Probing the effects of photodegradation

of acceptor materials in polymer solar

cells: bulk, surface, and molecular level

Vanja Blazinic

2019:30

Probing the effects of photodegradation of acceptor materials in polymer solar cells: bulk, surface, and molecular level

Increase of the global energy demand and the climate change are two factors motivating the study and use of renewable energy sources, such as the solar energy. Organic photovoltaics (OPV) is a technology that uses organic molecules to convert solar energy into electricity. These organic molecules can be kept in ink form, allowing OPV device manufacture via coating, and ultimately roll-to-roll printing techniques, resulting in inexpensive, light weight, portable, and mechanically flexible sources of electricity. OPV devices have reached over 15% in power conversion efficiency, but their operational lifetime has to increase.

In this work, the photostability of the active layer in organic solar cells and its molecular components was studied by a variety of spectroscopy, microscopy and electrical characterization techniques, with focus on the chemical changes that these materials undergo during exposure to light and air. The aim was to determine the relation between materials’ degradation and the device performance degradation.

DOCTORAL THESIS | Karlstad University Studies | 2019:30

Faculty of Health, Science and Technology

Physics

DOCTORAL THESIS | Karlstad University Studies | 2019:30

ISSN 1403-8099

ISBN 978-91-7867-064-2 (pdf)

ISBN 978-91-7867-054-3 (print)

DOCTORAL THESIS | Karlstad University Studies | 2019:30

Probing the effects of photodegradation of

acceptor materials in polymer solar cells:

bulk, surface, and molecular level

Vanja Blazinic

Print: Universitetstryckeriet, Karlstad 2019

Distribution:Karlstad University Faculty of Health, Science and TechnologyDepartment of Engineering and PhysicsSE-651 88 Karlstad, Sweden+46 54 700 10 00

© The author

ISSN 1403-8099

urn:nbn:se:kau:diva-75093

Karlstad University Studies | 2019:30

DOCTORAL THESIS

Vanja Blazinic

Probing the effects of photodegradation of acceptor materials in polymer solar cells: bulk, surface, and molecular level

WWW.KAU.SE

ISBN 978-91-7867-064-2 (pdf)

ISBN 978-91-7867-054-3 (print)

i

Abstract Polymer solar cells (PSC) have reached record power conversion efficiencies of over 15%. The operational lifetime of PSCs, however, has to increase for their use in large area outdoor applications. In this work, a set of spectroscopic techniques (UV-vis, FTIR, NEXAFS, XPS) was used to study the impact of exposure to light and air (photo-oxidation) on the photoactive layer and its components. We focused on the electron acceptor components: the fullerene derivatives, PC60BM and PC70BM, and the polymer N2200. A comparative study of photo-oxidized PC60BM and PC70BM thin films by UV-vis and FTIR spectroscopy has shown that both materials undergo similar photochemical transformation, with the process being faster in PC60BM, due to the greater curvature of the C60 cage. Comparing experimental FTIR, XPS and NEXAFS spectra of the photo-oxidized PC60BM thin films with the calculated spectra for a large variety of photo-oxidation products, it was found that dicarbonyl and anhydride groups attach to the C60 cage during photo-oxidation. The study of photo-oxidized TQ1:PC70BM blend films by spectroscopic and J-V measurements shows that deterioration of the charge transport in PC70BM is the major contributor to the device performance degradation. Kelvin Probe measurements demonstrated that the charge transport deterioration was due to upward band bending and gap states being formed on the surface of photo-oxidized PC70BM. The TQ1:PC70BM blends films were further studied by AFM-IR in order to determine the lateral distribution of pristine components, as well as the photo-oxidation products. It was found that anhydride oxidation products of PC70BM are equally distributed over the blend film surface. The PC70BM is replaced with the polymer N2200 in the blend with TQ1. The photostability in air of the blend and its pristine components was studied by UV-vis and FTIR spectroscopy. The spectra show that thermal annealing improves the photostability in air of both components.

ii

List of publications

The following publications are included in this thesis: I. Photo-degradation in air of spin-coated PC60BM and PC70BM

films Vanja Blazinic, Leif K. E. Ericsson, Stela Andrea Muntean, and Ellen Moons Synthetic Metals, 2018, 241, 26-30, DOI:10.1016/j.synthmet.2018.03.021

II. Impact of intentional photo-oxidation of donor polymer and PC70BM acceptor on solar cell performance Vanja Blazinic, Leif K. E. Ericsson, Igal Levine, Rickard Hansson, Andreas Opitz, and Ellen Moons Physical Chemistry Chemical Physics, 2019, 21, 22259-22271 DOI:10.1039/C9CP04384E

III. The Photo-oxidation of PC60BM: New Insights from Spectroscopy Iulia Emilia Brumboiu, Leif K. E. Ericsson, Vanja Blazinic, Rickard Hansson, Andreas Opitz, Barbara Brena, and Ellen Moons manuscript

IV. Effects of morphology on the photodegradation of TQ1:PC70BM films: an AFM-IR study Leif. K. E. Ericsson, Vanja Blazinic and Ellen Moons manuscript

V. Stability of TQ1:N2200 active layers for all-polymer solar cells Vanja Blazinic, Leif K. E. Ericsson and Ellen Moons manuscript

iii

My contributions to the publications I. Prepared all the samples, carried out all the measurements and

analysed the results. Wrote the first draft of the manuscript. Contributed to of editing the final version of the manuscript.

II. Prepared all the samples, except those for the NEXAFS measurements. Carried out the J-V measurements and the EQE, UV-vis and FT-IR spectroscopy, and analysed the corresponding data. Wrote the first draft of manuscript. Contributed to editing of the final version of the manuscript and submitted the manuscript.

III. Prepared the samples for FT-IR spectroscopy and carried out the FT-IR measurements, contributed to the writing of the experimental part of the manuscript, and contributed to the discussions and analysis of the results.

IV. Prepared all samples, analysed FT-IR data, contributed to the writing of the manuscript.

V. Contributed to the development of the project idea. Prepared all the samples, carried out all the measurements, analysed the FT-IR and AFM data. Wrote the manuscript.

Papers not included in this thesis: Opportunities and challenges in probing local composition of organic material blends for photovoltaics Rickard Hansson, Leif K. E. Ericsson, Natalie Holmes, Vanja Blazinic, Paul Dastoor and Ellen Moons Journal of Materials Research, 2017, 32, 1982-1992 DOI: 10.1557/jmr.2017.7

iv

Acknowledgements First and foremost, I would like to thank my supervisors, Ellen Moons and Leif Ericsson, for the patience and motivation they poured into this project and my PhD education. A very big thanks goes to the fellow co-authors: Iulia Brumboiu and Barbara Brena, Andrea Muntean, Rickard Hansson, Igal Levine and Andreas Opitz. Thank you for all the help, contribution and advice you shared. This work would not have come to its fruition without all of you. To the members of the research group, former and present, Dargie Deribew, Ishita Jalan, Zhiyuan Cong, Jan van Stam and Zafer Hawash, thank you for making the time spent in the lab funny and productive, and for occasional nudges of encouragement when things would get stuck. A warm thanks also goes to the colleagues at Department of Physics and Engineering, and Computer Science, for all those fikas, innebandy and bowling evenings. In the end, a warm hug to my friends and family, for all the emotional and moral support you selflessly gave over the years, and not just the last four. In Karlstad, on October 14th 2019. Yes, there is a secret message here.

v

List of acronyms and abbreviations AEY Auger Electron Yield AFM Atomic Force Microscopy AM1.5 Air Mass 1.5 BHJ Bulk heterojunction CdTe Cadmium telluride CIGS Copper indium gallium selenide CT Charge transfer CTL Charge transport layer D/A Donor/acceptor DFT Density functional theory Eg Energy gap EQE External Quantum Efficiency ETL Electron transport layer FF Fill factor ϕa, ϕanode Anode work function ϕc, ϕcathode Cathode work function FT-IR Fourier Transform Infrared HOMO Highest Occupied Molecular Orbital HTL Hole transport layer IDT Indaceno dithiophene IDTT Indaceno dithienothiophene In2O3 Indium(III) oxide IPCE Incident photon-to-converted electron IR Infrared ITO Indium tin oxide Jd Diode/dark current Jph Photocurrent LiF Lithium fluoride LUMO Lowest Unoccupied Molecular Orbital MoOx Molybdenum oxide MPP Maximum power point NDI Naphthalene diimide NEXAFS Near-edge X-ray Absorption Fine Structure NFA Non-fullerene acceptor OSC Organic solar cell

vi

P(NDI2OD-T2)

poly{[N,N’-bis(2-octyldodecyl)-naphthalene-1,4,5,8-bis(dicarboximide)-2,6-diyl]-alt-5,5’-(2,2’-bithiophene)}

P3HT Poly(3-hexylthiophene-2,5-diyl) PC60BM [6,6]-phenyl-C60-butyric acid methyl ester PC70BM [6,6]-phenyl-C70-butyric acid methyl ester PCE Power conversion efficiency PDI Perlyene diimide PEDOT:PSS Poly(3,4-ethylenedioxythiophene)/poly(styrene

sulfonate) PES Photoelectron Spectroscopy PET Poly(ethylbenzene-1,4-dicarboxylate) PEY Partial electron yield PIB Polyisobutylene PSC Polymer solar cells PV Photovoltaics SnO2 Tin oxide STM Scanning Tunneling Microscopy TEY Total Electron Yield TFT Thin film transistor TQ1 Poly[2,3-bis-(3-octyloxyphenyl)-quinoxaline-5,8-

diyl-alt-thiophene-2,5-diyl] UHV Ultra-high vacuum UV Ultraviolet UV-vis Ultraviolet-visible Vbi Built-in potential VOC Open circuit voltage XPS X-ray Photoelectron Spectroscopy

vii

Contents Abstract ................................................................................................................... i List of publications ................................................................................................ ii My contributions to the publications ................................................................... iii Acknowledgements ............................................................................................... iv List of acronyms and abbreviations....................................................................... v

CHAPTER 1: INTRODUCTION............................................................... 1

1.1 Historical overview ........................................................................................... 1 1.2 Challenges for PSC devices .............................................................................. 3

CHAPTER 2: ORGANIC SOLAR CELLS ................................................. 7

2.1 Conjugated polymers and organic semiconductors ......................................... 7 2.2 Physics of organic solar cells ........................................................................... 9 2.3 Morphology of the active layer of organic solar cells .................................... 12

2.3.1 Spin coating ........................................................................................... 13

CHAPTER 3: MATERIALS ................................................................... 15

3.1 Polymer TQ1 ................................................................................................... 15 3.2 Polymer N2200 ............................................................................................. 16 3.3 Fullerene derivatives ..................................................................................... 17

CHAPTER 4: EXPERIMENTAL TECHNIQUES .................................... 19

4.1 Electrical characterization ............................................................................. 19 4.1.1 Current-voltage measurements ............................................................. 19 4.1.2 External quantum efficiency ................................................................ 24

4.2 Spectroscopic characterization ...................................................................... 25 4.2.1 Ultraviolet-visible absorption spectroscopy ......................................... 25 4.2.2 Fourier transform infrared spectroscopy ............................................. 28

4.3 Synchrotron-based techniques ...................................................................... 31 4.3.1 X-ray photoelectron spectroscopy ........................................................ 33 4.3.2 Near-edge X-ray absorption fine structure spectroscopy .................... 34

4.4 Atomic force microscopy ............................................................................... 37 4.4.1 Atomic force microscopy infrared spectroscopy ................................... 39

CHAPTER 5: SUMMARY OF THE PAPERS .......................................... 41

CHAPTER 6: CONCLUSIONS AND OUTLOOK .................................... 46

BIBLIOGRAPHY ................................................................................ 48

1

Chapter 1: Introduction

1.1 Historical overview Since the First Industrial revolution in the late 18th century, the conversion of natural resources, primarily fossil fuels, into electricity has had a significant impact on our civilization. With even faster industrialization during the 20th century, it has been estimated that the fossil fuel reserves will be depleted within the next hundred years.1,2 Even more concerning issue are the anthropogenic effects on the climate change and the ecosystem3–5, emphasizing the transition to renewable, environmentally friendly energy sources as a necessary solution. While wind and hydropower have been extensively implemented and used over the last two decades6, the application of photovoltaic (PV) energy systems offers several advantages. They use a long-term source – sunlight. The solar panels are a practical off-grid solution in areas where energy distribution infrastructure is unavailable or difficult to deploy.7,8 Several PV technologies are available among the renewable energy sources. Traditionally, silicon-based PV cells (1st generation PVs) have dominated the market since their commercialization in the 1960s. Recent progress in this PV technology has led to 26% and 22% efficiencies for single-crystalline and multi-crystalline silicon-based materials, respectively.9 The silicon-based PV modules, while being cheap with a price falling below 0.02 USD/kWh, still have a significant carbon footprint, due to the required high temperature required for processing, purification, and crystallization of the raw material.6,10,11 In the 2nd generation PVs, three new structures were introduced: cadmium telluride (CdTe), copper-indium-gallium-selenium (CIGS), and amorphous silicon (a-Si), with record efficiencies of 23%, 21% and 12%, respectively.9 These structures use less material due to the thin film deposition, and were the first PVs to be deposited on flexible substrates. However, in the case of CdTe and CIGS solar cells, thinner films are somewhat offset by the usage of materials which are not as abundant as silicon (tellurium and indium).12 Solar cells based on organic molecules and halide perovskites belong to the 3rd generation PVs. The latter are currently the fastest developing

2

PV technologies, with certified efficiencies of 12%, in 2013, to impressive 25% in 2018.13 Perovskite solar cells, however, are struggling with instability and environmental issues, as they are based on lead and the research is focused on finding a proper lead substitute.14 Polymer solar cells (PSC) are an interesting alternative to the aforementioned technologies for several reasons. They are based on semiconducting polymers, which can be synthesized and processed at room temperature, or at least at temperatures significantly lower than those for inorganic materials (~100°C for polymer synthesis as opposed to 900 – 1000°C for silicon, 500 – 600°C for CdTe).15 Polymers, and accompanying fullerene derivatives, are solution processable, allowing the storage of these materials in ink form. Semiconducting polymers, in particular, have high absorption coefficients, so light absorption can be achieved with very thin films, in the thickness range 100 – 200 nmi. Thicker films will absorb more light, but at the expense of increased loss in charge transport due to charge recombination.16,17 Solution processing and thin film deposition allows the application of roll-to-roll printing techniques adapted from the paper printing industry.18,19 As a result, low weight, mechanically flexible, large area PSC modules can be manufactured at low cost, providing a mobile source of renewable electric energy. PSC modules could be integrated with existing infrastructure, such as building windows, but also on a smaller scale as power supplies in autonomous sensor units and small electronics, as it has been shown they can operate in low illumination conditions, such as indoor light.20–22 With a recent record values for the power conversion efficiency of over 15%23, PSC devices can become competitive with conventional technologies in terms of price per unit power.24,25

i This is about 1000 times thinner than a single human hair.

