Organic donor-acceptor systems Serbenta, Almis
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date:
2016
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Serbenta, A. (2016). Organic donor-acceptor systems: Charge generation and morphology. University of Groningen.
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
The publication may also be distributed here under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license.
More information can be found on the University of Groningen website: https://www.rug.nl/library/open-access/self-archiving-pure/taverne- amendment.
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
Download date: 15-10-2022
1
Chapter 1
Introduction
1. Abstract
Organic materials for optoelectronic applications have gained substantial interest both in the academia and industry due to their cheap solution processing, lightweight, and chemical tunability of various optoelectronic properties. This unique combination of features opens new possibilities in the area of light emitting and solar power harvesting technologies.
This Thesis discusses a number of fundamental processes which occur in organic semiconductor materials and devices, with a focus on those materials, which are designed and used in organic solar cell applications. This introductory Chapter presents the basic operating principles of organic solar cells, and introduces concepts used throughout the Thesis, such as bound electron-hole pairs (excitons) and their diffusion, the strongly bound electron-hole state formed at a donor-acceptor interface (interfacial charge transfer state), and nanostructured phase separation between donor and acceptor materials in the blends (morphology). Next, concepts of the core experimental method, non-linear spectroscopy, used throughout this Thesis, are discussed. Finally, the goals of this Thesis are formulated and a short overview of the Chapters is provided.
1.1. Organic solar cells and their principles of operation
Interest in electronic and optical properties of organic semiconducting materials began in the last century (around 1940s)1 when it was proposed that charges may be conducted between molecules. With the development of novel organic materials, the field of molecular electronics has emerged. Organic materials have gained attention both in academia and industry, and have become one of the hot topics in research and development in the past few decades2. The interest in organic semiconductors has arisen due to their potential for cheap mass-production via all solution processing and tunability of their optoelectronic properties via chemical engineering. One of the main advantages of organic materials as compared to their inorganic counterparts is the spectral tunability that allows maximizing solar power harvesting in solar cells or selecting the desired emission color for light-emitting diodes.
Another important advantage of the organic semiconducting materials is their low-weight and flexibility, which promotes a variety of new applications, for instance, integration into
2
clothing, bags, etc. These types of materials also have the potential to become completely biocompatible and environmentally friendly, meaning that all materials would be recyclable and not harmful for the environment, e.g. fully based on water/alcohol-soluble materials3.
Optoelectronic properties of organic materials can be tuned by designing molecules with desirable energy levels of molecular orbitals. In addition, the molecular arrangement and/or packing order strongly influence the overlap of molecular orbitals, thereby tuning the polarizability, transport properties, and stiffness. The molecular orbitals and their overlap play a crucial role in determining the spectral position of optical absorption and emission, as well as the efficiency of generation and recombination of charges, and the energy and the charge transfer processes. The spectral positions of absorption and emission are important for photon-to-charge conversion in solar cells, or the reverse, charge-to-photon, conversion processes in light emitting devices. On the other hand, the broad range of tunable parameters such as absorption, emission, charge generation, charge recombination, and charge transport introduces complexity when trying to optimize all of them simultaneously for a particular application, e.g. solar cells or light emitters.The complexity of the problem sets a huge challenge toward achieving the desired properties, but it also opens numerous opportunities for the development of novel optoelectronic devices.
This Thesis focuses on the organic materials designed for plastic solar cells (also called organic photovoltaics); however, many of the fundamental processes are generic to all organic semiconductors. The perspective to create competitive low-cost renewable energy sources was one of the main drivers behind the research of organic solar cells (OSC) in academia and industry4-8. One of the major factors that keep the energy costs of OSCs high is the low power conversion efficiency (PCE)4-7, 9-12. Fundamental understanding of the physical processes is the first step towards getting new ideas on how to change the approach to design solar cells that are more efficient. This Chapter is dedicated to (a) an explanation of the basic operating principles of the organic solar cells, (b) the most popular molecular design approaches of solar cell production, (c) nano-scale molecular structures (packing) of active layer in organic solar cells, (d) nonlinear ultrafast spectroscopy methods to investigate photophysical processes. Finally, an overview of the Thesis will be presented.
1.1.1. Operating principles
The conversion of photons into current involves a sequence of photophysical and physical processes: 1) light absorption, 2) diffusion of the photoexcitation (exciton) to the
3 interface, 3) charge generation at the interface by exciton dissociation, 4) charge transport towards electrodes, and 5) extraction of charges at the electrodes (see references4-8, 13-20
for more information about organic photovoltaics). This Thesis focuses on the first three stages of operation and does not discuss the charge transport towards electrodes and extraction, despite the obvious relevance of the two last steps for the overall solar cell efficiency.
The molecular orbitals that play a central role in the operation of OSCs are the π orbitals forming delocalized electron clouds over π-conjugated segments (fig. 1.1) of the molecules. The frontier molecular orbitals play a major role in energy and charge transfer processes. The highest occupied molecular orbital (HOMO) is a ground state occupied π orbital. The lowest unoccupied molecular orbital (LUMO) of the π-conjugated segment (called π*) becomes semi-occupied when the molecule is excited to the lowest excited state (e.g. after absorption of a photon by a material).
Fig. 1.1 Example of π-conjugation on the trans-polyacetylene polymer. The overlap of the atomic orbitals (orbitals in conjugation, i.e. on adjacent carbon atoms) form delocalized molecular π and π* orbitals.
1.1.2. Absorption of light and conditions for separation of electron-hole pairs When a material absorbs light, an electron occupying the HOMO is excited to the higher molecular orbitals, i.e. LUMO, leaving behind a hole in the HOMO. A state diagram can be drawn for processes such as absorption of light as shown in fig. 1.2. In the organic materials, this electron-hole pair (called a Frenkel exciton21) has a strong binding energy of the order 0.3-0.5 eV22-24, much higher than the room temperature energy kBT~25 meV. There are two types of excitons: the net spin 0, called singlet exciton (primary optical excitation), and the net spin 1 excitation called triplet exciton created after intersystem crossing25 (note that in the literature singlet exciton levels are usually abbreviated as S0, S1, S2 etc. while triplet levels are marked as T1, T2, T3, etc.).
4
Fig. 1.2 A Jablonski diagram representing energy levels of typical states corresponding to the ground state (S0), the singlet (S1, S2, ... Sn) and triplet (T1, T2, ... Tn) excitons and the transitions between these levels.
Electron spins are represented as arrows pointing up or down. Due to an intersystem crossing, the electron spin can flip and the singlet exciton becomes a triplet. Both singlet and triplet excitons can be further excited to higher energy levels by absorbing photons with energy lower than the bandgap of the material. hν is the minimal energy of a photon that can be absorbed, hν' is the energy of a photon exceeding the absorption bandgap, hν" is the energy of a photon below the absorption bandgap which is not absorbed. Excess excitation energy is quickly (<100 fs) lost by relaxation.
In a situation where the photon energy exceeds the HOMO-LUMO gap (hν’ fig. 1.2), the exciton initially has excess energy that is quickly dissipated into molecular reorganization. Naturally, if the HOMO-LUMO gap would be tuned to absorb as many photons emitted by the Sun (at the surface of Earth) as possible then large losses due to the excited electron relaxation to the LUMO would occur. In contrast, a too high energy of the absorption edge would result in most solar photons not being absorbed at all (hν” fig. 1.2).
