduct (IC60BA), and bis-[6,6]-phenyl C61-butyric acid methyl ester (bis-PC60BM), can improve the energetic alignment by increasing the LUMO energy level (see Error: Reference source not found and Error: Reference source not found for reference and chemical structures). For example, IC60BA (shown in Error: Reference source not foundB) has a 0.17 eV increase in the LUMO which increases the VOC by 0.2 V in OPV devices PC60BM, as seen in Error: Reference source not found.[220] Bis-PC60BM OPV devices have a similar effect whereby the LUMO energy level is 0.1 eV higher than PC60BM, which is enough to improve the VOC by 0.8 V.[221]

There has been considerably less research on new acceptors compared to the myriad of donor polymers available. In developing the next generation of organic acceptors the following should be considered: light absorption, electron mobility, solubility, and thermal stability.[216]

The morphology of the photoactive layer is critical to maximizing light absorption, exciton dissociation, limiting recombination, and efficient charge transport. Control over the morphology of the BHJ is important in optimizing the performance of OPVs. Loos et al. investigated the effect of different processing conditions on the morphology of the photoactive layer, as characterized with transmission electron microscopy (TEM) (Error: Reference source not found).[222] After spin-coating, the photoactive film has not reached thermodynamic equilibrium.[222] Thus, additional post-processing steps can improve blend morphology. Thermal annealing the deposited film for 20 minutes at 130 °C improves the polymer crystallinity. As seen in Error: Reference source not foundB, crystalline P3HT nanowires (NWs) are formed (white phase contrast).

A similar result is obtained with solvent assisted annealing, which consists of placing wet P3HT:PC60BM films in a closed Petri dish for 3 hours. This allows the ortho-dichlorobenzene solvent to slowly evaporate from the films. From Error: Reference source not foundC, we once again see crystalline P3HT NWs. OPV devices almost doubled in power conversion efficiency for both thermal annealing and solvent assisted annealing treatments. Both the JSC and the fill factor improved. The JSC increase is attributed to enhanced hole mobility in the crystalline P3HT NWs. The higher FF indicates improved morphology: better percolation pathways and fewer trap sites.[222] Further discussion on controlling and stabilizing the BHJ morphology of OPV devices can be found in Chapter 4.

Low band gap donor and fullerene acceptor materials with high mobilities and appropriate offsetting of energy levels maximize the short-circuit current density and open-circuit voltage. The photoactive layer should have high interfacial area between the donor and acceptor with ideal morphology consisting of domain sizes of ~20 nm, while providing bicontinuous pathways of charge extraction which results in higher power conversion efficiencies.

Reflective Electrodes

Reflective electrodes are the top contact in OPVs as depicted in Error: Reference source not found. Ideal top electrodes make ohmic contact with the photoactive layer,[223] and are reflective. This allows light which has not been absorbed by the photoactive layer to be reflected for a second opportunity for absorption. Higher work function metals such as Ag, and Au (anode) are used for inverted-mode, while a lower work function metal like Al (cathode) is used for forward-mode OPVs. Top electrodes are typically deposited via thermal evaporation of the metal under vacuum.[224] However, some solution processes have been developed. These include: slot-die coating,[61] screen printing,[225] and spin-coating PEDOT:PSS,[226] and spray-coating PEDOT:PSS.[227] In fact, Jen et al. uses PEDOT:PSS as both the transparent and top electrode.[227] This opens the potential for all-organic OPVs, however, PEDOT:PSS has poor reflectively limiting the photovoltaic performance.[227]

Tandem Organic Photovoltaics

Tandem OPVs consist of two photoactive layers stacked on top of each other (Error: Reference source not found). Requirements for high efficiency polymer tandem photovoltaics include: two donors with complementary absorption, efficient recombination layer, and orthogonal solubilities of adjacent layers.[228] When two donor polymers are selecting with different absorption ranges, a larger portion of the solar radiation can be absorbed. The recombination layer is located between the two photoactive layers. The purpose of this layer is to efficiently collect electrons from one sub-cell and holes from the other photoactive layers. This layer should form ohmic contact with each of the sub-cells.[228] Orthogonal solubilities of adjacent layers are required to prevent the dissolution of previously cast layers. Although challenging for a minimum of 6 layers, solution processed polymer tandem photovoltaics have been fabricated, obtaining 6% efficiency.[169, 229]

The serial connection of two photoactive layers leads to a sum of individual open-circuit voltages. While the short-circuit current density is limited to the sub-cell with the higher fill factor.[230, 231] The JSC is dependent on the recombination efficiency in the interlayer. If one of the photoactive layers is providing an excess of charge carriers, they will not be able to be recombined. This results in overall JSC losses.[230]

A high efficiency polymer tandem OPV was reported by Yang et al. in 2012.[232] This device combines P3HT with the low band gap polymer: poly{2,6'-4,8-di(5-ethylhexylthienyl)benzo[1,2-b;3,4-b]dithiophene-alt-5-dibutyloctyl-3,6-bis(5-bromothiophen-2-yl)pyrrolo[3,4-c]pyrrole-1,4-dione} (PBDTT-DPP) (Error: Reference source not foundA). PBDTTT-DPP has a band gap of 1.44 eV, enabling complementary absorption to P3HT as seen in Error: Reference source not foundB.[232] The inverted-mode tandem OPV consisted of a P3HT:ICBA bottom cell and PBDTT-DPP:PC70BM as the top photoactive layer. A certified tandem devices obtained a power conversion efficiency of 8.62%.[232] Although multi-junction OPVs are more difficult to fabricate with solution processing, and have higher material costs. Theory predicts that optimized tandem OPVs have the potential to obtain almost 15% efficiency, which would present a strong case for commercialization.[233]