3

1.2 Challenges for PSC devices One of the main challenges which inhibits faster and wider commercialization of PSC devices is their stability.26 Under protective laboratory conditions (storage, operation and testing in inert nitrogen atmosphere), the extrapolated lifetime of PSC devices has been shown to be up to 10-12 years.27,28 Outdoor tests have, however, shown that the lifetime of large area panels is roughly 1-2 years.29,30 A polymer solar cell is a device made of a photoactive layer, stacked between two electrodes of different materials, which are all sensitive to both internal influences, due interfacial reactions between materials, and external influences, such as heat, light, oxygen, moisture and mechanical stresses.30–32 Aluminium and indium tin oxide (ITO) are commonly used as the electrodes in PSC devices. ITO is often coated with poly(3,4-ethylene dioxythiophene)/poly(styrene sulfonate) (PEDOT:PSS), which has several roles: it adjusts the work function of ITO to higher value to improve charge collection, and it smoothens the rough ITO surface.33 PEDOT:PSS is prone to the water vapour uptake, so the PSS unit can develop an acidic compound, which etches the ITO under it. It has been demonstrated that clusters of indium may even migrate to other layers of the PSC device.33,34 Several transitional metal oxides have been proposed to replace PEDOT:PSS. These were thermally evaporated at first, but more recently several solution-deposited oxides have been reported as both electron transport and hole transport materials.35–38 As a replacement for ITO, solution-processed metal-doped zinc oxides39,40 and graphene-based materials41 are investigated. Aluminium will oxidize, leading to a loss of electrical conductivity. Encapsulation with a transparent sealant can reduce the uptake of atmospheric compounds, namely water vapour and oxygen. If it is to be used also as a supporting substrate in the roll-to-roll printed PSC modules, the sealant has to be resistant to aggressive chemicals, mechanically flexible and light weight in order to ensure modules’ easy transport and deployment. Poly(ethyl benzene-1,4-dicarboxylate)

4

(PET) has been commonly used as sealant, but there are a few promising alternatives such as polyisobutylene (PIB).42,43 In spite of the progress in the encapsulation materials’ research, it is vital to understand the impact of potential oxygen and moisture leak-in on the PSC devices and their individual components. Better understanding of how chemical processes negatively impact the optical and electronic properties of materials can lead to possible solutions of modifying existing, or synthesize new, more stable materials. While the search for more stable electrode and charge transport materials is ongoing, the greatest challenge to the PSC stability lies within the photoactive layer, sandwiched between the two electrodes. The photoactive layer consists of two materials, an electron donor polymer, which also commonly takes the main role as a light absorber, and an electron acceptor molecule, either a fullerene derivative, another polymer or non-fullerene small molecules. While efficiencies over 10% for polymer/fullerene OPV devices have been reached44, several studies have indicated that the degradation of devices is linked to the degradation of the electrical and chemical properties of the fullerene acceptor upon the exposure of these devices simultaneously to oxygen and light.45–51 Also, in the absence of atmospheric compounds, fullerenes were found to dimerise upon exposure to light or heat.52–59 Both dimerization and oxidation disrupt the chemical bonding, i.e. the alternating single bond – double bond character or the conjugation, which is necessary for (semi)conducting properties of the fullerene derivative. Given their inherent instability, poor absorption in the visible range and expensive synthesis, fullerene derivatives are being replaced with alternative electron acceptor materials, such as small molecules60,61 based on indaceno dithiophene (IDT), indaceno dithienothiophene (IDTT), and perlyene diimide (PDI), and polymers62–64 based on naphthalene diimide (NDI). The vulnerability of active layer components to light and air exposure (i.e. photo-oxidation) is one of the main challenges towards durable and efficient PSC devices. The aim of the work in this thesis was to investigate how the photo-oxidation affects the optical and chemical properties of the active layer components, particularly the electron acceptor materials, as well as the electrical performance of PSC devices

5

based on those active layers. By relating the changes in materials’ properties to the device performance, this work attempts to establish which processes in the PSC devices, from light absorption to electric current generation, are hindered by the photo-oxidation. Understanding this material degradation – device degradation link would lead to a better understanding of degradation processes, but could also provide insights towards improving the stability of the materials and the devices. Spectroscopic methods were employed to investigate changes in the chemical and optical properties of electron donor and acceptor materials upon photo-oxidation. The electrical measurements were performed to investigate the impact of the photo-oxidation of the photoactive layer on the PSC device performance, and the results were combined with the spectroscopic data to determine which processes, from light absorption to electrical current extraction, are hindered by the materials’ instability. Furthermore, a comparison of photochemical stability of two electron acceptor materials, two fullerene derivatives and a polymer, is made. Synchrotron-based spectroscopic techniques are employed with the aim of determining the nature of the photo-oxidation products of a fullerene derivative. The thesis is structured in the following way: Chapter 2 discusses the background of semiconducting polymers and physical processes that govern the light-to-current conversion in organic solar cells, Chapter 3 gives a brief description of the organic semiconducting materials studied in this work, Chapter 4 describes the experimental methods used in this work, Chapter 5 consists of a summary of the papers included in this thesis, followed by the conclusions in Chapter 6. The majority of experimental work was done at the Department of Engineering and Physics at Karlstad University, Sweden. NEXAFS measurements were performed at the Swedish national synchrotron facility MAX IV in Lund, Sweden, the Swiss Light Source at the Paul Scherrer Institute in Villigen, Switzerland, and the National Synchrotron Radiation Centre SOLARIS in Kraków, Poland. Kelvin Probe measurements and Surface Photovoltage Spectroscopy were performed by Prof. Ellen Moons and Dr. Igal Levine, respectively, at the Weizmann Institute of Science in Rehovot, Israel. For the

6

spectroscopic studies on the fullerene derivative PC60BM, theoretical calculations were performed by Dr. Barbara Brena at the Department of Physics and Astronomy at Uppsala University, Sweden, and Dr. Iulia Brumboiu at the Royal Institute of Technology (KTH) in Stockholm, Sweden, and presently on a post-doctoral stay at the Korea Advanced Institute of Science and Technology in Daejeon, South Korea.

7

Chapter 2: Organic solar cells

2.1 Conjugated polymers and organic semiconductors Among organic semiconductors used in solar cells, there are two classes of materials: small molecules and polymers. Small molecule-based solar cells were originally fabricated by vacuum deposition methods. Even though recent progress has led to solution processing alternatives, the non-fullerene acceptor (NFA) small molecule-based solar cells are still falling behind their polymer-based counterparts in efficiency, mainly due to the difficulties in charge transport and morphology control of NFA small molecule blends.65 Many polymers, whether naturally occurring, such as cellulose, or synthesized ones, such as PET, are classified as insulating materials. This insulating character comes from the characteristic chemical bonding in the polymer chains, i.e. single σ bonds between the carbon atoms. Due to this, electrons are strongly bound to their host atoms and localized, not allowing them any movement along the polymer chain. In contrast, (semi)conductive polymers have a bond structure that is comprised of alternating single and double bonds, forming a conjugated system. In each carbon atom, there are four valence electrons in the 2s22p2

configuration. The 2s and one of the 2p electrons form sp2 hybrid orbitals due to the overlap of their respective wave functions. These form three σ bonds and the corresponding σ* antibonds with neighbouring atoms. The three σ bonds spread in a shared plane, with an angle of 120° between them.16

Figure 2.1 Formation of the double carbon-carbon bonds.

8

The fourth electron is located in 2p orbital perpendicular to the plane formed by the three σ bonds, hence often designated as pz. If the two carbon atoms are close enough, two pz orbitals on neighbouring carbons will form a π bond, leading to the formation of the second bond between the two carbon atoms (C=C), as shown in Figure 2.1. While the electrons in σ molecular bonds are strongly bound and form the backbone of the polymer chain, the electrons in π bonds can be delocalized upon excitation and move between the monomer units. As the polymer chains form an organic solid through Van der Waals interactions, formerly individual bonding σ and π molecular orbitals and antibonding σ* and π* orbitals begin to overlap. Due to this overlap, the corresponding energy levels split, leading to the formation of energy bands (Figure 2.2), analogous to the bands in inorganic semiconductors. The molecular bands are not continuous, as the Van der Waals interactions between the monomer units are too weak for such formation. Of particular interest are the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO), which correspond to the valence and conduction band, respectively. The energy difference between HOMO and LUMO defines the bandgap Eg of the given material. The bandgap Eg is dependent on the effective conjugation length – the distance over which π electrons delocalize, and is shorter than the physical length of a polymer chain. As the effective conjugation length of the polymer increases, the bandgap Eg decreases, due to the wave function overlap and split of energy levels of adjacent molecules.16,66

Figure 2.2 The schematic of electronic structure formation: from two sp2 hybridized

carbon atoms, to a molecule, to a conjugated chain, to an organic solid.

9

As discussed in Section 2.2, two materials with different HOMO and LUMO are required in the organic solar cells. Materials with a higher LUMO will be defined as the electron donor, and materials with a lower LUMO will be defined as the electron acceptor.

2.2 Physics of polymer solar cells Polymer solar cells are multi-layered semiconductor devices (Figure 2.3), consisting of a photoactive layer, sandwiched between two electrodes. One electrode is semi-transparent and a commonly used material is indium tin oxide (ITO), an alloy of In2O3 and SnO2. On top of the ITO, PEDOT:PSS or other materials with high work function, such as transitional metal oxides (e.g. molybdenum oxide – MoOx67, vanadium oxide – V2Ox68, nickel oxide - NiOx37), are coated to facilitate the extraction of holes from the donor material into the ITO.33,34 The other electrode is a low work function material, most commonly aluminium with a thin layer of lithium fluoride (LiF), which facilitates the extraction of electrons from the acceptor material.69,70

Figure 2.3 Illustration of the layered structure of organic solar cells with typical thickness

ranges of each layer.

When the light is absorbed by an organic semiconductor, an electron is promoted from the HOMO, where it leaves a hole behind, to the LUMO (Figure 2.4a). Both are still bound by Coulomb forces, forming an electron-hole pair, or an exciton. Due to the low dielectric permittivity

10

in organic semiconductors (ε = 3 – 4) and the fact that the exciton is spatially confined to an individual conjugated segment of the molecule, the exciton binding energy in the organic semiconductors is high, in the range 0.4 – 1 eV,71,72 as compared to their inorganic counterparts. The thermal energy kBT at room temperature (~25 mV) is insufficient to promote the exciton dissociation in organic semiconductors.73,74 To ensure the exciton dissociation (i.e. charge separation), an energy level offset is needed. Two materials have to be introduced, and electron donor and an electron acceptor, each with different positions of HOMO and LUMO (Fig. 2.4b). The energy offset between the LUMOs of electron donor and electron acceptor must be greater than the exciton binding energy for charge separation to occur. Excitons can diffuse over the distance of 5-10 nm before they recombine.75,76 This distance defines the exciton diffusion length, and sets the limit how close the donor/acceptor interface should be to the exciton generation site. If the exciton is generated further away from the donor/acceptor interface, the electron and hole may recombine before reaching the interface. This type of recombination is called geminate recombination.77 Once the electron is transferred from the donor LUMO to the acceptor LUMO, it is still bound to the hole (which remains in the donor HOMO) in a charge transfer (CT) state, but not as tightly as in the exciton state. This allows for the CT state to separate into free charges (Figure 2.4b). The free charges are then transported away, electrons through the acceptor LUMO towards the cathode and holes through the donor HOMO towards the anode, under the influence of an internal electric field (Figure 2.4c). The internal electric field originates from the difference in the electrodes’ work functions, defined as the built-in potential:78

𝑉𝑉𝑏𝑏𝑏𝑏 =𝜙𝜙𝑐𝑐𝑐𝑐𝑐𝑐ℎ𝑜𝑜𝑜𝑜𝑜𝑜 − 𝜙𝜙𝑐𝑐𝑎𝑎𝑜𝑜𝑜𝑜𝑜𝑜

𝑒𝑒(2.1)

During the transport from the donor/acceptor interface to the respective electrode, free charge may recombine with an opposite charge or be trapped. The traps may be induced by energetic disorder, chemical impurities or structural defects. The recombination of free charges, regardless of the cause, is called non-geminate recombination.79

11

Figure 2.4 Energy diagram model of steps occurring during the photovoltaic effect, under

short circuit conditions, in organic solar cells: a) light absorption and exciton generation, followed by exciton diffusion to the donor/acceptor (D/A) interface; b) transfer of electron from donor LUMO to acceptor LUMO and exciton dissociation; c) charge transport from donor/acceptor interface to the respective electrodes; d) charge collection. Black circle represents an electron, white circle represents a hole. Oval encirclement indicates the electron-hole pair.

Charge collection, i.e. the transfer of charges from the photoactive layer to the respective electrode, is the final step in photocurrent extraction. The charge collection is most efficient when the acceptor LUMO and donor HOMO match the Fermi level of the cathode and anode, respectively, in such a way to form an ohmic contact at the photoactive layer/electrode interface when charges are being extracted (Figure 2.4d). As the aforementioned match is not frequently achieved, additional interlayers are inserted. These interlayers serve to modify the work function of the electrode to match the LUMO or the HOMO of the organic semiconductor, reduce contact (series) resistance, and modify the wettability of the surfaces.80 Due to the bulk heterojunction

12

morphology (discussed in Section 2.3) of the photoactive layer, both donor and acceptor components are usually in direct contact with both electrodes, so these interlayers also serve to promote (or block) one type of charge. Depending on the charge type, they are categorized as a hole transport layer (HTL) or electron transport layers (ETL).