Therefore, the absorption efficiency – (defined as the number of photons absorbed per number of incident photons) – should be tuned by designing materials with a good spectral overlap of their absorption spectrum with the solar spectrum. The theoretical limit of the overall PCE for an ideal single bandgap (~1.4 eV) solar cell was estimated by Shockley and Queisser to be slightly above 30%26 for the Sun spectrum at the surface of Earth.
Due to the strong binding energy of excitons in organic materials, efficient solar cells are made using at least two materials: electron donor, and electron acceptor molecules27, 28. The optimum bandgap of a solar cell composed of two materials depends on the particular selection of donor and acceptor materials, and on the alignment of HOMO-LUMO energy levels of the two molecules. So far, successful materials for organic solar cells have been mostly based on polymers or small molecules (donors) combined with various fullerene
5 derivatives (acceptors)8, 17, 20, 29, 30
. This is due to the unique property of fullerene derivatives to be extremely efficient in the charge separation processes as well as reasonably good charge transport properties. It was shown that one of the fullerene derivatives – the PC61BM – can be beneficial also for perovskite-based solution processable solar cells31. The most popular fullerene derivatives are soluble PC61BM and its successor PC71BM, which has stronger absorption in the visible (vide infra).
The most popular fullerene derivatives for OSCs, based on C60 and C70, have very similar energies of HOMO and LUMO32-36, hence, these levels can be tuned only for donor materials. The optimum bandgap for donor materials – conjugated polymers and small molecules – used in organic solar cells was estimated to be around ~1.5 eV37, 38. More photons can be absorbed by increasing the film thickness (in practice typically a few hundred nanometers). However, thicker films lead to fewer charges being able to reach the electrodes due to the charge recombination processes30, 39-41.
Over the past two decades, scientists have learned the way to shift the absorption spectral range of organic materials (polymers and small molecules) towards the infrared (IR) spectral region in order to collect IR photons of the solar spectrum without evoking significant losses in other aspects. As an example, fig. 1.3 shows the comparison between P3HT (regiorandom poly(3-hexylthiophene-2,5-diyl)) and PCPDTBT (Poly[2,1,3- benzothiadiazole-4,7-diyl[4,4-bis(2-ethylhexyl)-4H-cyclopenta[2,1-b:3,4-b']dithiophene-2,6- diyl]]) polymers. On the other hand, fullerenes and their soluble derivatives like PC61BM have relatively high symmetry (as compared to many organic donor materials used in organic solar cells), and therefore, transitions in the visible spectral range are forbidden by symmetry.
It has been shown that reducing the symmetry of these materials leads to an enhancement of the absorption cross section in the visible range (see fig. 1.3 comparing PC61BM ([6,6]- Phenyl C61 butyric acid methyl ester) with PC71BM ([6,6]-Phenyl C71 butyric acid methyl ester)).
6
500 1000 1500
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 PC61BM
Solar irradiation (arb. u.) Absorbance (arb. u.)
Wavelength (nm) P3HT
PCPDTBT PC71BM
Solar irradiation AM 1.5
4.0 3.5 3.0 2.5 2.0 1.5 1.0
Energy (eV)
Fig. 1.3 Solar spectrum42 (green line) and example absorption spectra of popular organic materials used in solar cells: donor polymers P3HT and PCPDTBT, acceptor fullerene derivatives PC61BM and PC71BM.
1.1.3. Exciton diffusion
After an exciton is created by the light absorption, it has to reach the interface between the two materials by diffusion and subsequently dissociate into charges. The exciton diffusion length, L, is related to the diffusion coefficient D and exciton lifetime τ as43, 44:
, (1.1)
where Z is the spatial dimensionality. Equation 1.1 is a general description of diffusion for random walking particles that connects an average diffusion length of excitons with their lifetime. Typical diffusion lengths of excitons in organic materials are very short, 5-15 nm45-
55, with some exceptions56, 57.
Exciton diffusion occurs by energy transfer from one molecule to another. Energy transfer may occur via dipole interaction between adjacent molecules or by exchange of electrons. Diffusion of excitons through dipole-dipole interaction can be described using the Förster resonant energy transfer (FRET) model58. On the other hand, diffusion by exchange of electrons is usually modeled using Dexter theory59. In practice, exciton diffusion in an inhomogeneous energy landscape is often simulated by the Monte Carlo approach. The energy transfer rate is assumed to be a thermally-activated process with an initial transfer rate modified by Boltzmann factor60, 61:
, (1.2)
7 where ki→j is the transfer rate from the i-th to the j-th molecule, kEnT is the energy transfer rate from a site with higher energy to a site with lower energy (no thermal activation required), Ei
and Ej are the energies of excitons on these sites. Equation 1.2 represents that when excitons have to "hop" from a site with lower energy to a site with higher energy, the energy transfer rate decreases. During the exciton diffusion process, the majority of excitons go ‘downhill’
(in energy) to the lowest available energies within the density of states (DOS). Consequently, the hopping rate decreases with time, rendering the diffusion coefficient D time-dependent and decreasing with time. After the majority of excitons reach the lowest energy levels of the DOS, the diffusion reaches a quasi-equilibrium, where D becomes a constant value within a certain time window. Note that the Boltzmann factor can be also used in traditional FRET and Dexter-based models to account for the energy-disordered landscape of exciton diffusion43, 61.
1.1.4. Exciton dissociation into charges: Charge transfer
The electron donating and accepting materials are designed to efficiently dissociate excitons. Different energy levels of molecular orbitals of the electron donor and the acceptor materials create the energy level offset. The energy offset acts as a driving force for exciton dissociation and as an energy barrier (ΔE1 or ΔE2 in fig. 1.4a) for an electron to return to the donor or for a hole to move to acceptor materials. The energy offset needs to be larger than exciton binding energy, which is only one of the requirements for efficient exciton dissociation.
Figure 1.4 represents a typical example of HOMO-LUMO energy level alignment of donor-acceptor materials used in organic photovoltaics. The LUMO level of the acceptor is situated lower than that of the donor, which allows electron transfer from the excited donor LUMO to the acceptor LUMO, but not vice versa. Analogously, when the acceptor material is excited, a hole appears in the original HOMO. Therefore the former HOMO becomes a lower singly-occupied or semi-occupied orbital (SOMO) and the former LUMO becomes a higher SOMO. The lower SOMO of an electron acceptor is positioned lower than the HOMO of an electron donor allowing a hole transfer from this SOMO to the HOMO. The hole transfer process can be viewed as an electron transfer from the electron donor HOMO to the electron acceptor lower SOMO. The hole transfer process is the focus of Chapter 2 and is used in the modeling of Chapter 3.
8
Fig. 1.4 Schematic representation of electron (shown as blue circles with spins pointing up or down) transfer from the electron donor to the acceptor materials between the LUMOs (a) and HOMOs (b). The electron transfer between the two HOMOs is called hole transfer because from the perspective of a hole quasiparticle it is transferred from electron acceptor to the donor material. Absorption of light (black arrow) photoexcites the electron into a higher energy level and subsequently the electron or hole transfer occurs. The energy barrier between donor and acceptor materials is either LUMO-LUMO (ΔE1) or HOMO-HOMO (ΔE2).