The ability to print OPVs on lightweight, flexible substrates, opens up new potential products and markets, which traditional photovoltaic technology cannot compete.[29, 30] The ability to tune the colour of OPVs and printing on fabrics adds aesthetic appeal and a closer integration of consumers with their power sources.[59, 234] Flexible and mechanically robust OPVs may find application in portable consumer electronics, reducing or potentially eliminating the need to charged devices.[59] Other applications include off-grid power, outdoor recreation, lightweight power for the military, building integrated photovoltaics, and the potential for lowlight and indirect power applications.[59]

Krebs et al. projects a 500 million $US market by 2018 with the main applications in consumer electronics and residential building applications.[32] This is backed by an industrial market report by IDTechEx, which projects a 630 million $US market by 2022.[235] The projected market segments of OPVs is presented in Error: Reference source not found. Consumer electronics, and products designed for developing countries are the first major sectors for OPVs. One example of an OPV integrated product for developing countries is an OPV powered LED lamp used in the "Lighting Africa" initiative.[236] Mature applications for OPVs include residential, commercial, industrial, military and emergency applications.[32, 235] In Nanomarkets assessment of the OPV market, they state that a big breakthrough in reduction of costs, increase in efficiency and lifetime, is required.[237]

OPVs have the potential to produce cost effective electrical power from solar radiation.[41, 51, 61, 239, 240] To encourage widespread commercial acceptance, flexible OPVs should be efficient, mechanically robust, lightweight, and fabricated with techniques scalable for mass manufacturing.[48, 51] A combination of all of these factors has yet to be achieved, limiting the commercialization of the technology. The scalability of OPVs, has yet to overcome limitations in efficiencies and lifetimes compared to other photovoltaic systems.

Theory predicts that, single junction organic photovoltaics may be able to achieve power conversion efficiencies of 10-20%.[212, 241] In addition, a study has demonstrated projected lifetimes of 7 years,[47] with OPV modules encapsulated in glass. While work in improving the synthesis of high molecular weight semiconducting polymers aims to provide economical and environmentally friendly coupling polymerizations.[242] There is considerable fundamental research still occurring in this field, which are leading to daily reports of new materials and architectures for OPVs. If modules of 10% and lifetimes in excess of 10 years are achieved, there will be considerable applications for this technology.[30, 32]

Organization of Thesis

This thesis presents an integrative approach to improving the mechanical properties and stability of high efficiency organic photovoltaics with the use of new semiconducting polymers and device architectures. The thesis is organized to 'build-up' OPV devices starting with the transparent electrode.

Transparent conductors have utility in a number of applications including electrodes for OPVs. A scalable spray-coating method for PEDOT:PSS transparent electrodes on glass and plastic substrates is developed in Chapter 2. This fabrication method leads to low sheet resistance films with high transparency and have superior mechanical properties compared to ITO. Chapter 2 was reproduced in part with permission from: a) J.G. Tait, B.J. Worfolk, S.A. Maloney, T.C. Hauger, A.L. Elias, J.M. Buriak, K.D. Harris, Sol. Energy Mater. Sol. Cells. 2013, 4, xxxx-xxxy. Copyright © 2013 Elsevier.

A new architecture for cathodic interfacial buffer layers is introduced in Chapter 3. Using water-soluble polymers, stable and high efficiency inverted-mode organic photovoltaics are achieved. The chapter discusses stability and degradation issues of OPVs, as well as electrostatic layer-by-layer assembly, which is used to fabricate cathodic interfacial films. Chapter 3 was reproduced in part with permission from: a) B.J. Worfolk, T.C. Hauger, K.D. Harris, D.A. Rider, J.A.M. Fordyce, S. Beaupré, M. Leclerc, J.M. Buriak, Adv. Energy Mater. 2012, 2, 361-368. Copyright © 2012 Wiley-VCH Verlag GmbH & Co. b) D.A. Rider, B.J. Worfolk, K.D. Harris, A. Lalany, K. Shahbazi, M.D. Fleischauer, M.J. Brett, J.M. Buriak, Adv. Funct. Mater. 2010, 20, 2404-2415. Copyright © 2010 Wiley-VCH Verlag GmbH & Co. c) Q. Chen, B.J. Worfolk, T.C. Hauger, U. Al-Atar, K.D. Harris, J.M. Buriak, ACS Appl. Mater. Interfaces 2011, 3, 3962-3970. Copyright © 2011 American Chemical Society.

In Chapter 4, carboxylated polythiophenes are introduced into the bulk heterojunction photoactive layer of OPVs. This enables morphological control of the photoactive layer through the use of hydrogen bonding. The effect of hydrogen bonding on the mechanical properties of the photoactive layer is discussed. Strategies for controlling and stabilizing the morphology, as well as the use of hydrogen bonding in the photoactive layer are outlined. Chapter 4 was reproduced in part with permission from: a) B.J. Worfolk, D.A. Rider, A.L. Elias, M. Thomas, K.D. Harris, J.M. Buriak, Adv. Funct. Mater. 2011, 21, 1816-1826. Copyright © 2011 Wiley-VCH Verlag GmbH & Co. b) B.J. Worfolk, W. Li, P. Li, T.C. Hauger, K.D. Harris, J.M. Buriak, J. Mater. Chem. 2012, 22, 11354-11363. Copyright © 2012 The Royal Society of Chemistry.

Chapter 5 summarizes the research in each chapter, and discusses future research directions.