2.3 Morphology of the active layer of organic solar cells Although the alignment of HOMO and LUMO levels at the donor/acceptor interface is important for exciton diffusion and dissociation into free charges, the microstructure (morphology) of the photoactive layer influences the free charge transport from the donor/acceptor interface to the electrodes. Initially, the organic solar cell devices featured a bilayer junction (shown in Figure 2.5a) and essentially mimicked the p-n junction of inorganic solar cells. C. W. Tang demonstrated first such organic solar with a bilayer heterojunction in 1986.81

Figure 2.5 Structures of photoactive layers in organic solar cells: a) bilayer; b) bulk

heterojunction (BHJ); c) ideal bulk heterojunction.

In such a junction, excitons may recombine if they are generated at the distance from the donor/acceptor interface that is greater than their diffusion length. The solution to the exciton diffusion length limit was introduced in the form of bulk heterojunction (BHJ) interface (shown in Figure 2.5b), which was demonstrated by Yu et al.82, for organic solar cells, and by Halls et al.83, for organic photodiodes. The bulk heterojunction is commonly described as an inter-percolating77 or interpenetrating84 bicontinuous network with a high interfacial area between the donor and acceptor phases. Such a mixture is more favourable than the bilayer structure as the probability of exciton

13

generation within 10 nm from the donor/acceptor interface is more probable, and so is the exciton dissociation into free charges. Ideally, distinctive and highly ordered phases (shown in Figure 2.5c) are be preferred. The width of the columns should not exceed exciton diffusion length. The formation of the bulk heterojunction, the dispersion of one blend component in a matrix of another and the phase separation into domains depend on several of parameters, such as: solubility of donor and acceptor materials in the selected solvent, solvent boiling point, usage of solvent additives or co-solvents, polymer molecular weight, blend component ratio, etc.85–89 These phases are not necessarily pure, but are rather donor-rich or acceptor-rich.86,90 Upon coating, one can apply solvent or thermal annealing to induce (re)crystallization of one or both components and facilitate a favourable microstructure.91,92

2.3.1 Spin coating On the laboratory scale, spin coating is a commonly used method for photoactive layer deposition as it allows reproducible formation of uniform thin films. Upon a (hydrophilic) substrate, mounted on a rotating disk, thin liquid film from the donor/acceptor blend solution is deposited (Fig. 2.6a). The disk rapidly accelerates to a high angular speed (500-2000 rpm) and solid thin film begins to form. The formation of a solid thin film consists of several steps shown in Figure 2.6b-d.

Figure 2.6 Illustration of spin coating: a) solution is deposited on a substrate; b) the

substrate is accelerated angularly and excess solution is ejected. c) the remaining solution radially spreads out over the substrate, while solvent begins to evaporate; d) the film thickness has reduced due to evaporation, the flow stops and film solidification begins.

14

During the initial angular acceleration, excess solution is ejected (Fig. 2.6b). The remaining liquid film flows radially outwards due to the liquid’s inertia and the adhesive forces between the liquid and the substrate, leading to the film thickness reduction. At the same time, solvent evaporation also contributes to the thickness reduction. As a result of these contributing processes, the solute concentration and the film viscosity increase and phase separation may occur (Fig. 2.6c). In the final step, the remaining solvent evaporates and viscous forces prevent further flow, eventually leading to the solidification of the film (Fig. 2.6d).85,93,94 As many mechanisms are involved in the film formation, modelling and predicting the film thickness and morphology have led to a variety of experimental62,85–87,91 and theoretical95–97 work. It has been established that the parameters such as initial amount of deposited solution, deposition rate or total spin time have little effect on the final film thickness, while angular velocity and solution concentration play a significant role. Empirically established mathematical relationship links the final film thickness to angular velocity93:

ℎ = 𝐶𝐶𝜔𝜔𝛼𝛼 (2.2) where: h – film thickness, ω – angular velocity, and C and α are empirically determined constants which depend on properties of the solutes, chosen solvent, and substrate. For industrial production, coating techniques such as doctor blading or slot die coating are applied.18

15

Chapter 3: Materials

3.1 Polymer TQ1 Poly[2,3-bis-(3-octyloxyphenyl)quinoxaline-5,8-diyl-alt-thiophene-2,5-diyl] (TQ1), an electron donor material, is a conjugated copolymer comprised of alternating quinoxaline electron acceptor and thiophene electron donor units, as shown in Figure 3.1. The insertion of alternating acceptor-donor units along the conjugated polymer backbone causes a spectral shift in absorption towards the near-IR, allowing for synthesis of low-bandgap polymers. Furthermore, it has been shown that the proper choice of donor and acceptor units can contribute to the reduction of repulsive steric interaction between neighbouring units, thus contributing to the planar stacking of the polymer chains and increased π-conjugation length.98,99 Simplified synthesis of this polymer and introduction to polymer solar cells was accomplished by Wang et al.100 in 2010. The record power conversion efficiency of 7% in TQ1:PC70BM organic solar cells was achieved by Kim et al.101

Figure 3.1. Chemical structure of TQ1.

Several derivatives of this polymer have been synthesized, showing an increased open circuit voltage102, stability against photobleaching103,

and increased overall performance104, as compared to the 7% achieved by Kim et al.101

16

3.2 Polymer N2200 The polymer poly{[N,N’-bis(2-octyldodecyl)-naphthalene-1,4,5,8-bis(dicarboximide)-2,6-diyl]-alt-5,5’-(2,2’-bithiophene)} P(NDI2OD-T2), commonly known as Polyera ActivInkTM N2200105, is an electron acceptor material, a copolymer consisting of naphthalene diimide (NDI) acceptor unit and dithiophene donor unit, which both ensure the geometric planarity of the polymer (Fig. 3.2). It was originally synthesized for use in organic thin film transistors as n-channel component. Due to the polymer’s high electron mobility and considerable stability of N2200-based TFT devices in storage ambient conditions, the N2200 was introduced to the usage in polymer solar cells. High electron mobility has been credited to the lamellar packing of the N2200 molecular chains.106,107 The degree of stacking can be reduced by the presence of a second polymer, typically in the blend. The stacking can be recovered by thermal or solvent annealing of the blend.108–110

Figure 3.2. Chemical structure of N2200.

The power conversion efficiencies of devices based on P3HT:N2200 active layer were poor due to morphological hindrances. With the development of new polymer donor materials and optimized morphologies, organic solar cells using this polymer have reached power conversion efficiencies over 8% and have demonstrated improved stability, even when processed under ambient conditions.111–113

17

3.3 Fullerene derivatives Fullerenes are allotropes of carbon, consisting of carbon atoms connected by single and double bonds in a way to form fused rings, closed into a cage (Fig. 3.3). The cage formation conforms to the isolated pentagon rule, meaning no two pentagon rings share an edge, i.e. they are surrounded by hexagons.114 The smallest possible fullerene to be formed in that way is C20. Fullerene C60 was discovered by Kroto et al.115 in 1985, while Hare et al.116 developed a procedure to separate C60 from C70, another fullerene. Due to their high electron affinity117,118 and good conductivity119,120, fullerenes have been considered a promising material in (organic) electronics. However, the fullerenes in their original form are not very soluble in commonly used solvents, so fullerene-based devices were, at first, made by the thermal evaporation in vacuum.118 The problem of low solubility was overcome in 1995, when Hummelen et al.121 synthesized [6,6]-phenyl-C60-butyric acid methyl ester (PC60BM), a C60 derivative shown in Figure 3.3a.

Figure 3.3. Chemical structure of the fullerene derivatives a) PC60BM and b) PC70BM.

Another commonly used derivative is [6,6]-phenyl-C70-butyric acid methyl ester (PC70BM), shown in Figure 3.3b. It has a lower symmetry compared to the to PC60BM, which results in the increased absorption in the visible region122, thus contributing to a better light harvesting in polymer/fullerene active layers and improved device performance.123,124

18

In the presence of light or heat, fullerenes and fullerene derivatives are prone to oxygen take-up from ambient air and can react chemically with it, which then leads to the breakage of C=C bonds and the formation of carbonyl structures on the fullerene cage.46,125–128 Power conversion efficiencies over 11% for polymer/fullerene devices have been reached.129 The fullerene derivatives suffer from two drawbacks: expensive synthetization (cca 130 €/g for PC60BM, cca 600 €/g for PC70BM) and lower absorption in the visible range.130,131 As opposed to fullerenes, polymers and small molecules, both electron acceptor and donor, can be tailored in order to facilitate more suitable optical and electrical properties.

19

Chapter 4: Experimental techniques

4.1 Electrical characterization

4.1.1 Current-voltage measurements The J-V measurement is the basic method for characterization of solar cells. It is performed by shining light on the device and measuring the current I or current density J at a given applied voltage V. The J-V characterization of solar cells is also performed in the absence of illumination, since the devices behave as diodes, in order to determine electrical properties such as parasitic resistances. For the ideal diodes, the dark current Jd can be described with the Shockley diode equation:

𝐽𝐽𝑜𝑜 = 𝐽𝐽0 ⋅ exp �𝑒𝑒𝑉𝑉𝑛𝑛𝑘𝑘𝐵𝐵𝑇𝑇

− 1� (4.1)

where: J0 - reverse saturation current density, e - elementary charge, V - voltage applied across the diode, kB - Boltzmann constant, n - diode ideality factor, and T - absolute temperature. Operation of an polymer solar cell in the dark can be explained with the help of the energy diagrams in Figure 4.1. In the high forward bias region (Fig. 4.1a), V > Vbi, the electric field is strong enough to tilt the energy bands towards the anode, thus allowing injection of electrons from the cathode into to the acceptor LUMO and holes from the anode into the donor HOMO to flow towards the donor/acceptor interface. At the interface, electrons and holes may recombine, which results in radiative recombination (generation of photons) or non-radiative recombination. As the applied voltage is reduced below the voltage Vbi (i.e. 0 < V < Vbi), the energy bands will tilt in the opposite direction, towards the cathode. The probability of injecting the electrons and holes is lower, leading to a significant drop in current density Jd. At short circuit conditions V = 0 (Fig. 4.1b) no net current is flowing through the diode. The energy bands are still tilted due to the built-in potential Vbi, which is caused by the difference between the work function of the anode Φa and the cathode Φc (Eq. 2.1). In the reverse bias V < 0 (Fig. 4.1c), only a very small current flows, due to barrier that hinders charge injection.

20

Figure 4.1. Energy diagram of a simplified model for the donor/acceptor interface of a

BHJ solar cell at different applied voltages V across the polymer solar cell in the absence of light (a-c). a) In the forward bias region V > 0, electrons (black circles) are injected from the cathode to the acceptor’s LUMO and holes (white circles) are injected to the donor’s HOMO; b) At short circuit V = 0, the bands are tilted due to the existence of a built-in electric field, represented by the potential Vbi; c) In the reverse bias region V < 0, the voltage across the diode causes a strong electric field which prevents the charges from flowing; d) Dark current density graph of a polymer solar cell device, with different voltage regions indicated.

Should the active layer absorb a photon with energy above the energy gap (E > Eg), this will generate an exciton which may dissociate at the donor/acceptor interface into a free electron and a hole. These photogenerated charges will be accelerated towards the electrodes, i.e. in the opposite direction of the dark current density Jd, due to the built-in electric field, and constitute the photogenerated current density Jph. The total current density will then be J = Jd - Jph. In the high forward bias region (V > VOC), the charges are injected into the PSC device. The excitonic pairs are still generated, but tilting of the energy bands is unfavourable for the exciton dissociation. Total current density J is dominated by the Jd term. As the applied voltage is reduced, the

21

injected current density also reduces. It will reach a point where Jd and Jph are equal and cancel each other out, leading to the total current density J = 0, and at this point applied voltage is equal to the open circuit voltage VOC.

Figure 4.2. a) Typical J-V characteristics of a solar cell in dark (grey) and under

simulated sunlight (black). Ratio between grey shaded and dashed rectangle areas defines the fill factor FF. b) Equivalent electrical circuit of the solar cell under illumination.