When charge transfer occurs, the uncompensated charges induce geometrical reorganization of atoms in both the donor and the acceptor molecules62. Due to the geometric relaxation of the molecules, the initial HOMO and LUMO energy positions, shown in fig.
1.5a, shift up and down respectively as depicted in figure 1.5b63. This deformation, bound to the uncompensated charge inducing it, is called a polaron. The HOMO, which was shifted in energy due to uncompensated charge, contains only a single electron and, therefore, is also called SOMO.
Optimized solar cells can have very fast (<100 fs) and efficient exciton dissociation via charge transfer. For instance, polymer blends with fullerene derivatives have been demonstrated to exhibit electron and hole transfer times of ~30-40 fs64-66. Theoretical estimates also support the ultrafast character of the charge transfer process67. Such extremely fast transfer rates indicate that the process occurs without any barrier or via efficient electron tunneling. Due to the complex molecular structure and geometrical reorganization after charge transfer in films of organic photovoltaic blends, an accurate theoretical calculation of the charge transfer rate (or speed) is very challenging. In addition, it has been argued43, 65 that resonant energy transfer may also occur instead of the charge transfer process.
b ) a
)
a) b)
9
Fig. 1.5 Schematic representation of energy levels of the molecular orbitals of donor and acceptor materials in a neutral state (a) and in the presence of charges (b). The LUMO+1 and LUMO+2 represent molecular orbitals, which have higher energy than LUMO. When the electron is withdrawn by the acceptor, molecular orbitals of the donor are shifted due to the geometrical relaxation, and additional transitions become available in the energy gap(b). LE and HE stand for low energy and high energy charge-induced (polaronic) transitions, respectively. SOMO is an energy level of a semi-occupied molecular orbital.
1.1.5. Excited states and Charge recombination
The dissociated excitons are not free charges, but instead, they form bound charge pairs, called geminate charge pairs68-70. These geminate charge pairs are called charge- transfer (CT) states when the electron resides on the electron-acceptor and the hole remains on the electron-donor material at the donor-acceptor interface. Coulomb’s law tells that the electron and the hole of the geminate pair attract each other and may recombine. The force of attraction is inversely proportional to the distance between opposite charges and the dielectric constant (material property, screening of charges). As soon as the distance becomes significant enough that the binding energy becomes comparable to the room temperature energy ~25 meV, the probability of geminate recombination becomes low and charges are considered no longer bound. These charges form charge separated (CS) states.
A schematic representation of typical excited states in donor-acceptor materials is summarized in fig. 1.6. Frenkel (singlet, fig. 1.6a) excitons are formed either on single molecules (localized exciton) or on several adjacent molecules, for example nanocrystals (delocalized exciton). In all cases, the central positions of the electron and hole distributions
a) b)
10
coincide. CT excitons (fig. 1.6b), on the other hand, form either on a single molecule (i.e. on electron donating-accepting units of the same molecule, called push-pull), or between two molecules of a same type, or may even form at an interface of donor-acceptor molecules for a short time. For CT excitons, the central positions of the electron and hole distributions are displaced. In contrast to CT excitons, CT states (fig. 1.6c) are typically formed on donor-acceptor molecules situated next to each other with a very weak, if any, wavefunction spatial overlap of electrons and holes due to energy barrier between the two materials. Non- bound charges (electrons and holes in fig. 1.6d) are separated by at least few molecules;
therefore, they are called charge separated (CS) states.
Fig. 1.6 Schematic representation of typical excited states in donor-acceptor materials. Green zigzag lines represent donor polymer, black spherical shapes represent acceptor (e.g. PC61BM or PC71BM), gray circles represent the (delocalized) wavefunction of a Frenkel exciton, red and blue circles represent (delocalized) displaced wavefunctions of holes and electrons, respectively
Charges may also recombine by meeting opposite non-geminate charges (the so-called bimolecular recombination). Bimolecular recombination, unlike geminate recombination, is dependent on the charge density70, which is initially proportional to absorbed photon density.
The more charges are generated, the higher probability for them to meet the opposite charge and recombine bimolecularly. Under Sun-driven operational conditions of solar cells, the photon density is low, and therefore, the density of generated charges is relatively low as compared to a majority of experiments involving ultrafast lasers as a light source. Under ambient light conditions, bimolecular recombination is more likely to occur not by a random encounter of opposite charges but via the assistance of traps and/or dead-ends where no
a)
b)
c)
d)
11 pathways are provided towards electrodes (see next Section). The longer the trapped charges stay, the higher becomes the probability of recombining with an opposite charge.
1.2. The nanostructural design of donor-acceptor materials
1.2.1. Bulk heterojunction: the concept and the challenges
The short exciton diffusion length scale (5-15 nm) in organic materials sets a useful upper limit to the thickness of a bilayer structure of donor and acceptor solar cells (fig. 1.7).
However, the absorption of layers as thin as 10-30 nm is too low to make a solar cell efficient. This is why the bilayer donor-acceptor organic solar cells were not successful28. To solve the problem, the solution processing method was developed28 where the two solutions of donor and acceptor materials are mixed together in a single solution and deposited on a substrate. Such preparation method results in self-organized structures of donor- and acceptor-enriched areas inside the whole bulk of the film (fig. 1.8). This mixed nanostructure is called bulk heterojunction (BHJ)71. Self-organization may (or may not) lead to finely intermixed donor-acceptor materials. Mixing of the two materials on the nano-scale (when self-organization is optimal) provides a large donor-acceptor interface. The large interface area has to be balanced by sufficient pathways for charge transport towards electrodes.
Meeting both requirements simultaneously is not a trivial task. So far the mixed solution deposition method is the most extensively used due to its simplicity (not expensive) and high power conversion efficiencies achievable, exceeding 10%12.
Fig. 1.7 Representation of a bilayer donor/acceptor architecture. The short diffusion length of excitons (5-15 nm) limits the thickness of the active layers to be well below the optical absorption length, and hence its total efficiency.
12
Fig. 1.8 Nanostructured donor-acceptor bulk heterojunction (BHJ). The donor and acceptor materials self- organize into semi-random structures with a characteristic phase separation scale.
Fig. 1.8 shows a schematic representation of the BHJ nanostructure after deposition of the mixed solution on a substrate. The research field that studies such nanostructures is called morphology, from the Ancient Greek μορφή (morphé) which means a form and λόγος (lógos) which means a study or a research. The BHJ morphology is one of the key parameters determining the efficiencies of both charge generation and collection. Charge generation is affected by morphology due to limited exciton diffusion length, whereas charge collection efficiency depends on the available percolation pathways for charges towards the electrodes72. Optimization of the bulk heterojunction morphology is not a trivial task 73 because of the lack of active control and the need to balance between the relative interface area and trapless pathways for charge transport. The difficulty of morphology optimization is the price that needs to be paid for cheap all-solution processing. Carefully chosen parameters in the film preparation procedure can lead to an optimized bulk-heterojunction, which is efficient for both the charge generation and transport.
Each particular combination of donor-acceptor materials self-organizes into a particular morphology because of the differences in various physical and chemical properties. An ideal morphology is depicted in fig. 1.928. However, such morphology has not yet been achieved.
One of the major issues trying to achieve an ideal morphology is that simple solution processing is not sufficient, and more complicated methods need to be used (e.g. mechanical nano-imprinting74), which sacrifice the major advantage of simple solution processing: the low price. Moreover, the donor and the acceptor materials are highly inter-miscible not only during the solution drying process but also after film deposition75, 76, which revokes (at least partially) the gain obtained by using complex preparation methods. Making predictions of
13 self-organized structures by using a theoretical model is an extremely challenging and complex matter. Such predictions have been successfully achieved only qualitatively for some types of donor-acceptor materials77, 78. The significance of the BHJ morphology for the PCE of solar cells, as well as the difficulties of morphology prediction, calls for a good quantitative experimental characterization of the BHJ nanomorphology aiming to optimize the PCE.
Fig. 1.9 Ideal nanostructure of donor-acceptor materials. Columns should have a width of twice the diffusion length providing a maximized exciton dissociation yield. The interleaved structure provides ideal pathways for charge transport towards electrodes.
Numerous new polymers and small molecules synthesized every year with better optoelectronic properties than the previous ones8, 17, 20, 29, 30
put forward the necessity to optimize the morphology with different parameters and/or new methods. The material properties often lead to different molecular packing and formation of different nanostructures.
The number of interplaying parameters, e.g. optical, electronic, and nanostructural, which need to be balanced simultaneously, make the development of more efficient organic solar cells very challenging.
1.2.2. Optimization of nanostructures
Since the invention of the BHJ, the research field for studying nanoscale structures in blended organic films has been developing. The aim is to gain insight into the nature of the operation of organic solar cells and to understand what improvements or optimizations lead to a higher PCE. It was found that variation of solvents, the molecular weight of the donor polymer, and its regioregularity (if applicable) strongly affect the self-organization of BHJ nanostructures79. More complications arise in some polymers due to their tendency to crystallize leading to a morphology containing nanoscale crystalline regions mixed with completely amorphous regions. Nanocrystals improve the exciton diffusion length but when
14
the nanocrystals become too large, the solar cell suffers losses due to excitons unable to reach an interface where they can dissociate80, 81. Moreover, nanocrystals can mediate charge transport82 but when nanocrystals are too small they act as the boundaries or traps for charges, which reduces the overall amount of collected charges83.
Various techniques were developed aiming to influence the self-organization towards a more optimal morphology. Post-treatment with thermal or solvent assisted annealing was found to be advantageous for some types of blends, e.g. RRe-P3HT:PCBM79. Annealing these blends improves crystallization of the polymer and optimizes the phase separation.
Usage of solvent additives (e.g. 1,8-octanedithiol, ODT) was advantageous for other types of blends, e.g. PCPDTBT:PCBM, because it leads to a more balanced solubility of donor and acceptor materials79, 84, 85
. It was found that the morphology can be influenced and even stabilized by attachment of long side groups and/or cross-linking units to the donor molecules79. Due to this stabilization, the morphology remains unchanged for an extended period of time at elevated temperatures, e.g. at environmental conditions frequently encountered in the daily operation of solar cells. Since profound theories for the formation of the morphology are not available, the research field of BHJ morphology is mostly based on the “know-how” approach, which is developed by trial and error methods rather than through a systematic theoretical approach.
1.2.3. Characterization of nanostructures
Various techniques for the characterization of nanoscale structures are available, some of which have been developed over the past two decades. The most widely used techniques are based on scattering/diffraction/reflection/absorption/transmission of X-rays/neutrons/electrons72. In addition, scanning probe methods like atomic force microscope (AFM) as well as optical microscopies72 are commonly used techniques. Optical microscopy and AFM techniques cannot resolve nanostructures of organic materials on length scales below 200 nm and 10 nm, respectively. Other techniques pose huge challenges or limitations: for instance, they are applicable to specific combinations of materials or require removal of one of the electrodes or preparation of very thin free standing films separately. The factors that limit the spatial resolution depend on the particular method.
Contrary to the inorganic materials, the spatial resolution is often limited by the relatively poor contrast between the constituent materials. In addition, the organic blends are quite susceptible to damage due to overexposure (e.g. electron microscopy). One of the main
15 drawbacks of these microscopy techniques is that they cannot be applied directly for operational devices.
1.3. Nonlinear ultrafast spectroscopy
1.3.1. Linear optical spectroscopy – a starting point for nonlinear spectroscopy The fundamental processes involved in the operation of organic photovoltaics (OPV) can be investigated using the interaction of light with matter. For instance, the linear absorption spectrum contains information about this interaction of light with matter in a thermal equilibrium state. In other words, linear spectroscopy is a useful tool to investigate the available transitions of a material from the ground state to an excited state. For instance, the absorption edge indicates the optical bandgap between the HOMO and LUMO. The presence of several absorption peaks represents vibronic transitions or transitions to higher energy levels, such as from HOMO to LUMO+1. Materials that can form nanocrystals have an additional shoulder of absorption at low energies around the bandgap due to the HOMO-LUMO energy splitting38. Moreover, the so-called ground state charge transfer complexes (CTCs) can be identified by the renormalization of the HOMO and LUMO levels of the whole system86-88. However, for accessing the changes in optical properties induced by light absorption, one should move to nonlinear spectroscopy.
1.3.2. Nonlinear spectroscopy
A nonlinear spectroscopy is capable of disclosing the state of the material when it is not in the equilibrium state (ground state). Amid the popular nonlinear spectroscopy methods are photoluminescence measurements, which are based on the light absorption by a material and subsequent luminescence of the photoexcited state. Another group of popular measurements is based on the photoexcitation of the material and subsequent detection of the change of transmission of light, e.g. pump-probe spectroscopy. In the following sections, a brief overlook will be provided on the light-induced processes and their spectroscopic signatures.
1.3.3. Singlet excitons and their spectroscopic signatures
Typical primary photoexcitations in organic materials for solar cells are singlet excitons that have a net spin of 0 (fig. 1.2). Singlet excitons may be detected by emission if the quantum yield is high enough. The detection of emission is a popular method48-52 to characterize pristine donor and acceptor materials separately and in some cases their BHJs.
Another signature of singlet excitons is their absorption from the lowest excited state to
16
higher energy levels, e.g. S1 to S2, which has lower energy than the bandgap of the material (fig. 1.2, example spectrum is shown in fig. 1.10). Singlet excitons can also be detected by the decreased probability of the S0 to S1 transitions due to the depletion of the ground state (called Ground State Bleaching, GSB).
Fig. 1.10 Example of typical IR PIA spectra of organic photovoltaic materials.
The formation of a BHJ is usually detrimental for the emission yield due to an efficient exciton dissociation rendering a very low signal to noise (e.g.<< 1) in photoluminescence measurements. The drawback of detecting transitions from S1 to the higher energy levels for spectroscopy is that these energies often overlap with other available transitions. For instance, charge induced transitions89 often have some overlap with singlet transitions. The GSB is present when the material is not in the ground state, therefore, it cannot be attributed to singlet excitons alone.
1.3.4. Charges: polarons and their spectroscopic signatures
The LE and HE charge-induced (polaronic) transitions have already been introduced in Section 1.1.4 (and fig. 1.5). However, LE and HE transitions are only a simplified representation. The shifted molecular orbitals form intrinsically broad photoinduced absorption bands (PIA) because electronic transitions are mixed with vibrational levels (example spectra are shown in fig. 1.10). These multiple charge-induced (polaronic)
LE HE
Exciton
Charge (Polaron)
CT state Charge (Polaron)
17 transitions forming PIA bands are called the LE and HE polaron bands. Due to the very broad nature of polaron bands, the HE polaron band usually overlaps with PIA of excitonic transitions, whereas LE polaron band often has overlap with PIA of transitions from CT states.