In the lower forward bias region (0 < V < VOC), the total current density changes direction – it begins to flow outward from the OSC device, as the photocurrent Jph begins to compete with injection current Jd. As the applied voltage is reduced further, it will reach a point Vmax at which maximum power of the solar cell can be gained. When the applied voltage equates the built-in potential Vbi, maximum current density can be obtained from the solar cell, i.e. short circuit current density JSC. The current density-voltage output measured in dark and under illumination (AM1.5) is shown in Figure 4.2a. The J-V characteristic is modelled with an equivalent circuit, as shown in Figure 4.2b. In the equivalent circuit, the parasitic resistances, parallel (or shunt) RP and series RS, are introduced to account for various processes that may interfere with the device performance. The parallel resistance RP accounts for current leakage, for instance due to manufacturing imperfections (e.g. pinhole protrusions)66. The series resistance RS accounts for materials’ resistances, interface resistances between individual layers and degradation of electrode contacts. Its greatest impact is on reduction of solar cell output current132. From J-V measurements in the dark (i.e. Jph = 0), it is possible to extract the parameters J0, n, and RS. In the first approximation, the parallel

22

resistance RP can be estimated using Ohm’s law, assuming linear J-V dependence in the region around V = 0:

𝑅𝑅𝑃𝑃 = �𝑑𝑑𝑉𝑉𝑑𝑑𝑑𝑑�𝑉𝑉→0

(4.2)

Taking these contributions into account, the output current density J of a non-ideal solar cell under illumination can be written as:

𝐽𝐽 = 𝐽𝐽0 ⋅ exp ��𝑒𝑒(𝑉𝑉 − 𝐽𝐽𝑅𝑅𝑺𝑺𝑛𝑛𝑘𝑘𝐵𝐵𝑇𝑇

� − 1� +𝑉𝑉 − 𝐽𝐽𝑅𝑅𝑺𝑺𝑅𝑅𝑃𝑃

− 𝐽𝐽𝑝𝑝ℎ (4.3)

The quality and efficiency of OSC devices is evaluated from the measured J-V characteristics under illumination (AM1.5 spectrum), and several parameters are extracted from these measurements.77,133,134 Open circuit voltage VOC is the voltage at which no net current J flows, since the electrodes are not electrically connected into the closed circuit and the dark current Jd and photogenerated current Jph cancel each other out. The VOC depends on the choice of materials in the active layer, namely the difference of energy positions of the donor HOMO and the acceptor LUMO. The short circuit current density JSC is the current generated when the applied external voltage is zero (V = 0). The short circuit current is strongly influenced by the optical properties of the light source (intensity and spectrum), the absorption properties of active layer materials, but also by charge transport and charge extraction efficiency. Maximum output power density Pmax = JmaxVmax, is the power per unit area generated by the solar cell at the maximum power point (MPP), shown in Figure 4.2a. Fill factor (FF) is defined as the ratio of maximum power provided by the solar cell and the product JSCVOC:

𝐹𝐹𝐹𝐹 = 𝑃𝑃𝑚𝑚𝑐𝑐𝑚𝑚

𝐽𝐽𝑆𝑆𝑆𝑆 ⋅ 𝑉𝑉𝑂𝑂𝑆𝑆(4.4)

It indicates how efficiently photogenerated charges are extracted from the solar cell. It has been shown that the fill factor is related to the

23

active layer morphology, namely the dimensions of donor and acceptor material domains and the selected materials, but also charge generation and recombination and charge mobilities.135,136 Graphically, it defines the squarness the of J-V characteristics of a given solar cell device. From the VOC, JSC and FF, the power conversion efficiency (PCE) is calculated as a ratio of maximum output power density Pmax of the solar cell and the power density of incident light Pin:

𝑃𝑃𝐶𝐶𝑃𝑃 = 𝑃𝑃𝑚𝑚𝑐𝑐𝑚𝑚𝑃𝑃𝑏𝑏𝑎𝑎

= 𝐽𝐽𝑆𝑆𝑆𝑆 ⋅ 𝑉𝑉𝑂𝑂𝑆𝑆 ⋅ 𝐹𝐹𝐹𝐹

𝑃𝑃𝑏𝑏𝑎𝑎(4.5)

As discussed previously, the characterization of solar cells requires J-V measurements performed on illuminated devices, as well as in the dark. For comparable and reproducible reports on the solar cell performances, the illumination source has to accurately simulate the solar spectrum. As a standard, the solar spectrum measured at 48° to the zenith is taken (Fig. 4.3). The solar spectrum intensity is normalized to produce an integrated irradiance of 1000 W/m2. It is designated as Air Mass coefficient 1.5 - AM1.5, meaning that the optical length path of solar radiation is 50% greater than that at the sea level when sun is in zenith at equator (AM1). The Air Mass coefficient is calculated from zenith angle Θ as AM = 1/cos Θ.66

Figure 4.3. The standard solar spectrum AM1.5.137

24

4.1.2 External quantum efficiency The external quantum efficiency (EQE) or incident photon-to-converted electron (IPCE) measurement, is a method used to determine the spectral response of the current generation in solar cells. If Ne is the number of electrons generated by the solar cell and Nph is the number of photons incident on the sample, then EQE at a given wavelength λ will be:

𝑃𝑃𝐸𝐸𝑃𝑃(𝜆𝜆) =𝑁𝑁e𝑁𝑁ph

⋅ 100 =𝑑𝑑(𝜆𝜆)𝑃𝑃(𝜆𝜆)

ℎ𝑐𝑐𝜆𝜆 ⋅ 𝑒𝑒

⋅ 100 (4.6)

where: I (λ) is the generated current, P(λ) is power of monochromatic light beam at wavelength λ, h is Planck’s constant, and c is a speed of light.138,139 While the J -V measurements under white light (AM1.5) illumination provide the information on overall current generation and device PCE, the EQE measurements provide the means to measure how specific regions of the solar spectrum contribute to the current generation. The intensity of the EQE signal will depend on probabilities of individual processes that lead to current generation: photon absorption, exciton diffusion to the donor/acceptor interface and its dissociation, and free charge transport. The EQE measurements are performed by shining monochromatic light from a monochromator on the solar cell device while measuring the current generated at the given wavelength. To ensure that the solar cells under consideration are characterized in standard illumination conditions, the devices are illuminated with white bias (AM1.5) light. In order to distinguish the contribution of white light from the monochromatic component, the latter is sent through an optical chopper, which is connected to a lock-in amplifier. By chopping the monochromatic light, the lock-in amplifier can measure the difference in current under illumination of both monochromator and bias light, Imono+AM1.5, and the current under bias light only, IAM1.5, i.e. I = Imono+AM1.5 – IAM1.5.138,139 In this way, the differential EQE is measured. It has been shown that for light intensities

25

of 1000 W/m2 and lower, the difference between differential EQE and real EQE is negligible.139 The EQE measuring setup is usually calibrated with respect to a standard silicon or germanium photodiode.140

4.2 Spectroscopic characterization In physics and chemistry, selection rules define the probability of transition between two energy states of an atom or a molecule. Selection rules, like transitions, can be divided into electronic, vibrational and rotational, depending on which spectroscopic method is used to investigate and observe those transitions. In spectroscopy, the underlying principle, is the Fermi golden rule (Eq. 4.7). For UV-vis, IR and X-ray spectroscopy, it can be derived from oscillating perturbation theory. The Fermi golden rule defines transition probability wfi as:

𝑤𝑤𝑓𝑓𝑏𝑏 = 2𝜋𝜋ℏ�𝑉𝑉𝑓𝑓𝑏𝑏�2𝜌𝜌�𝑃𝑃𝑓𝑓𝑏𝑏� (4.7)

where: Vfi is transition probability matrix, and ρ(Efi) is density of states at energy Efi corresponding to the transition from initial state ψi to final state ψf. The Fermi golden rule is applicable to different cases of electromagnetic excitation of a molecule, assuming that proper Vfi is used. For different excitations, i.e. different spectroscopic methods, additional rules may apply.141

4.2.1 Ultraviolet-visible absorption spectroscopy Organic compounds usually absorb light in the ultraviolet or visible region of the electromagnetic spectrum. Absorption will occur only if the photon energy of incident radiation is identical to the difference between the energy levels of allowed electronic transitions in a molecule. Upon absorption of a photon of energy E = hν, molecule will be excited to higher electronic, vibrational or rotational energy states. The total energy increase can be expressed as:

ℎ𝜈𝜈 = Δ𝑃𝑃e + Δ𝑃𝑃vib + Δ𝑃𝑃rot (4.8)

26

where: ∆Ee, ∆Evib, ∆Erot are energy differences between electronic, vibrational and rotational state, respectively. The excitations to higher rotational energy states are observable in the microwave region, while the vibrational ones in the infrared region (discussed further in Section 4.2.2). There are several classes of transitions in organic molecules. The σ-σ* transitions occur in the far UV region (below 200 nm) and cannot be observed with commonly used UV-vis spectrophotometers. The π → π* transitions occur in conjugated molecules. The absorption features will depend on the extent of conjugation of the molecule.142

Figure 4.4. a) Illustration of an energy level diagram. Energy levels E0, E1, E2 correspond

to electronic states, while e0, e1, e2 correspond to vibrational states. Vibrational states e’ and e’’ are associated with higher electronic states E1, E2 respectively. Adapted from reference.143 b) UV-vis absorption spectra of thin films of the materials used in this work.

Although the energy levels are discrete (Fig. 4.4a), the absorption features do not appear as sharp lines, as is the case with atomic line spectra, but rather as broad bands (Fig. 4.4b). This is due to the fact that electromagnetic radiation in UV-visible range (300-800 nm) is energetic enough to excite all three energy states (electronic, vibrational and rotational). For each electronic state there are several closely positioned vibrational and rotational states. As they cannot be resolved by UV-vis spectrophotometer, the resulting spectra are the sum of all transitions and appear as continuous bands. The absorption spectra are obtained in the following way (Fig. 4.5).

27

Figure 4.5. Illustration of UV-vis spectrophotometer optics.

Light is focused, through a set of mirrors, onto the rotational grating of the monochromator. By rotating the grating, a specific wavelength can be selected, giving rise to the monochromatic light. Monochromatic light is then shone onto the sample (e.g. thin film spin coated on quartz substrate). Intensity of light passing through the sample, I, is measured by the photodetector and compared to that of the reference light beam, I0.143 The absorbance at the given wavelength is then calculated via equation:

𝐴𝐴 = − log10 �𝑑𝑑𝑑𝑑0� (4.9)

As the absorption at a given wavelength is related to the electronic transition within the sample, absorption spectra can give insight into electronic properties of studied materials, as well as molecular configurations, aggregate formation, etc. This is particularly important in conjugated molecules where the shifts of the absorption maxima are indicative of changes in the conjugation length of the system. In general, longer conjugation length of the material redshifts the absorption maxima. UV-vis spectroscopy is often employed to study effects of PSC materials’ processing conditions144, exposure to light145–

147, heat55,148, humidity and different atmospheres on organic materials149. In this work, UV-vis absorption spectroscopy was used to

28

monitor photochemical stability of solar cell materials upon exposure to simulated sunlight (AM1.5) in ambient air. Changes in the absorption spectrum are usually the first indicator for degradation of the photoactive materials in the solar cell.

4.2.2 Fourier transform infrared spectroscopy Infrared (IR) radiation is usually not strong enough to excite the electronic transitions in photoactive layer materials. A molecule will absorb IR radiation if the photon energy matches the energy difference between two discrete vibrational states and only if it causes a net change in the molecular dipole. The vibrational frequency, ν, at which a part of the molecule vibrates, can be estimated on a model diatomic molecule (e.g. C=O), in which the molecular bond can be represented by a spring. The frequency of bond vibration can be calculated from:

𝜈𝜈 =1

2𝜋𝜋�𝐾𝐾𝜇𝜇

(4.10)

where: K is the spring constant representing the bond strength, and µ is the reduced mass. In IR spectroscopy, the frequency is often expressed in wavenumber value k = ν/c = 1/λ.150 Equation 4.10 can be derived both from classical and quantum mechanical treatment of a model diatomic molecule. Molecular bonds can undergo two types of vibrations: stretching - bond length changes, and bending - bond angle changes. Illustrations of different vibrations are shown in Figure 4.6.

29

Figure 4.6. Illustrations of different vibrational modes of a triatomic molecule. a)

Stretching modes. b) In-plane bending modes. c) Out-of-plane bending modes. Assuming the plane is in the page, “-“ indicates an atom is moving away from the reader, “+” indicates an atom is moving towards the reader. Adapted from reference143.

These modes of vibration are commonly observed in the mid-IR spectral region of 4000-400 cm-1 (2.5-25 μm) and are used for chemical identification. The IR spectrum can, in principle, be obtained with an IR spectrometer with a similar optical system as the UV-vis spectrometer (using an appropriate IR source and a detector). However, the Fourier-transform Infrared (FT-IR) spectrophotometer has been a significant development, with better signal-to-noise ratio, higher resolution, and shorter data acquisition time.143 Today FT-IR spectroscopy is commonly used. The schematic of a typical FT-IR spectrophotometer is shown in Figure 4.7.

30

Figure 4.7. Schematic of interferometer used in FT-IR spectroscopy. Adapted from

reference.150

The central part of the FT-IR spectrophotometer is a Michelson interferometer. The incident IR radiation is divided by a beam splitter into two beams, which are directed towards a stationary mirror M1 and moving mirror M2. The reflected beams meet back at the beam splitter and interfere constructively or destructively, dependent on their optical path length difference. If the IR source had been monochromatic, the dependence of the resulting beam intensity on position x of the moving mirror M2 would be:

𝑑𝑑(𝑥𝑥) = 12𝑑𝑑(𝑘𝑘) cos(2𝜋𝜋𝑘𝑘𝑥𝑥) (4.11)

where I (k) is the intensity of IR source. The resulting dependence from Eq. (4.11) is called an interferogram.

In an FT-IR spectrophotometer, the IR source emits a continuum of wavenumbers and for a given mirror M2 position x, only components at specific wavenumber will interfere constructively (i.e. those satisfying condition x = nλ) and be transmitted further. In FT-IR information on all emitted wavelengths is acquired simultaneously. For a different mirror M2 position x, different components of the IR source radiation will be transmitted through or absorbed by the sample, before reaching the detector. The raw data have to be processed by the computer

31

through a Fourier transform algorithm for the spectra to be obtained.142 The position x of mirror M2 is measured with the help of a helium-neon laser, leading to improved accuracy of the method. The obtained spectra are used for chemical identification of unknown compounds or tracking of chemical changes in the established samples46,151.

4.3 Synchrotron-based techniques Following their discovery by W. Röntgen in 1895, X-rays (or Röntgen rays) have been used to characterize materials. While the initial sources were X-rays tubes, synchrotrons are used as a primary source of X-ray radiation for scientific research in modern times, due to their wide spectrum (from UV to X-ray region), ability to generate higher photon flux and higher brilliance, a physical quantity which describes the angular distribution of photon flux. It is defined as:

[Brilliance] = �photons/second

(mrad2)(mm2 source area)(0.1% bandwidth)�

In a synchrotron, an electron beam is bent by dipole (bending) magnets into a circular motion. This will cause a radial acceleration of the charges and, consequently, produce an X-ray radiation along the tangent of the electron’s path. Since the electrons move at velocities close to the speed of light, the angular distribution of radiation will be narrow, described by natural opening angle Θ:

𝛩𝛩 ∝ 𝛾𝛾−1, 𝛾𝛾 =𝑃𝑃

𝑚𝑚e𝑐𝑐2

where: γ is the Lorentz factor, and mec2 is the electron rest energy.152 Synchrotron system is shown in Figure 4.8.

32

Figure 4.8. Illustrative representation of a synchrotron and additional facilities

needed to perform X-ray spectroscopy.

A source generates electrons through thermal emission from a heated filament. Electrons are then linearly accelerated to a booster ring. The purpose of the booster ring is to further accelerate electrons to a higher energies before they are delivered to the storage ring. Depending on the type of synchrotron, the electrons are injected periodically to maintain the specified storage ring current or, in more recently developed facilities, they are delivered into the storage ring in smaller, but more frequent bursts. The latter allows operation of the beamlines with almost no interruptions. Once in the storage ring, electrons are kept in a closed path by a set of magnets:152

a) Dipole or bending magnets keep electrons in the closed path within the storage ring.

b) Quadrupole magnets are used to focus the electron beam and minimize the Coulomb repulsion between electrons.

c) Sextupole magnets compensate for the chromatic aberrations caused by quadrupole focusing.