1.3.5. Spectral positions of excited states in typical materials for organic photovoltaics
Typical absorption spectra of organic donor materials for photovoltaics span the range of visible to near-infrared (NIR). Emission by singlets is red-shifted with respect to absorption due to the vibrational relaxation. Further absorption by the excitons to their higher energy levels occurs at spectral energies lower than the original bandgap of the material.
These energies often overlap with the emission spectra. The charge-induced (polaron- induced) HE absorption is also below the bandgap of the material, usually in the NIR region.
The LE charge-induced absorption, on the other hand, is situated more in the middle-IR (Mid-IR) range. Therefore, the LE band is advantageous for selective detection of charges.
However, in some combinations of materials transitions related to CT states (or CT excitons) may overlap with the LE polaron absorption86-88, 90,91
(see example spectra in fig. 1.10), which requires spectral decomposition. In principle, CT states and CS states (separated charges) should have slightly different absorption spectra due to the difference of molecular geometry or deformation of the molecules to the different extent. However, discrimination between them may not be a trivial task due to the broad spectral response of polaron bands. In contrast, when LE transitions have no overlap with the transitions related to CT states, the data analysis becomes more straightforward as compared to, for instance, probing of the HE polaron band.
1.3.6. Concept of the pump-probe technique
In the presence of any excited state, be it singlet or triplet excitons or charges, all of them induce additional optical transitions in a spectral energy range below the bandgap energy. Simultaneously, the ground state is bleached and emission from the excited state may occur. One of the most popular methods to follow the time evolution of different excited states is the so-called pump-probe technique. Time-resolved pump-probe spectroscopy is capable of addressing the transient phenomena.
The basic concept of TR pump-probe experiments is that one laser pulse, called excitation or pump pulse, creates a population of excited states while a second, time-delayed
18
probe pulse detects the changes induced by the pump pulse (fig. 1.11). Changing the delay between the two pulses allows monitoring the dynamics of excited states. The shorter the laser pulses, the faster the processes that can be monitored. The time resolution of the processes that can be resolved is limited by the combination of the signal-to-noise ratio and the temporal width of the convolution of the pump and probe pulses, defined as the full width at half maximum (FWHM) or standard deviation σ of the time profile of the measured response, where for Gaussian profiles, typical for laser pulses.
Fig. 1.11 Schematic representation of a pump-probe setup. The pump pulse induces changes in the transmission ΔT as imposed by the chopper, which are detected by the time-delayed probe pulse.
1.3.7. Polarization-sensitive detection: photoinduced transient anisotropy
Another useful approach based on the pump-probe scheme is to monitor the polarization of the transition dipole moment of excited states. This technique allows getting information that is otherwise hidden in the simple pump-probe technique, where only the population of excited states is monitored.
The concept of polarization sensitive technique is that correlations are measured between the two transient dipole moments of the transition to the excited state induced by excitation pulse and from the excited state induced by the probe pulse (fig. 1.12). This concept is realized in the following way: the pump pulse that provides a photoexcitation is linearly polarized in the vertical direction (fig. 1.12). The probe pulse has its polarization rotated by 45 ° with respect to the pump pulse. After the probe pulse passes the sample, it is split into two replicas. Polarizers are placed in front of the detectors in order to select the components of both replicas that are parallel and perpendicular to the pump pulse. The
τ 2
τ 1
Chopper Pump
Probe
Delay (t)
Sample Detector
19 correlation of the two transient dipole moments is defined as the photoinduced transient anisotropy92:
, (1.4)
where and are the transmission changes that are probed with parallel and perpendicular components, respectively. For disordered organic materials r(t) varies from 0.4 to 092.
Fig. 1.12 Schematic representation of the polarization sensitive pump-probe detection method (a).
Polarizations of the pump (blue wave) and the probe (red wave) pulses are tilted with respect to each other by 45 °. Two polarization components of the probe pulse (parallel and perpendicular) are detected after having passed the sample (b). The transient anisotropy is calculated from the two components of probe pulse according to Eq. 1.4 (c). Anisotropy contains information about the photoinduced polarizations in the sample, which is schematically depicted in the zoomed in sample image (green wiggles are polymer chains, blue-red ellipses are excitons).
The localized excited states typically exhibit a direction of transition dipole moments upon excitation that fully correlates with the direction of the probed transition dipole moment. The anisotropy value of such a combination is high and may reach a maximum value of 0.492 assuming that the decrease of population of excited states does not depend on vibration state. In contrast to the localized states, the delocalized states typically exhibit low anisotropy values86, 87, 93, 85
. When the time delay between the pump and probe pulses increases, the excitation may be transferred to other molecules either through exciton or
b ) a
)
a) b)
20
charge diffusion. During this hopping process, the correlation of the direction of the new transition dipole moment with the initial one is gradually lost, which is reflected in a decrease of the anisotropy. Another scenario for the correlation loss is a geometrical relaxation of the molecule. The hopping process60 and the geometrical relaxation94, 95 occur on a timescale from ps to hundreds of ps.
Anisotropy has been shown as an indispensable parameter to investigate the electron vs.
hole transfer processes66. In the case of electron transfer from the donor polymer, the direction of the induced transient dipole moment of the LE charge-induced (polaronic) transition does correlate with the direction of the dipole moment of the excited transition.
Therefore, the anisotropy value is essentially nonzero (although it might not reach the 0.4 value due to some angle between directions of the two dipole moments). In contrast, in the case of hole transfer from e.g. PC61BM acceptor to the polymer, the direction of the dipole moment of the excited transition at the PC61BM molecule does not correlate with the direction of the dipole moment of the LE charge-induced (polaronic) transition because of random orientations between the PC61BM and polymer molecules. Therefore, the anisotropy value is zero. This creates a perfect contrast parameter to distinguish between the electron and hole transfer processes that will be used in Chapter 2.
1.3.8. Utilization of ultrafast spectroscopy to reveal nanoscale structures
The photophysical processes and the particular morphology of a BHJ are closely related as is explained in Chapter 3. For instance, for a very coarse phase separation, exceeding the exciton diffusion length (e.g.>>10 nm) only interface excitons would be harvested leading to a very few charges being generated. At the other extreme, donor-acceptor molecules are mixed at the molecular level causing exciton dissociation immediately after photoexcitation.
In contrast to these two cases, if the molecules form phase separated domains small enough for excitons to reach the interface, exciton dissociation is delayed by exciton diffusion.
Therefore, studying exciton diffusion from polymer and/or fullerene domains in BHJ provides valuable information on BHJ morphology and even the characteristic size of phase separation.
Different techniques can be applied to study exciton diffusion44, 73, 96-98
. The most common technique is the time-resolved photoluminescence quenching. Molecular systems, where primary photoexcited states are singlet excitons, usually exhibit a reasonably high photoluminescence yield. As soon as singlet excitons reach the interface with another
21 material, exciton quenching, and/or dissociation occurs causing quenching of the photoluminescence. Careful modeling allows extracting the exciton diffusion related parameters. Triplet exciton diffusion in molecular materials was visualized experimentally by monitoring the time-dependent and spatially resolved photoluminescence in tetracene99. However, visualization of exciton diffusion after deconvolution was limited to 30 ps time resolution and 250 nm spatial resolution. Exciton diffusion is usually resolved by modeling statistically averaged experimental data73. These methods, based on statistics are not affected by diffraction limit and the ultimate time resolution depends on the laser pulse duration (down to sub-ps) (see Chapter 3 of this Thesis).