The synchrotron storage ring is not circular, but rather consists of straight line segments, connected at the bending magnet joints. The straight line segments are housing insertion devices, which produce intense X-ray radiation, along with bending magnets. Insertion devices are divided into wigglers and undulators. Both are comprised of

33

periodic magnetic structures set in a way to bend the beam trajectory into an oscillatory fashion. With each deviation, radiation is emitted and superimposed on the radiation generated in the next segment. The main difference in wigglers and undulators is that the latter consist of more, closely packed segments thus providing an X-ray beam of smaller angular spread. The emitted radiation is then sent down the beamline, where it is focused and monochromated before being delivered to the end station.

4.3.1 X-ray photoelectron spectroscopy Photoelectron spectroscopy (PES) is an analytical technique used to investigate electronic structure of occupied states in the sample surface by analysing the energy spectra of the photoelectrons. The X-ray photoelectron spectroscopy (XPS), with the energy range 0.1-10 keV for the incident beam, probes the core atomic levels, as shown in Figure 4.9a.

Figure 4.9. a) Processes in XPS. b) Processes in NEXAFS.

Electron, initially in a state with binding energy EB, absorbs a photon of energy hν and leaves the sample with kinetic energy:

𝑃𝑃kin = ℎ𝜈𝜈 − 𝑃𝑃B − 𝜑𝜑S (4.12)

34

where φS is the work function of a sample material. The experimental setup of PES (Fig. 4.10) has a light source, a sample in UHV conditions, and an electron energy analyser. Reaching the source slit of the electron analyser, emitted electrons enter a region of lenses where they are accelerated and focused at the entrance of the detector. An electric potential is applied to the analyser hemispherical electrodes to allow only electrons within a certain kinetic energy range to pass through and reach the detector. The outer electrode is biased negatively with respect to the inner one. Sample and detector are in joint electrical contact to common ground potential which ensures they have the same Fermi level.153

Figure 4.10. Schematic of XPS system. Adapted from reference.154

As the binding energies of the chemical elements are known, XPS is commonly used for identification of elemental constituents. It also yields information on the type of chemical bonds in the material. For organic solar cells, XPS has been used to investigate stability of materials for active layers48,50,155, processing of charge transport layer (CTL) materials37,156 and organic/CTL interfaces36,157.

4.3.2 Near-edge X-ray absorption fine structure spectroscopy NEXAFS is commonly used to study chemical and structural properties of the thin films. Unlike XPS, where the energy of the incident X-ray beam is kept constant, and it is high enough to promote core electron

35

above the vacuum level, the incident X-ray beam in NEXAFS is varied across the energy interval around the binding energy of core electrons, i.e. energy required to promote a core electron to the unoccupied states (Fig. 4.9b). Another electron, from an outer shell, will fill the core-hole, leading to an excess energy within the atom. The atom will relax to a lower energy state through one of the two processes, as shown in Figure 4.9b.158,159 The transition of an outer shell electron to the core level will emit a fluorescence photon (radiative emission) or an Auger electron (non-radiative emission). The radiative emission is common in high atomic number elements, since increased number of protons in the nucleus bounds core electrons more tightly. The non-radiative emission is typical for low atomic number elements, threshold between the two processes being roughly at Z = 30.152 As carbon constitutes the majority of semiconducting organic materials (followed by oxygen and nitrogen), the non-radiative emission is of greater interest. X-ray absorption in NEXAFS can be measured directly, by measuring the intensity of X-ray light transmitted through the sample, or by detecting the emission of electrons or fluorescence yield. The electron yield can be measured in three different modes, each with a different surface sensitivity.159,160 The Auger electrons, detected in electron yield, are emitted with characteristic kinetic energies which are independent of incident X-ray beam energy and can be detected using an electron energy analyser. As the electrons may be inelastically scattered before leaving the sample, only those emitted from very close to the surface will be detected. The distance that electrons can cover before inelastic scattering is defined an electron mean free path, and its dependence on electron kinetic energy is shown by a graph in Figure 4.11.

Figure 4.11. Mean free path as a function of electron kinetic energy.161

36

The energy electron analyser can be adjusted to detect only the electrons within a certain energy range of characteristic Auger transitions. This makes Auger electron yield (AEY) very sensitive to surface modifications, in order of 1 nm below the sample surface. Partial electron yield (PEY) is measured by a channeltron (i.e. electron multiplier). In front of the channeltron, a grid with retarding voltage is installed to block slow electrons. More energetic electrons, coming deeper from the sample (3-5 nm below the surface160) can be detected. In total electron yield (TEY), all electrons ejected from the sample are counted by measuring the drain current needed to keep it electrically neutral (Fig. 4.12). Information depth of TEY is around 10 nm.90

Figure 4.12. A schematic of the detection modes in NEXAFS. Adapted from reference.160

As the incident X-ray light is promoting electrons from core-level states to unoccupied states in NEXAFS, the energy of detected electrons will not only depend on the specific chemical element, but also on the chemical surrounding, i.e. types of bonds, and structural properties (bond orientation). In the past, NEXAFS has been extensively used to study molecular stacking of conjugated polymers162–164, effects of thermal annealing on conjugated polymers and fullerenes91,165, effects of blend morphology on PSC active layer electronic structure and subsequent impact on PSC device performance87,89 and photo-oxidation of fullerene C60 and its derivative PC60BM128,166.

37

4.4 Atomic force microscopy Atomic force microscopy (AFM) belongs to a family of scanning probe microscopy techniques. While this field began with the development of scanning tunnelling microscopy (STM) in 1982, by G. Binning and H. Rohrer, the introduction of AFM gave an offset to this family of techniques, as it extended the possibility of surface characterization into insulating materials under ambient atmosphere conditions.167

Since STM is based on measurements of tunnelling current, it is limited to the conductive surfaces and not suitable for organic samples, namely polymer/fullerene and polymer/polymer blend thin films. In the AFM, a sharp tip is attached to a flexible cantilever, which is used as a force sensor. Interaction between the tip and the sample surface will cause cantilever deflection which is measured by the laser beam light reflected from the cantilever into position calibrated photodiode, as shown in Figure 4.13.

Figure 4.13. Schematic of typical AFM setup.

The type tip-sample interaction will depend on the operational mode of AFM used, but in most general case the interaction can be approximated with the Lennard-Jones potential168:

𝑉𝑉𝐿𝐿−𝐽𝐽 = 4𝜀𝜀 ��𝜎𝜎𝑟𝑟�12− �

𝜎𝜎𝑟𝑟�6

� (4.13)

38

where: r is the distance between the tip and the sample surface, ε is the depth of the potential and the point of forces’ equilibrium, and σ is a distance where the potential is zero. The 12th-power term originates from the Pauli repulsion and the 6th-power term represents the attractive Van der Waals forces. The Lennard-Jones potential, along with distance range of commonly used operational modes in AFM, is given in Figure 4.14. In the contact mode, the cantilever is kept at constant deflection (constant force) by adjusting the tip-sample distance through a feedback loop. As the repulsive forces vary greatly at the short tip-sample distances, height profiles of constant force will correspond to the local topography of the sample. Since the forces in contact mode are strong, it is preferable to use this mode on hard surfaces.167

Figure 4.14. The Lennard-Jones potential with ranges of operational AFM modes.

For soft sample surfaces, such as the active layer thin films, intermittent contact mode (also known as tapping mode) is used. In this mode, the cantilever is driven at its resonance frequency with a large oscillation amplitude. The large oscillation amplitude ensures that the cantilever comes into contact with the sample once per oscillation cycle, but also to move the tip further away, which reduces the attractive tip-sample interaction and decreases the probability of

39

sample being damaged. The amplitude, due to the periodic tip-sample contact, is used as a feedback parameter for topography imaging. Along with the amplitude shift, the phase shift between the driving frequency (from piezo-controllers) and the actual cantilever oscillations can be recorded.168 In the non-contact mode, the tip is far from the sample, at distances where attractive potentials dominate. The cantilever is maintained near its resonance frequency so that the frequency shift is used as the feedback parameter in imaging. Since the tip does not come into direct contact with the sample surface as in the two aforementioned cases, the non-contact mode is almost completely non-destructive for the sample. However, this also makes it very sensitive to the possible surface contaminations which might mask the topography of the sample. Therefore, the measurements in this mode are commonly performed in vacuum conditions.167

4.4.1 Atomic force microscopy infrared spectroscopy Atomic force microscopy infrared spectroscopy (AFM-IR) is an emerging technique that combines the chemical analysis of the FT-IR spectroscopy and the spectral resolution of AFM, thus allowing highly resolved chemical composition mapping of the surfaces at submicron scale. The first attempt of combined AFM/FT-IR characterization was done by Hammiche et al.169 in 1999. While running the AFM in contact mode, a narrow sample region, directly below the AFM tip, was heated with continuous, low intensity IR radiation from a conventional FT-IR spectrometer source. Upon heating of the sample, the resulting temperature increase would change the electrical resistance of the AFM tip. The electrical resistance was measured via Wheatstone bridge and analysed through a Fourier transform algorithm. An alternative approach, used in modern AFM-IR, was developed by Anderson170 around the same time. Instead of measuring the electrical resistance change due to the temperature gradient, Anderson had measured the AFM tip deflection, which was due to the direct contact of the tip with the thermally expanded sample. Both approaches had a poor spatial

40

resolution, as the tip would be affected by the thermally induced changes from the area around the targeted point, due to the heat dissipation. Improvement was achieved by Lu et al.171 in 2014. They had introduced the quantum cascade laser (QCL) as an IR radiation source. By heating the sample region with short IR pulses from a QCL source, they were able to reduce the heat dissipation to the area just below the AFM tip. This allowed the utilization of the AFM spatial resolution. By tuning the IR pulse repetition rate to the AFM cantilever resonance frequency, Lu et al. were able to increase the IR detection efficiency (i.e. probe sensitivity). This is called the resonance enhanced AFM-IR. The AFM-IR works in the following way. The AFM tip is brought above the targeted spot and the IR pulse repetition rate is set to the AFM cantilever resonance frequency. As the IR heats the spot under the AFM tip, the material begins to thermally expand. As the AFM tip is in the contact with the sample, thermal expansion will cause the tip to periodically deflect. The change in the deflection is tracked by the photodiode, and the process is repeated by varying the IR frequency of the QCL pulse, while keeping the AFM tip at the same location. The deflection signal is then Fourier transformed, and the local IR spectrum is obtained. If the IR source is kept at the frequency specific to some molecular vibration in the sample (usually obtained from the standard transmission FT-IR measurement), and the tip is scanned across a certain area, it is possible to obtain chemical composition maps.172,173 Just as in the case of standard AFM, the method can be applied in contact mode (described thus far) and tapping mode. In tapping mode, AFM-IR is used for soft or loosely adhesive sample, as direct contact could deform or delaminate the scanned region. Tapping is performed at the cantilever free resonance frequency, but the pulse repetition rate is adjusted to the difference between this free resonance frequency and the 1st harmonic of the AFM cantilever.174

41

Chapter 5: Summary of the papers

Paper I Photo-degradation in air of spin-coated PC60BM and PC70BM films, Vanja Blazinic, Leif K. E. Ericsson, Stela Andrea Muntean, and Ellen Moons Synthetic Metals, 2018, 241, 26-30 DOI:10.1016/j.synthmet.2018.03.021 This paper deals with the comparative study of degradation of PC60BM and PC70BM thin films upon their exposure to simulated sunlight (AM1.5) in ambient air (i.e. photo-oxidation). Thin films were characterized by UV-vis and FT-IR spectroscopy for various degradation times. The PC70BM thin film has a more pronounced loss of absorption in the visible region (400-700 nm), but the intensity decrease of absorption features of PC60BM thin film in the near-UV region (below 400 nm) indicates faster degradation of this fullerene derivative. The FT-IR spectra of thin films show that, upon exposure, both materials develop new absorption features in the carbonyl region. Tracking of carbonyl features’ intensities for different exposure times indicates that the photo-oxidation rate of PC60BM thin film is 10-15% greater than that of PC70BM. Two possible causes for different photo-oxidation rates are suggested: different oxygen diffusion rate through the film, due to the differences in molecular packing of the molecules or differences in the inherent stability of individual fullerene derivative molecules. Both the molecular packing and the stability are affected by the curvature of fullerene cages. The difference in materials’ densities, and hence molecular packing and oxygen diffusion rates are too small to account for the difference in photo-oxidation rates. Therefore, we conclude that the photochemical stability of individual molecules plays a more significant role.

42

Paper II Impact of intentional photo-oxidation of donor polymer and PC70BM acceptor on solar cell performance Vanja Blazinic, Leif K. E. Ericsson, Igal Levine, Rickard Hansson, Andreas Opitz, and Ellen Moons Physical Chemistry Chemical Physics, 2019, 21, 22259-22271 DOI: 10.1039/C9CP04384E In this paper, the impact of photo-oxidation of TQ1:PC70BM (v/v 1:3 ratio) photoactive layer, spin coated on ITO/MoO3 substrates, on the device performance is investigated. Upon photo-oxidation, LiF/Al cathode was evaporated. J-V measurements show that the exposure of photoactive layer has a detrimental effect on OSC device performance – short circuit current JSC, open circuit voltage VOC and fill factor FF decrease and devices’ series resistance RS increases with the increasing active layer photo-oxidation time. The EQE shows the loss in current generation through the entire recorded spectrum, with the loss more pronounced in the visible-near-IR region for shorter exposure times (less than 2 hours). The thin films of TQ1, PC70BM and their 1:3 blend were characterized with UV-vis and FT-IR spectroscopy before and upon photo-oxidation for the same times. FT-IR spectra indicate development of the carbonyl products both in pure TQ1 and PC70BM films, as well as the blend films, upon exposure. Comparison of EQE spectra of complete device and UV-vis spectra of the photoactive layer upon exposure indicates that the photobleaching of the photoactive layer is not the major contribution to the degraded device performance, but that the charge transport and collection is also hampered. Since the TQ1 is known to be electronically stable from previous NEXAFS study, focus is turned to the electronic properties of the fullerene derivative PC70BM. The C1s NEXAFS spectra show that in photo-oxidized PC70BM thin films, the density of unoccupied molecular states, expressed by the intensity of the π* resonances, decrease with increased exposure time, leading to the loss of the C70 cage conjugation. This effect is more pronounced at the surface than in the bulk material. Further study of the PC70BM thin film surface, with Kelvin probe, indicates that the photo-oxidation leads to the formation of filled gap states near the HOMO and upward band bending. It is concluded that the loss of conjugation, appearance of the gap states, which serve as

43

charge traps, and the degradation of charge transport and collection properties in the PC70BM are the main causes of the device degradation. Paper III The Photooxidation of PC60BM: New Insights from Spectroscopy Iulia Emilia Brumboiu, Leif K. E. Ericsson, Vanja Blazinic, Rickard Hansson, Andreas Opitz, Barbara Brena, and Ellen Moons manuscript In this work, a spectroscopic study of a spin coated thin films of PC60BM, upon their exposure to light and air, was performed by the C1s and O1s core level XPS, the C K-edge and O K-edge NEXAFS, and FT-IR, in order to gain an insight into photo-oxidation of PC60BM. Obtained spectroscopic results were compared with the corresponding calculated spectra for a large variety of possible photo-oxidation products of PC60BM in order to identify the chemical nature of the products observed in the experiment. From the C1s XPS and FT-IR we identify C=O bond types as degradation products. We further show from C1s NEXAFS spectra, both theoretical and experimental, that photo-oxidation causes breaking of C=C bonds of the C60 cage leading to the loss of conjugation of the fullerene derivative. The O1s NEXAFS spectra, both theoretical and experimental, show that the photo-oxidation products have a significant C=O character. From the comparison of theoretical and experimental FT-IR spectra, it is deduced that the photo-oxidation of PC60BM leads to opening of the cage, and formation of products which are a combination of PC58BM anhydride and PC58BM dicarbonyl. The experimental XPS and NEXAFS results show a match with the calculated spectra for a combination of pristine PC60BM, PC58BM anhydride and PC58BM dicarbonyl. Furthermore, the O1s XPS spectra indicate a contribution of the molecular oxygen diffused into the film.

44

Paper IV Effects of morphology on the photodegradation of TQ1:PC70BM films: an AFM-IR study Leif. K. E. Ericsson, Vanja Blazinic, and Ellen Moons manuscript In this manuscript, an emerging method, Atomic Force Microscopy Infrared spectroscopy (AFM-IR), is applied to the well-studied polymer/fullerene TQ1:PC70BM (1:3) blend. The TQ1:PC70BM thin films were spin coated from chlorobenzene, known to form phase separated PC70BM-rich domains with a submicron scale. Blend films were degraded under the simulated sunlight in ambient air. By using the individual components’ molecular IR fingerprint, known from previous FT-IR study, we were able to chemically map the distribution of TQ1 and PC70BM in the phase separated blend film with a nanometer spatial resolution. Furthermore, by adjusting the IR laser frequency to 1780 cm-1, which corresponds to a vibration of the anhydride photo-oxidation product on the PC70BM cage, we mapped the distribution of photo-oxidation products in the intentionally degraded film. We established that the anhydride photo-oxidation product of PC70BM is uniformly distributed over the blend film surface. Paper V Stability of TQ1:N2200 active layers for all-polymer solar cells Vanja Blazinic, Leif. K. E. Ericsson, Ellen Moons manuscript In this manuscript, the photo-oxidation of un-annealed and annealed TQ1, N2200 and their (2:1) blend is investigated with FT-IR, UV-vis and AFM. UV-vis spectra of neat polymer films shows that un-annealed TQ1 film is more stable than un-annealed N2200 film against photobleaching. Thermal annealing of both neat films slow the photobleaching down to a similar rate. The FT-IR spectroscopy results show that, on the other hand, the un-annealed N2200 film is more stable against formation of carbonyl photo-oxidation products, as compared to un-annealed TQ1. Thermal annealing slows down the photochemical changes, due to the exposure to light and air, of the TQ1

45

thin film, while the photochemical effects of the exposure of N2200 film remains largely unchanged. Comparison of neat un-annealed polymers’ FT-IR spectra with that of un-annealed blend in the carbonyl region indicates that the TQ1 component is contributing to the photo-oxidation of the blend more than N2200 component. Thermal annealing of neat polymer films and their 2:1 blends increases their stability against photo-oxidation, as shown both from FT-IR and UV-vis spectroscopy. The differences in UV-vis and FT-IR results of un-annealed neat polymer films indicates that TQ1 and N2200 undergo different photodegradation processes when exposed to light and ambient atmosphere. AFM imaging of the surface of the samples before and upon photo-oxidation does not show any observable changes in the surface morphology.

46

Chapter 6: Conclusions and outlook

In the work presented in this thesis, impact of exposure to light and air (i.e. photo-oxidation) on active layer components, namely the electron acceptor materials, was investigated. Two commonly used electron acceptor materials, the fullerene derivatives PC60BM and PC70BM, were investigated. Photostability in ambient air of spin coated thin films of these materials was comparatively studied by readily available spectroscopic methods, UV-vis absorption spectroscopy and FT-IR spectroscopy. The FT-IR results show that these fullerene derivatives share a similar photochemical degradation pathway upon the exposure to air and light, with the degradation rate being higher in PC60BM. It was found that degradation rate was governed by the individual chemical reactivity of molecules, and not by materials’ density and oxygen diffusion. The greater chemical reactivity of PC60BM was found to be directly related to the greater curvature of C60 cage. In order to investigate how the photo-oxidation affects the performance of polymer/fullerene solar cell, the active layer, based on TQ1:PC70BM blend, was spin coated onto ITO/MoO3 electrode and intentionally exposed to simulated sunlight and air, prior to LiF/Al electrode deposition. J-V measurements under illumination show that performance of these devices drops rapidly with the exposure time. Comparison of various spectroscopic results has distinguished 3 key issues: 1) degradation of charge transport and extraction properties of the PC70BM is plays a dominant role in device performance degradation, and the light absorption, 2) photo-oxidation of PC70BM, hampering charge transport and extraction, occurs mostly at the film surface, 3) surface gap states and band bending are responsible for poor device performance, even if the bulk material may be unaffected. We returned to PC60BM, in order to determine how the modification of the fullerene derivative occurs during the photo-oxidation. By comparing the spectroscopic results from XPS, NEXAFS and FT-IR with the calculated spectra for variety of photo-oxidation products, we have determined that the photodegradation of PC60BM films occurs due to the breaking of the conjugation, followed by opening of the C60 cage, and formation of dicarbonyl and anhydride photo-oxidation

47

products, along with the with the presence of the molecular oxygen in the material. As it was established that the photo-oxidation at the surface of the active layer hinders the PSC device performance, we employed an emerging spectro-microscopic method, AFM-IR, to investigate the photo-oxidation of TQ1:PC70BM films. We demonstrated the ability of this method to chemically and spatially map the distribution of photo-oxidation sites at the surface of the active layer. After investigating the photo-oxidation of the fullerene derivative degradation on bulk, surface and molecular level, and its influence on the device performance, it was replaced with a commonly used electron acceptor polymer from the family of naphthalene diimide polymers, N2200. The comparative study of impact of exposure to light and ambient atmosphere on thin films of TQ1 and N2200 indicates that these two polymers, one being an electron donor and other an electron acceptor, photodegrade in different ways. As the electron acceptor polymers become more widely used, it will be important to investigate how these materials degrade. Understanding their degradation pathways should provide us with the guidelines how to chemically tailor them to be more stable and more efficient.

48

Bibliography

1. Shafiee, S. & Topal, E. When will fossil fuel reserves be diminished? Energy Policy 37, 181–189 (2009).

2. Doron, P., Karni, J. & Slocum, A. A generalized approach for selecting solar energy system configurations for a wide range of applications. MRS Energy Sustain. 6, E11 (2019).

3. Keller, D. P. et al. The Effects of Carbon Dioxide Removal on the Carbon Cycle. Curr. Clim. Chang. Reports 4, 250–265 (2018).

4. Stroeve, J. & Notz, D. Changing state of Arctic sea ice across all seasons. Environ. Res. Lett. 13, 103001 (2018).

5. Walker, W. H., Meléndez‐Fernández, O. H ., Nelson, R. J. & Reiter, R. J. Global climate change and invariable photoperiods: A mismatch that jeopardizes animal fitness. Ecol. Evol. ece3.5537 (2019). doi:10.1002/ece3.5537

6. IRENA. Renewable Power Generation Costs in 2018. (2019). 7. Breyer, C. et al. Solar photovoltaics demand for the global energy

transition in the power sector. Prog. Photovoltaics Res. Appl. 26, 505–523 (2018).

8. IRENA. Measurement and estimation of off-grid solar , hydro and biogas energy. (2018).

9. Green, M. A. et al. Solar cell efficiency tables (Version 53). Prog. Photovoltaics Res. Appl. 27, 3–12 (2019).

10. Fthenakis, V. M., Kim, H. C. & Alsema, E. Emissions from Photovoltaic Life Cycles. Environ. Sci. Technol. 42, 2168–2174 (2008).

11. Branker, K., Pathak, M. J. M. & Pearce, J. M. A review of solar photovoltaic levelized cost of electricity. Renew. Sustain. Energy Rev. 15, 4470–4482 (2011).

12. Bermudez, V. An overview on electrodeposited Cu(In,Ga)(Se,S)2 thin films for photovoltaic devices. Sol. Energy 175, 2–8 (2018).

13. National Renewable Energy Laboratory (USA). Available at: https://www.nrel.gov/pv/cell-efficiency.html. (Accessed: 13th September 2019)

14. Wang, R. et al. Opportunities and Challenges of Lead-Free Perovskite Optoelectronic Devices. Trends Chem. 1, 368–379 (2019).

15. Drachev, A. I. & Gil’man, A. B. The synthesis of semiconducting polymers in a low-temperature plasma. Russ. J. Phys. Chem. A 82, 1733–1741 (2008).

16. Petty, M. C. Molecular Electronics: From Principles to Practice. (Wiley, 2008).

17. Xu, L., Lee, Y.-J. & Hsu, J. W. P. Charge collection in bulk heterojunction organic photovoltaic devices: An impedance spectroscopy study. Appl. Phys. Lett. 105, 123904 (2014).

49

18. Krebs, F. C. Fabrication and processing of polymer solar cells: A review of printing and coating techniques. Sol. Energy Mater. Sol. Cells 93, 394–412 (2009).

19. Søndergaard, R., Hösel, M., Angmo, D., Larsen-Olsen, T. T. & Krebs, F. C. Roll-to-roll fabrication of polymer solar cells. Mater. Today 15, 36–49 (2012).

20. Sommer-Larsen, P., Jørgensen, M., Søndergaard, R. R., Hösel, M. & Krebs, F. C. It is all in the Pattern-High-Efficiency Power Extraction from Polymer Solar Cells through High-Voltage Serial Connection. Energy Technol. 1, 15–19 (2013).

21. Jayaraman, T. et al. Recent development on carbon based heterostructures for their applications in energy and environment: A review. J. Ind. Eng. Chem. 64, 16–59 (2018).

22. Bergqvist, J. et al. Asymmetric photocurrent extraction in semitransparent laminated flexible organic solar cells. npj Flex. Electron. 2, 4 (2018).

23. Cui, Y. et al. Over 16% efficiency organic photovoltaic cells enabled by a chlorinated acceptor with increased open-circuit voltages. Nat. Commun. 10, 1–8 (2019).

24. Azzopardi, B. & Mutale, J. Life cycle analysis for future photovoltaic systems using hybrid solar cells. Renew. Sustain. Energy Rev. 14, 1130–1134 (2010).

25. Yan, J. & Saunders, B. R. Third-generation solar cells: a review and comparison of polymer:fullerene, hybrid polymer and perovskite solar cells. RSC Adv. 4, 43286–43314 (2014).

26. Stability and Degradation of Organic and Polymer Solar Cells. (John Wiley & Sons, Ltd, 2012). doi:10.1002/9781119942436

27. Roesch, R., Eberhardt, K.-R., Engmann, S., Gobsch, G. & Hoppe, H. Polymer solar cells with enhanced lifetime by improved electrode stability and sealing. Sol. Energy Mater. Sol. Cells 117, 59–66 (2013).

28. Berger, P. R. & Kim, M. Polymer solar cells: P3HT:PCBM and beyond. J. Renew. Sustain. Energy 10, 013508 (2018).

29. Zhang, Y., Samuel, I. D. W., Wang, T. & Lidzey, D. G. Current Status of Outdoor Lifetime Testing of Organic Photovoltaics. Adv. Sci. 5, 1800434 (2018).

30. Mateker, W. R. & McGehee, M. D. Progress in Understanding Degradation Mechanisms and Improving Stability in Organic Photovoltaics. Adv. Mater. 29, 1603940 (2017).

31. Jørgensen, M. et al. Stability of Polymer Solar Cells. Adv. Mater. 24, 580–612 (2012).

32. Gevorgyan, S. A. et al. Improving, characterizing and predicting the lifetime of organic photovoltaics. J. Phys. D. Appl. Phys. 50, 103001 (2017).

33. de Jong, M. P., van IJzendoorn, L. J. & de Voigt, M. J. A. Stability

50

of the interface between indium-tin-oxide and poly(3,4-ethylenedioxythiophene)/poly(styrenesulfonate) in polymer light-emitting diodes. Appl. Phys. Lett. 77, 2255–2257 (2000).

34. Friedel, B. et al. Effects of Layer Thickness and Annealing of PEDOT:PSS Layers in Organic Photodetectors. Macromolecules 42, 6741–6747 (2009).

35. Chen, S., Manders, J. R., Tsang, S.-W. & So, F. Metal oxides for interface engineering in polymer solar cells. J. Mater. Chem. 22, 24202 (2012).

36. Voroshazi, E. et al. Influence of cathode oxidation via the hole extraction layer in polymer:fullerene solar cells. Org. Electron. 12, 736–744 (2011).