Westenhoff et al.60 have demonstrated that a pump-probe setup can be utilized for morphology characterization, e.g. monitoring exciton diffusion in a polymer-polymer blend with pump-probe spectroscopy can be used to get insights of the characteristic size of nanoscale structures. In contrast to Westenhoff et al., we selectively photoexcited the PC71BM acceptor and probed the LE transitions in donor polymers for the appearance of charges (see Chapter 3 of this Thesis). The appearance of charges is the direct outcome of the dissociation of excitons. Characteristic sizes of the PC71BM domains are derived by observation of the exciton diffusion dynamics and the dissociation yield. The method is especially effective with low bandgap donor polymers or small molecules that have absorption shifted towards the NIR region, clearly separated from UV-Vis absorption spectra of PC61BM or PC71BM acceptor.
As is shown in this Thesis, nonlinear ultrafast spectroscopy provides a complementary method to study morphology. It also features a number of important advantages: 1) possibility to investigate complete devices on-the-fly (non-invasively); 2) possibility to extract information of nanostructure sizes below 10 nm. In contrast to some microscopy techniques (e.g. AFM or transmission electron microscopy), nonlinear ultrafast spectroscopy data do not produce direct images but rather give an average view of the morphology.
1.4. Overview of the Thesis
Understanding the fundamental processes in organic solar cells is crucial for further design and improvement of PCE. Multiple parameters have to be tuned simultaneously to find an optimal balance between the efficient generation of excitons, their dissociation to charges and subsequent charge collection. These parameters strongly depend not only on particular materials and their molecular orbitals but also on the interaction between the
22
selected donor-acceptor materials and the overall morphology. The aim of this Thesis is to shed light on the capability of optical spectroscopy to reveal the PC71BM exciton dissociation to charges time-scale, the characteristic PC71BM cluster size relevant for understanding morphology and to study the excited states (e.g. excitons, charges) of organic solar cells.
Following this introductory Chapter, Chapter 2 discusses the hole transfer dynamics measured by implementing ultrafast pump-probe spectroscopy. The method relies on selective excitation of PC71BM at energies below the polymer bandgap and selective detection of the appearance of charges on the polymer by probing the LE polaron band. The experiments were performed for three different polymers (RRa-P3HT, MDMO-PPV, and RRe-P3HT) to highlight the universal character of the ultrafast (30-200 fs) hole transfer time.
Different weight ratios between the polymer and PC71BM allowed to investigate the influence of PC71BM aggregation on the hole transfer process. The hole transfer time was found to be slightly faster, e.g. ~30 fs, when most PC71BM molecules were isolated islands within the polymer matrix as compared to the slower 200 fs when PC71BM molecules formed aggregates. The RRe-P3HT:PC71BM blend was the only one that exhibited direct excitation of the CT states; however, the disappearance of RRe-P3HT nanocrystals, when blended with high loads of PC71BM, resulted in a vanishing contribution of the CT states to the transients.
Chapter 3 extends the method of studying hole transfer described in Chapter 2 by revealing the bulk heterojunction morphology of polymer:fullerene blends. This method allows extracting information on the fullerene cluster size. An extensive modeling of dissociated excitons using Monte Carlo simulations revealed the typical size of small PC71BM aggregates (≤ 7 nm). We have also demonstrated that a simple approach considering surface exciton dissociation (within the first ps) relative to the total amount of dissociated excitons is sufficient to determine the approximate size of PC71BM aggregates. Significant exciton losses were observed in some polymer:PC71BM blends due to the formation of large (> 15 nm) PC71BM domains. Significant differences of the BHJ morphology were observed for the three polymers studied: RRa-P3HT, MDMO-PPV, and RRe-P3HT.
Chapter 4 discusses the time-resolved photoinduced absorption spectra of polymers, PC71BM, and their blends in the energy range of 0.1-1 eV (or 1-10 µm wavelength). The following donor polymers were investigated: RRa-P3HT, PCDTBT, PTB7, PCPDTBT PBDTTPD, and BTT-DPP. The spectra were captured at 3 and 400 ps after photoexcitation for a better discrimination of different excited states. The signatures of excitons, CT excitons,
23 CT states, and charges were observed. Each polymer and its blend with PC71BM exhibited unique spectral features of photoinduced absorption. Chapter 4 serves as a useful reference for other research based on (time-resolved) pump-probe spectroscopy.
References
1. N. S. Hush, Ann. N.Y. Acad. Sci. 1006, 1-20 (2003).
2. S. R. Forrest, Nature 428 (6986), 911-918 (2004).
3. C. Duan, K. Zhang, C. Zhong, F. Huanga and Y. Cao, Chem. Soc. Rev. 42, 9071- 9104 (2013).
4. M. C. Scharber and N. S. Sariciftci, Prog. Polym. Sci. 38, 1929-1940 (2013).
5. J. Ahmad, K. Bazaka, L. J. Anderson, R. D. White and M. V. Jacob, Renew. Sustain.
Energy Rev. 27, 104–117 (2013).
6. Y.-W. Su, S.-C. Lan and K.-H. Wei, Materials Today 15 (12), 554-562 (2012).
7. C. J. Brabec, S. Gowrisanker, J. J. M. Halls, D. Laird, S. Jia and S. P. Williams, Adv.
Mater. 22, 3839–3856 (2010).
8. W. Cao and J. Xue, Energy Environ. Sci. 7, 2123–2144 (2014).
9. M. A. Green, K. Emery, Y. Hishikawa, W. Warta and E. Dunlop, D., Prog. Photovolt:
Res. Appl. 23 (1), 1-9 (2015).
10. W. Nie, H. Tsai, R. Asadpour, J.-C. Blancon, A. J. Neukirch, G. Gupta, J. J. Crochet, M. Chhowalla, S. Tretiak, M. A. Alam, H.-L. Wang and A. D. Mohite, Science 347, 522-525 (2015).
11. N. J. Jeon, J. H. Noh, W. S. Yang, Y. C. Kim, S. Ryu, J. Seo and S. I. Seok, Nature 517, 476–480 (2015).
12. http://www.nrel.gov/ncpv/.
13. M. Pope and C. E. Swenberg, Electronic Processes in Organic Crystals and Polymers. (Oxford University Press New York Oxford, 1999).
14. S.-S. Sun and N. S. Sariciftci, Organic Photovoltaics: Mechanisms, Materials, and Devices. (CRC Press, 2005).
15. D. F. J. Kavulak, Structure-Function Relationships in Semiconducting Polymers for Organic Photovoltaics. (UC Berkeley Electronic Theses and Dissertations, 2010).
16. O. Ostroverkhova, Handbook of organic materials for optical and (opto)electronic devices : properties and applications.
17. L. Dou, J. You, Z. Hong, Z. Xu, G. Li, R. A. Street and Y. Yang, Adv. Mater. 25, 6642–6671 (2013).
18. O. A. Abdulrazzaq, V. Saini, S. Bourdo, E. Dervishi and A. S. Biris, Part. Sci.
Technol. 31, 427–442 (2013).