37. Chavhan, S. D. et al. Low temperature processed NiO x hole transport layers for efficient polymer solar cells. Org. Electron. 44, 59–66 (2017).

38. Meyer, J., Khalandovsky, R., Görrn, P. & Kahn, A. MoO3 Films Spin-Coated from a Nanoparticle Suspension for Efficient Hole-Injection in Organic Electronics. Adv. Mater. 23, 70–73 (2011).

39. Ihn, S.-G. et al. ITO-free inverted polymer solar cells using a GZO cathode modified by ZnO. Sol. Energy Mater. Sol. Cells 95, 1610–1614 (2011).

40. Liu, H. et al. Efficient and ultraviolet durable inverted organic solar cells based on an aluminum-doped zinc oxide transparent cathode. Appl. Phys. Lett. 103, (2013).

41. Chen, R. et al. Co-percolating graphene-wrapped silver nanowire network for high performance, highly stable, transparent conducting electrodes. Adv. Funct. Mater. 23, 5150–5158 (2013).

42. Bag, M., Banerjee, S., Faust, R. & Venkataraman, D. Self-healing polymer sealant for encapsulating flexible solar cells. Sol. Energy Mater. Sol. Cells 145, 418–422 (2016).

43. Mozumder, M. S., Mourad, A.-H. I., Pervez, H. & Surkatti, R. Recent developments in multifunctional coatings for solar panel applications: A review. Sol. Energy Mater. Sol. Cells 189, 75–102 (2019).

44. Chen, J.-D. et al. Single-Junction Polymer Solar Cells Exceeding 10% Power Conversion Efficiency. Adv. Mater. 27, 1035–1041 (2015).

45. Wohlers, M. et al. The mechanism of oxidation of fullerenes with molecular oxygen. ACS Div. Fuel Chem. Prepr. 41, 108–112 (1996).

46. Chambon, S., Rivaton, A., Gardette, J.-L. & Firon, M. Photo- and thermal degradation of MDMO-PPV:PCBM blends. Sol. Energy Mater. Sol. Cells 91, 394–398 (2007).

47. Perthué, A. et al. New insights into polymer solar cells stability: The crucial role of PCBM oxidation. J. Mater. Res. 33, 1868–

51

1878 (2018). 48. Kettle, J., Waters, H., Ding, Z., Horie, M. & Smith, G. C. Chemical

changes in PCPDTBT:PCBM solar cells using XPS and TOF-SIMS and use of inverted device structure for improving lifetime performance. Sol. Energy Mater. Sol. Cells 141, 139–147 (2015).

49. Endale, T. et al. Investigation of photodegradation in polymer solar cells blended with different fullerenes derivatives. Sol. Energy Mater. Sol. Cells 123, 150–158 (2014).

50. Speller, E. M. et al. Impact of Aggregation on the Photochemistry of Fullerene Films: Correlating Stability to Triplet Exciton Kinetics. ACS Appl. Mater. Interfaces 9, 22739–22747 (2017).

51. Lee, H. K. H. et al. The role of fullerenes in the environmental stability of polymer:fullerene solar cells. Energy Environ. Sci. 11, 417–428 (2018).

52. Rao, A. M. et al. Photoinduced Polymerization of Solid C60 Films. Science (80-. ). 259, 955–957 (1993).

53. Wang, Y., Holden, J. M., Dong, Z.-H., Bi, X.-X. & Eklund, P. C. Photo-dimerization kinetics in solid C60 films. Chem. Phys. Lett. 211, 341–345 (1993).

54. Dzwilewski, A., Wågberg, T. & Edman, L. Photo-induced and resist-free imprint patterning of fullerene materials for use in functional electronics. J. Am. Chem. Soc. 131, 4006–4011 (2009).

55. Piersimoni, F. et al. Influence of fullerene photodimerization on the PCBM crystallization in polymer: Fullerene bulk heterojunctions under thermal stress. J. Polym. Sci. Part B Polym. Phys. 51, 1209–1214 (2013).

56. Wang, J., Enevold, J. & Edman, L. Photochemical transformation of fullerenes. Adv. Funct. Mater. 23, 3220–3225 (2013).

57. Heumueller, T. et al. Morphological and electrical control of fullerene dimerization determines organic photovoltaic stability. Energy Environ. Sci. 9, 247–256 (2016).

58. Mateker, W. R., Sachs-Quintana, I. T., Burkhard, G. F., Cheacharoen, R. & McGehee, M. D. Minimal Long-Term Intrinsic Degradation Observed in a Polymer Solar Cell Illuminated in an Oxygen-Free Environment. Chem. Mater. 27, 404–407 (2015).

59. Pont, S., Foglia, F., Higgins, A. M., Durrant, J. R. & Cabral, J. T. Stability of Polymer:PCBM Thin Films under Competitive Illumination and Thermal Stress. Adv. Funct. Mater. 28, 1802520 (2018).

60. An, Y. et al. Nonhalogen Solvent-Processed Asymmetric Wide-Bandgap Polymers for Nonfullerene Organic Solar Cells with Over 10% Efficiency. Adv. Funct. Mater. 28, 1706517 (2018).

61. Wang, C., Ni, S., Braun, S., Fahlman, M. & Liu, X. Effects of water vapor and oxygen on non-fullerene small molecule acceptors. J.

52

Mater. Chem. C 7, 879–886 (2019). 62. Gu, X. et al. Comparison of the Morphology Development of

Polymer-Fullerene and Polymer-Polymer Solar Cells during Solution-Shearing Blade Coating. Adv. Energy Mater. 6, 1601225 (2016).

63. Kim, M. et al. Improving the Photovoltaic Performance and Mechanical Stability of Flexible All-Polymer Solar Cells via Tailoring Intermolecular Interactions. Chem. Mater. 31, 5047–5055 (2019).

64. Wadsworth, A. et al. Critical review of the molecular design progress in non-fullerene electron acceptors towards commercially viable organic solar cells. Chem. Soc. Rev. 48, 1596–1625 (2019).

65. Huo, Y., Zhang, H.-L. & Zhan, X. Nonfullerene All-Small-Molecule Organic Solar Cells. ACS Energy Lett. 4, 1241–1250 (2019).

66. Geoghegan, M. & Hadziioannou, G. Polymer Electronics. (OUP Oxford, 2013).

67. Züfle, S., Hansson, R., Katz, E. A. & Moons, E. Initial photo-degradation of PCDTBT:PC70BM solar cells studied under various illumination conditions: Role of the hole transport layer. Sol. Energy 183, 234–239 (2019).

68. Shrotriya, V., Li, G., Yao, Y., Chu, C.-W. & Yang, Y. Transition metal oxides as the buffer layer for polymer photovoltaic cells. Appl. Phys. Lett. 88, 073508 (2006).

69. Brabec, C. J., Shaheen, S. E., Winder, C., Sariciftci, N. S. & Denk, P. Effect of LiF/metal electrodes on the performance of plastic solar cells. Appl. Phys. Lett. 80, 1288–1290 (2002).

70. Kim, H., Shin, M. & Kim, Y. Influence of thermal annealing on the deformation of a lithium fluoride nanolayer in polymer : fullerene solar cells. EPL (Europhysics Lett. 84, 58002 (2008).

71. Knupfer, M. Exciton binding energies in organic semiconductors. Appl. Phys. A 77, 623–626 (2003).

72. Puschnig, P. & Ambrosch-Draxl, C. Excitons in organic semiconductors. Comptes Rendus Phys. 10, 504–513 (2009).

73. Cao, W. & Xue, J. Recent progress in organic photovoltaics: device architecture and optical design. Energy Environ. Sci. 7, 2123 (2014).

74. Kirchartz, T., Bisquert, J., Mora-Sero, I. & Garcia-Belmonte, G. Classification of solar cells according to mechanisms of charge separation and charge collection. Phys. Chem. Chem. Phys. 17, 4007–4014 (2015).

75. Tamai, Y., Ohkita, H., Benten, H. & Ito, S. Exciton Diffusion in Conjugated Polymers: From Fundamental Understanding to Improvement in Photovoltaic Conversion Efficiency. J. Phys.

53

Chem. Lett. 6, 3417–3428 (2015). 76. Mikhnenko, O. V., Blom, P. W. M. & Nguyen, T.-Q. Exciton

diffusion in organic semiconductors. Energy Environ. Sci. 8, 1867–1888 (2015).

77. Nelson, J. Polymer:fullerene bulk heterojunction solar cells. Mater. Today 14, 462–470 (2011).

78. Manda, P. K., Ramaswamy, S. & Dutta, S. Extraction of the Built-in Potential for Organic Solar Cells From Current–Voltage Characteristics. IEEE Trans. Electron Devices 65, 184–190 (2018).

79. Proctor, C. M., Kuik, M. & Nguyen, T.-Q. Charge carrier recombination in organic solar cells. Prog. Polym. Sci. 38, 1941–1960 (2013).

80. Lai, T.-H., Tsang, S.-W., Manders, J. R., Chen, S. & So, F. Properties of interlayer for organic photovoltaics. Mater. Today 16, 424–432 (2013).

81. Tang, C. W. Two‐layer organic photovoltaic cell. Appl. Phys. Lett. 48, 183–185 (1986).

82. Yu, G., Gao, J., Hummelen, J. C., Wudl, F. & Heeger, A. J. Polymer Photovoltaic Cells: Enhanced Efficiencies via a Network of Internal Donor-Acceptor Heterojunctions. Science (80-. ). 270, 1789–1791 (1995).

83. Halls, J. J. M. et al. Efficient photodiodes from interpenetrating polymer networks. Nature 376, 498–500 (1995).

84. Joshi, D., Shivanna, R. & Narayan, K. S. Organic photovoltaics: key photophysical, device and design aspects. J. Mod. Opt. 61, 1703–1713 (2014).

85. Nilsson, S., Bernasik, A., Budkowski, A. & Moons, E. Morphology and Phase Segregation of Spin-Casted Films of Polyfluorene/PCBM Blends. Macromolecules 40, 8291–8301 (2007).

86. Kouijzer, S. et al. Predicting Morphologies of Solution Processed Polymer:Fullerene Blends. J. Am. Chem. Soc. 135, 12057–12067 (2013).

87. Hansson, R. et al. Vertical and lateral morphology effects on solar cell performance for a thiophene–quinoxaline copolymer:PC 70 BM blend. J. Mater. Chem. A 3, 6970–6979 (2015).

88. Lei, Y. et al. Controlling molecular weight of naphthalenediimide-based polymer acceptor P(NDI2OD-T2) for high performance all-polymer solar cells. J. Mater. Sci. Technol. 33, 411–417 (2017).

89. Deshmukh, K. D. et al. Tuning the Molecular Weight of the Electron Accepting Polymer in All-Polymer Solar Cells: Impact on Morphology and Charge Generation. Adv. Funct. Mater. 28, 1707185 (2018).

90. Hansson, R. et al. Opportunities and challenges in probing local

54

composition of organic material blends for photovoltaics. J. Mater. Res. 32, 1982–1992 (2017).

91. Bartelt, J. A. et al. The Importance of Fullerene Percolation in the Mixed Regions of Polymer-Fullerene Bulk Heterojunction Solar Cells. Adv. Energy Mater. 3, 364–374 (2013).

92. Long, G. et al. New Insights into the Correlation between Morphology, Excited State Dynamics, and Device Performance of Small Molecule Organic Solar Cells. Adv. Energy Mater. 6, 1600961 (2016).

93. Norrman, K., Ghanbari-Siahkali, A. & Larsen, N. B. 6 Studies of spin-coated polymer films. Annu. Reports Sect. ‘C’ (Physical Chem. 101, 174 (2005).

94. Heriot, S. Y. & Jones, R. A. L. An interfacial instability in a transient wetting layer leads to lateral phase separation in thin spin-cast polymer-blend films. Nat. Mater. 4, 782–786 (2005).

95. Yada, T., Maejima, T., MasaruAoki & Umesaki, M. Thin-Film Formation by Spin Coating: Characteristics of a Positive Photoresist. Jpn. J. Appl. Phys. 34, 6279–6284 (1995).

96. Budkowski, A., Bernasik, A., Cyganik, P., Rysz, J. & Brenn, R. Surface-directed phase separation in nanometer polymer films: self-stratification and pattern replication. e-Polymers 2, 1–21 (2002).

97. Frigerio, F. et al. Molecular dynamics simulations of the solvent- and thermal history-dependent structure of the PCBM fullerene derivative. J. Mater. Chem. 22, 5434 (2012).

98. Bian, L., Zhu, E., Tang, J., Tang, W. & Zhang, F. Recent progress in the design of narrow bandgap conjugated polymers for high-efficiency organic solar cells. Prog. Polym. Sci. 37, 1292–1331 (2012).

99. Braunecker, W. A. et al. Ethynylene-Linked Donor–Acceptor Alternating Copolymers. Macromolecules 46, 3367–3375 (2013).

100. Wang, E. et al. An Easily Synthesized Blue Polymer for High-Performance Polymer Solar Cells. Adv. Mater. 22, 5240–5244 (2010).

101. Kim, Y., Yeom, H. R., Kim, J. Y. & Yang, C. High-efficiency polymer solar cells with a cost-effective quinoxaline polymer through nanoscale morphology control induced by practical processing additives. Energy Environ. Sci. 6, 1909 (2013).

102. Dutta, G. K. et al. Synthesis of fluorinated analogues of a practical polymer TQ for improved open-circuit voltages in polymer solar cells. Polym. Chem. 5, 2540 (2014).

103. Carlé, J. E. et al. Comparative studies of photochemical cross-linking methods for stabilizing the bulk hetero-junction morphology in polymer solar cells. J. Mater. Chem. 22, 24417–

55

24423 (2012). 104. Sun, C. et al. A low cost and high performance polymer donor

material for polymer solar cells. Nat. Commun. 9, 743 (2018). 105. Chen, Z., Zheng, Y., Yan, H. & Facchetti, A.

Naphthalenedicarboximide- vs Perylenedicarboximide-Based Copolymers. Synthesis and Semiconducting Properties in Bottom-Gate N-Channel Organic Transistors. J. Am. Chem. Soc. 131, 8–9 (2009).