19. M. Wright and A. Uddin, Sol. Energy Mater. Sol. Cells 107, 87–111 (2012).
20. J. Yu, Y. Zheng and J. Huang, Polymers 6, 2473-2509 (2014).
21. A. Moliton and R. C. Hiorns, Polym. Int. 53 (10), 1397-1412 (2004).
22. M. Scheidler, U. Lemmer, R. Kersting, S. Karg, W. Riess, B. Cleve, R. F. Mahrt, H.
Kurz, H. Bässler, E. O. Göbel and P. Thomas, Phys. Rev. B 54, 5536 (1996).
23. A. Ruini, M. J. Caldas, G. Bussi and E. Molinari, Phys. Rev. Lett. 88, 206403 (2002).
24. J.-L. Brédas, D. Beljonne, V. Coropceanu and J. Cornil, Chem. Rev. 104 (11), 4971–
5004 (2004).
25. PAC, 68, 2223 (Glossary of terms used in photochemistry (IUPAC Recommendations 1996)) on page 2248.
24
26. W. Shockley and H. J. Queisser, J. Appl. Phys. 32 (3), 510-519 (1961).
27. G. Yu, J. Gao, J. C. Hummelen, F. Wudl and A. J. Heeger, Science 270 (5243), 1789- 1791 (1995).
28. S. Gunes, H. Neugebauer and N. S. Sariciftci, Chem. Rev. 107 (4), 1324-1338 (2007).
29. H.-C. Liao, C.-C. Ho, C.-Y. Chang, M.-H. Jao, S. B. Darling and W.-F. Su, Materials Today 16 (9), 326–336 (2013).
30. Y. Liu, J. Zhao, Z. Li, C. Mu, W. Ma, H. Hu, K. Jiang, H. Lin, H. Ade and H. Yan, Nat. Commun. 5, 5293 (2014).
31. J. Seo, S. Park, Y. C. Kim, N. J. Jeon, J. H. Noh, S. C. Yoon and S. I. Seok, Energ. &
Environ. Sci. 7, 2642-2646 (2014).
32. N. S. S. a. J. C. H. C. J. Brabec, Adv. Funct. Mater. 11 (1), 15-26 (2001).
33. S. Sensfuss, M. Al-Ibrahim, A. Konkin, G. Nazmutdinova, U. Zhokhavets, G.
Gobsch, D. A. M. Egbe, E. Klemm and H.-K. Roth, Proc. SPIE 5215, 129 (2004).
34. M. Al-Ibrahim, H.-K. Roth, U. Zhokhavets, G. Gobsch and S. Sensfuss, Sol. Energ.
Mater. Sol. Cells 85 (1), 13-20 (2005).
35. J. Y. Kim, K. Lee, N. E. Coates, D. Moses, T.-Q. Nguyen, M. Dante and A. J. Heeger, Science 317, 222-225 (2007).
36. M. K. Siddiki, J. Li, D. Galipeaua and Q. Qiao, Energy Environ. Sci. 3, 867-883 (2010).
37. M. C. Scharber, D. Wuhlbacher, M. Koppe, P. Denk, C. Waldauf, A. J. Heeger and C.
L. Brabec, Adv. Mater. 18 (6), 789-794 (2006).
38. G. Dennler, M. C. Scharber and C. J. Brabec, Adv Mater 21 (13), 1323-1338 (2009).
39. A. C. Stuart, J. R. Tumbleston, H. Zhou, W. Li, S. Liu, H. Ade and W. You, J. Am.
Chem. Soc. 135 (5), 1806–1815 (2013).
40. W. Ma, J. R. Tumbleston, M. Wang, E. Gann, F. Huang and H. Ade, Adv. Energy Mater. 3 (7), 864–872 (2013).
41. S. Albrecht, J. R. Tumbleston, S. Janietz, I. Dumsch, S. Allard, U. Scherf, H. Ade and D. Neher, J. Phys. Chem. Lett. 5 (7), 1131–1138 (2014).
42. R. E. Bird and R. L. Hulstrom, Solar Energy 30 (6), 563-573 (1983).
43. K. Feron, W. J. Belcher, C. J. Fell and P. C. Dastoor, Int. J. Mol. Sci. 13 (12), 17019- 17047 (2012).
44. O. V. Mikhnenko, H. Azimi, M. Scharber, M. Morana, P. W. M. Blom and M. A. Loi, Energ. Environ. Sci. 5 (5), 6960-6965 (2012).
45. A. Haugeneder, M. Neges, C. Kallinger, W. Spirkl, U. Lemmer, J. Feldmann, U.
Scherf, E. Harth, A. Gugel and K. Mullen, Phys. Rev. B. 59 (23), 15346-15351 (1999).
46. L. A. A. Pettersson, L. S. Roman and O. Inganas, J. Appl. Phys. 86 (1), 487-496 (1999).
47. M. Theander, A. Yartsev, D. Zigmantas, V. Sundstrom, W. Mammo, M. R.
Andersson and O. Inganas, Phys. Rev. B. 61 (19), 12957-12963 (2000).
48. D. E. Markov, E. Amsterdam, P. W. M. Blom, A. B. Sieval and J. C. Hummelen, J.
Phys. Chem. A. 109 (24), 5266-5274 (2005).
49. Y. Wu, Y. C. Zhou, H. R. Wu, Y. Q. Zhan, J. Zhou, S. T. Zhang, J. M. Zhao, Z. J.
Wang, X. M. Ding and X. Y. Hou, Appl. Phys. Lett. 87 (4), 154105 (2005).
50. Y. C. Zhou, Y. Wu, L. L. Ma, J. Zhou, X. M. Ding and X. Y. Hou, J. Appl. Phys. 100 (2), 023712 (2006).
51. C. Goh, S. R. Scully and M. D. McGehee, J. Appl. Phys. 101 (11), 114503 (2007).
52. O. V. Mikhnenko, F. Cordella, A. B. Sieval, J. C. Hummelen, P. W. M. Blom and M.
A. Loi, J. Phys. Chem. B. 112 (37), 11601-11604 (2008).
25 53. F. Monestier, J. J. Simon, P. Torchio, L. Escoubas, B. Ratier, W. Hojeij, B. Lucas, A.
Moliton, M. Cathelinaud, C. Defranoux and F. Flory, Appl. Optics. 47 (13), C251- C256 (2008).
54. S. Cook, A. Furube, R. Katoh and L. Y. Han, Chem. Phys. Lett. 478 (1-3), 33-36 (2009).
55. M. C. Fravventura, J. Hwang, J. W. A. Suijkerbuijk, P. Erk, L. D. A. Siebbeles and T.
J. Savenije, J. Phys. Chem. Lett. 3 (17), 2367-2373 (2012).
56. R. C. Powell and Z. G. Soos, J. Lumin. 11, 1-45 (1975).
57. B. Verreet, P. Heremans, A. Stesmans and B. P. Rand, Adv. Mater. 25, 5504–5507 (2013).
58. T. Förster, Discuss. Faraday Soc. (27), 7-17 (1959).
59. D. L. Dexter, J. Chem. Phys. 21 (5), 836-850 (1953).
60. S. Westenhoff, I. A. Howard and R. H. Friend, Phys. Rev. Lett. 101 (1), 016102 (2008).
61. K. Feron, X. Zhou, W. J. Belcher and P. C. Dastoor, J. Appl. Phys. 111, 044510 (2012).
62. J. L. Bredas, J. E. Norton, J. Cornil and V. Coropceanu, Accounts Chem. Res. 42 (11), 1691-1699 (2009).
63. Z. V. Vardeny, Ultrafast Dynamics and Laser Action of Organic Semiconductors.
(CRC Press, 2009).
64. C. J. Brabec, G. Zerza, G. Cerullo, S. De Silvestri, S. Luzzati, J. C. Hummelen and S.
Sariciftci, Chem. Phys. Lett. 340 (3-4), 232-236 (2001).
65. S. Gelinas, A. Rao, A. Kumar, S. L. Smith, A. W. Chin, J. Clark, T. S. van der Poll, G. C. Bazan and R. H. Friend, Science 343 (6170), 512-516 (2014).
66. A. A. Bakulin, J. C. Hummelen, M. S. Pshenichnikov and P. H. M. van Loosdrecht, Adv. Funct. Mater. 20 (10), 1653-1660 (2010).
67. S. M. Falke, C. A. Rozzi, D. Brida, M. Maiuri, M. Amato, E. Sommer, A. D. Sio, A.
Rubio, G. Cerullo, E. Molinari and C. Lienau, Science 344, 1001-1005 (2014).
68. A. Pivrikas, N. S. Sariciftci, G. Juška and R. Österbacka, Prog. Photovoltaics 15 (8), 677–696 (2007).
69. S. R. Cowan, N. Banerji, W. L. Leong and A. J. Heeger, Adv. Funct. Mater. 22 (6), 1116–1128 (2012).
70. G. Lakhwani, A. Rao and R. H. Friend, Annu. Rev. Phys. Chem. 65, 557-581 (2014).
71. I. Etxebarria, J. Ajuria and R. Pacios, J. Photon. Energy 5, 057214 (2015).
72. W. Chen, M. P. Nikiforov and S. B. Darling, Energ. Environ. Sci. 5 (8), 8045-8074 (2012).
73. F. S. U. Fischer, D. Trefz, J. Back, N. Kayunkid, B. Tornow, S. Albrecht, K. G.
Yager, G. Singh, A. Karim, D. Neher, M. Brinkmann and S. Ludwigs, Adv. Mater.
27, 1223–1228 (2014).
74. J. E. Slota, X. M. He and W. T. S. Huck, Nano Today 5 (3), 231-242 (2010).
75. B. A. Collins, J. R. Tumbleston and H. Ade, J. Phys. Chem. Lett. 2 (24), 3135-3145 (2011).
76. N. D. Treat, M. A. Brady, G. Smith, M. F. Toney, E. J. Kramer, C. J. Hawker and M.
L. Chabinyc, Adv. Energy Mater. 1 (1), 82-89 (2011).
77. J. M. Y. Carrillo, R. Kumar, M. Goswami, B. G. Sumpter and W. M. Brown, Phys.
Chem. Chem. Phys. 15 (41), 17873-17882 (2013).
78. S. Kouijzer, J. J. Michels, M. van den Berg, V. S. Gevaerts, M. Turbiez, M. M. Wienk and R. A. J. Janssen, J. Am. Chem. Soc. 135 (32), 12057-12067 (2013).
79. M. T. Dang, L. Hirsch, G. Wantz and J. D. Wuest, Chem. Rev. 113 (5), 3734-3765 (2013).
26
80. Y. Tamai, H. Ohkita, H. Benten and S. Ito, J. Phys. Chem. Lett. 6, 3417−3428 (2015).
81. Y. Tamai, K. Tsuda, H. Ohkita, H. Benten and S. Ito, Phys.Chem.Chem.Phys. 16, 20338-20346 (2014).
82. V. D. Mihailetchi, H. X. Xie, B. de Boer, L. J. A. Koster and P. W. M. Blom, Adv.
Funct. Mater. 16 (5), 699–708 (2006).
83. C. Menelaou, S. Tierney, N. Blouin, W. Mitchell, P. Tiwana, I. McKerracher, C.
Jagadish, M. Carrasco and L. M. Herz, J. Phys. Chem. C 118, 17351−17361 (2014).
84. F. Etzold, I. A. Howard, N. Forler, D. M. Cho, M. Meister, H. Mangold, J. Shu, M. R.
Hansen, K. Mullen and F. Laquai, J. Am. Chem. Soc. 134 (25), 10569-10583 (2012).
85. Y. Gu, C. Wang and T. P. Russell, Adv. Energy. Mater. 2 (6), 683-690 (2012).
86. A. A. Bakulin, D. S. Martyanov, D. Y. Paraschuk, M. S. Pshenichnikov and P. H. M.
van Loosdrecht, J. Phys. Chem. B. 112 (44), 13730-13737 (2008).
87. A. A. Bakulin, D. Martyanov, D. Y. Paraschuk, P. H. M. van Loosdrecht and M. S.
Pshenichnikov, Chem. Phys. Lett. 482 (1-3), 99-104 (2009).
88. A. A. Bakulin, S. A. Zapunidy, M. S. Pshenichnikov, P. H. M. van Loosdrecht and D.
Y. Paraschuk, Phys. Chem. Chem. Phys. 11 (33), 7324-7330 (2009).
89. R. Österbacka, C. P. An, X. M. Jiang and Z. V. Vardeny, Science 287 (5454), 839- 842 (2000).
90. T. Drori, J. Holt and Z. V. Vardeny, Phys. Rev. B. 82 (7), 075207 (2010).
91. K. Vandewal, A. Gadisa, W. D. Oosterbaan, S. Bertho, F. Banishoeib, I. Van Severen, L. Lutsen, T. J. Cleij, D. Vanderzande and J. V. Manca, Adv. Funct. Mater. 18 (14), 2064-2070 (2008).
92. R. G. Gordon, J. Chem. Phys. 45, 1643-1648 (1966).
93. S. Permogorov, A. Naumov, C.Gourdon and P.Lavallard, Solid State Commun. 74 (10), 1057-1061 (1990).
94. J. Clark, T. Nelson, S. Tretiak, G. Cirmi and G. Lanzani, Nat. Phys. 8 (3), 225-231 (2012).
95. A. A. Bakulin, A. Rao, V. G. Pavelyev, P. H. M. van Loosdrecht, M. S.
Pshenichnikov, D. Niedzialek, J. Cornil, D. Beljonne and R. H. Friend, Science 335 (6074), 1340-1344 (2012).
96. O. V. Mikhnenko, R. Ruiter, P. W. M. Blom and M. A. Loi, Phys. Rev. Lett. 108, 137401 (2012).
97. O. V. Mikhnenko, M. Kuik, J. Lin, N. v. d. Kaap, T.-Q. Nguyen and P. W. M. Blom, Adv. Mater. 26, 1912–1917 (2014).
98. O. V. Mikhnenko, P. W. M. Blom and T.-Q. Nguyen, Energy Environ. Sci. 8, 1867 (2015).
99. G. M. Akselrod, P. B. Deotare, N. J. Thompson, J. Lee, W. A. Tisdale, M. A. Baldo, V. M. Menon and V. Bulovic, Nat. Commun. 5, 3646 (2014).