106. Rao, M., Ponce Ortiz, R., Facchetti, A., Marks, T. J. & Narayan, K. S. Studies of Photogenerated Charge Carriers from Donor-Acceptor Interfaces in Organic Field Effect Transistors. Implications for Organic Solar Cells. J. Phys. Chem. C 114, 20609–20613 (2010).

107. Schuettfort, T., Thomsen, L. & McNeill, C. R. Observation of a Distinct Surface Molecular Orientation in Films of a High Mobility Conjugated Polymer. J. Am. Chem. Soc. 135, 1092–1101 (2013).

108. Xia, Y. et al. Inverted all-polymer solar cells based on a quinoxaline–thiophene/naphthalene-diimide polymer blend improved by annealing. J. Mater. Chem. A 4, 3835–3843 (2016).

109. Kang, H. et al. From Fullerene–Polymer to All-Polymer Solar Cells: The Importance of Molecular Packing, Orientation, and Morphology Control. Acc. Chem. Res. 49, 2424–2434 (2016).

110. Porzio, W., Scavia, G., Barba, L., Arrighetti, G. & McNeill, C. R. On the packing and the orientation of P(NDI2OD-T2) at low molecular weight. Eur. Polym. J. 61, 172–185 (2014).

111. Fabiano, S. et al. Role of photoactive layer morphology in high fill factor all-polymer bulk heterojunction solar cells. J. Mater. Chem. 21, 5891 (2011).

112. Li, Z. et al. High Performance All-Polymer Solar Cells by Synergistic Effects of Fine-Tuned Crystallinity and Solvent Annealing. J. Am. Chem. Soc. 138, 10935–10944 (2016).

113. Xu, Y. et al. Ambient Processable and Stable All-Polymer Organic Solar Cells. Adv. Funct. Mater. 29, 1806747 (2019).

114. Li, H. & Zhang, H. The isolated-pentagon rule and nice substructures in fullerenes. Ars Math. Contemp. 15, 487–497 (2018).

115. Kroto, H. W., Heath, J. R., O’Brien, S. C., Curl, R. F. & Smalley, R. E. C60: Buckminsterfullerene. Nature 318, 162–163 (1985).

116. Hare, J. P., Kroto, H. W. & Taylor, R. Reprint of: Preparation and UV/visible spectra of fullerenes C60 and C70. Chem. Phys. Lett. 589, 57–60 (2013).

117. Pichler, T., Matus, M., K?rti, J. & Kuzmany, H. Electronic transitions in KxC60 (0 ? x ? 6) from in situ absorption spectroscopy. Solid State Commun. 81, 859–862 (1992).

56

118. Srdanov, V. I., Lee, C. H. & Sariciftci, N. S. Spectral and photocarrier dynamics in thin films of pristine and alkali-doped C60. Thin Solid Films 257, 233–243 (1995).

119. Pevzner, B., Hebard, a. F. & Dresselhaus, M. S. Role of molecular oxygen and other impurities in the electrical transportand dielectric properties of C60 films. Phys. Rev. B 55, 16439–16449 (1997).

120. Habuchi, H., Nitta, S., Han, D. & Nonomura, S. Localized electronic states related to O2 intercalation and photoirradiation on C60 films and C70 films. J. Appl. Phys. 87, 8580–8588 (2000).

121. Hummelen, J. C. et al. Preparation and Characterization of Fulleroid and Methanofullerene Derivatives. J. Org. Chem. 60, 532–538 (1995).

122. Arbogast, J. W. & Foote, C. S. Photophysical properties of C70. J. Am. Chem. Soc. 113, 8886–8889 (1991).

123. Zhang, F. et al. Influence of PC60BM or PC70BM as electron acceptor on the performance of polymer solar cells. Sol. Energy Mater. Sol. Cells 97, 71–77 (2012).

124. Nogimura, A., Akaike, K., Nakanishi, R., Eguchi, R. & Kanai, K. Electronic structure and surface morphology of [6,6]-phenyl-C71-butyric acid methyl ester films. Org. Electron. 14, 3222–3227 (2013).

125. Eklund, P. C., Rao, A. M., Zhou, P., Wang, Y. & Holden, J. M. Photochemical transformation of C60 and C70 films. Thin Solid Films 257, 185–203 (1995).

126. Wohlers, M., Werner, H., Belz, T., Röhle, T. & Schlögl, R. C60: A host lattice for the intercalation of oxygen? Mikrochim. Acta 125, 401–406 (1997).

127. Anselmo, A. S., Dzwilewski, A., Svensson, K. & Moons, E. Photodegradation of the electronic structure of PCBM and C 60 films in air. Chem. Phys. Lett. 652, 220–224 (2016).

128. E. Brumboiu, I. et al. Near-edge X-ray absorption fine structure study of the C60-derivative PCBM. Chem. Phys. Lett. 568–569, 130–134 (2013).

129. Aqoma, H. et al. 11% Organic Photovoltaic Devices Based on PTB7-Th: PC 71 BM Photoactive Layers and Irradiation-Assisted ZnO Electron Transport Layers. Adv. Sci. 5, (2018).

130. Anctil, A., Babbitt, C. W., Raffaelle, R. P. & Landi, B. J. Material and Energy Intensity of Fullerene Production. Environ. Sci. Technol. 45, 2353–2359 (2011).

131. He, Y. & Li, Y. Fullerene derivative acceptors for high performance polymer solar cells. Phys. Chem. Chem. Phys. 13, 1970–1983 (2011).

132. van Dyk, E. E. & Meyer, E. L. Analysis of the effect of parasitic

57

resistances on the performance of photovoltaic modules. Renew. Energy 29, 333–344 (2004).

133. Facchetti, A. Polymer donor–polymer acceptor (all-polymer) solar cells. Mater. Today 16, 123–132 (2013).

134. Elumalai, N. K. & Uddin, A. Open circuit voltage of organic solar cells: an in-depth review. Energy Environ. Sci. 9, 391–410 (2016).

135. Gao, K. et al. Multi-Length-Scale Morphologies Driven by Mixed Additives in Porphyrin-Based Organic Photovoltaics. Adv. Mater. 28, 4727–4733 (2016).

136. Mukherjee, S., Jiao, X. & Ade, H. Charge Creation and Recombination in Multi-Length Scale Polymer:Fullerene BHJ Solar Cell Morphologies. Adv. Energy Mater. 6, 1600699 (2016).

137. National Renewable Energy Laboratory (USA). Available at: https://www.nrel.gov/grid/solar-resource/spectra-am1.5.html. (Accessed: 8th October 2019)

138. Brenner, T. J. K. et al. White-light bias external quantum efficiency measurements of standard and inverted P3HT : PCBM photovoltaic cells. J. Phys. D. Appl. Phys. 45, 415101 (2012).

139. Wehenkel, D. J., Hendriks, K. H., Wienk, M. M. & Janssen, R. A. J. The effect of bias light on the spectral responsivity of organic solar cells. Org. Electron. 13, 3284–3290 (2012).

140. Bentham Instruments Limited (UK). Available at: https://www.bentham.co.uk/products/systems/systems-search/pve300-photovoltaic-eqe-ipce-and-iqe-solution-16/#resources. (Accessed: 17th June 2019)

141. Atkins, P. W. & Friedman, R. S. Molecular quantum mechanics. (Oxford University Press, 1997).

142. Lambert, J. B., Shurvell, H. F., Lightner, D. A. & Cooks, R. G. Introduction to Organic Spectroscopy. Introduction to Organic Spectroscopy (Macmillan Publishing Company, 1987).

143. Skoog, D. A., Holler, F. J. & Crouch, S. R. Principles of Instrumental Analysis. (Thomson Brooks/Cole, 2007).

144. van Stam, J., Lindqvist, C., Hansson, R., Ericsson, L. & Moons, E. Fluorescence and UV/VIS absorption spectroscopy studies on polymer blend films for photovoltaics. in Physical Chemistry of Interfaces and Nanomaterials XIV (eds. Hayes, S. C. & Bittner, E. R.) 9549, 95490L (2015).

145. Reese, M. O. et al. Photoinduced Degradation of Polymer and Polymer-Fullerene Active Layers: Experiment and Theory. Adv. Funct. Mater. 20, 3476–3483 (2010).

146. Manceau, M. et al. Photochemical stability of π-conjugated polymers for polymer solar cells: a rule of thumb. J. Mater. Chem. 21, 4132 (2011).

147. Hoke, E. T. et al. The Role of Electron Affinity in Determining

58

Whether Fullerenes Catalyze or Inhibit Photooxidation of Polymers for Solar Cells. Adv. Energy Mater. 2, 1351–1357 (2012).

148. Kanai, K. et al. Effect of annealing on the electronic structure of poly(3-hexylthiophene) thin film. Phys. Chem. Chem. Phys. 12, 273–282 (2010).

149. Cook, S. et al. A photophysical study of PCBM thin films. Chem. Phys. Lett. 445, 276–280 (2007).

150. Stuart, B. H. Infrared Spectroscopy: Fundamentals and Applications. (Wiley, 2004).

151. Dettinger, U. et al. FTIR Study of the Impact of PC[60]BM on the Photodegradation of the Low Band Gap Polymer PCPDTBT under O 2 Environment. Chem. Mater. 27, 2299–2308 (2015).

152. Willmott, P. An Introduction to Synchrotron Radiation: Techniques and Applications. (Wiley, 2011).

153. Oura, K., Lifshits, V. G., Saranin, A. A., Zotov, A. V & Katayama, M. Surface Science: An Introduction. (Springer-Verlag, 2013).

154. Lüth, H. Surfaces and interfaces of solid materials. (Springer, 1995).

155. Madogni, V. I. et al. Comparison of degradation mechanisms in organic photovoltaic devices upon exposure to a temperate and a subequatorial climate. Chem. Phys. Lett. 640, 201–214 (2015).

156. Andreoli, E., Liao, K.-S., Haldar, A., Alley, N. J. & Curran, S. A. PPy:PSS as alternative to PEDOT:PSS in organic photovoltaics. Synth. Met. 185–186, 71–78 (2013).

157. Irfan, I., James Turinske, A., Bao, Z. & Gao, Y. Work function recovery of air exposed molybdenum oxide thin films. Appl. Phys. Lett. 101, (2012).

158. Stöhr, J. NEXAFS Spectroscopy. Springer series in surface sciences 25 25, (Springer, 1992).

159. Watts, B., Thomsen, L. & Dastoor, P. C. Methods in carbon K-edge NEXAFS: Experiment and analysis. J. Electron Spectros. Relat. Phenomena 151, 105–120 (2006).

160. Nahid, M. M., Gann, E., Thomsen, L. & McNeill, C. R. NEXAFS spectroscopy of conjugated polymers. Eur. Polym. J. 81, 532–554 (2016).

161. Zangwill, A. Physics at Surfaces. (Cambridge University Press, 1988).

162. Lee, S. S. & Loo, Y.-L. Structural Complexities in the Active Layers of Organic Electronics. Annu. Rev. Chem. Biomol. Eng. 1, 59–78 (2010).

163. Schuettfort, T. et al. Surface and Bulk Structural Characterization of a High-Mobility Electron-Transporting Polymer. Macromolecules 44, 1530–1539 (2011).

164. Bazylewski, P. F. et al. Side-chain effects on electronic structure

59

and molecular stacking arrangement of PCBM spin-coated films. Chem. Phys. Lett. 508, 90–94 (2011).

165. Giussani, E., Fazzi, D., Brambilla, L., Caironi, M. & Castiglioni, C. Molecular Level Investigation of the Film Structure of a High Electron Mobility Copolymer via Vibrational Spectroscopy. Macromolecules 46, 2658–2670 (2013).

166. Wohlers, M. et al. Reaction of C60 and C70 with molecular oxygen. Synth. Met. 77, 299–302 (1996).

167. Meyer, E., Hug, H. J. & Bennewitz, R. Scanning Probe Microscopy: The Lab on a Tip. (Springer Berlin Heidelberg, 2003).

168. Pérez, R. & García, R. Dynamic atomic force microscopy methods - Surface Science Reports.pdf. Surface Science Reports 47, (2002).

169. Hammiche, A. et al. Photothermal FT-IR Spectroscopy: A Step towards FT-IR Microscopy at a Resolution Better Than the Diffraction Limit. Appl. Spectrosc. 53, 810–815 (1999).

170. Anderson, M. S. Infrared Spectroscopy with an Atomic Force Microscope. Appl. Spectrosc. 54, 349–352 (2000).

171. Lu, F., Jin, M. & Belkin, M. A. Tip-enhanced infrared nanospectroscopy via molecular expansion force detection. Nat. Photonics 8, 307–312 (2014).

172. Dazzi, A. et al. AFM–IR: Combining Atomic Force Microscopy and Infrared Spectroscopy for Nanoscale Chemical Characterization. Appl. Spectrosc. 66, 1365–1384 (2012).

173. Dazzi, A. & Prater, C. B. AFM-IR: Technology and Applications in Nanoscale Infrared Spectroscopy and Chemical Imaging. Chem. Rev. 117, 5146–5173 (2017).

174. Mathurin, J. et al. How to unravel the chemical structure and component localization of individual drug-loaded polymeric nanoparticles by using tapping AFM-IR. Analyst 143, 5940–5949 (2018).

Probing the effects of photodegradation of acceptor materials in polymer solar cells: bulk, surface, and molecular level

Increase of the global energy demand and the climate change are two factors motivating the study and use of renewable energy sources, such as the solar energy. Organic photovoltaics (OPV) is a technology that uses organic molecules to convert solar energy into electricity. These organic molecules can be kept in ink form, allowing OPV device manufacture via coating, and ultimately roll-to-roll printing techniques, resulting in inexpensive, light weight, portable, and mechanically flexible sources of electricity. OPV devices have reached over 15% in power conversion efficiency, but their operational lifetime has to increase.

In this work, the photostability of the active layer in organic solar cells and its molecular components was studied by a variety of spectroscopy, microscopy and electrical characterization techniques, with focus on the chemical changes that these materials undergo during exposure to light and air. The aim was to determine the relation between materials’ degradation and the device performance degradation.

DOCTORAL THESIS | Karlstad University Studies | 2019:30

ISSN 1403-8099

ISBN 978-91-7867-064-2 (pdf)

ISBN 978-91-7867-054-3 (print)