Transient Absorption Spectroscopy of Organic Semiconductors

(1)

Master Physics and Astronomy

Research Project 1

Charge Carrier Dynamics in Block

Copolymer Organic Semiconductors

Laura E. Schleeper BSc

February 9, 2020

Supervisor: dr. E.L. von Hauff

Second examiner: dr. R. Frese

(2)

Abstract

In the past three years, significant efficiency improvements have been observed in single component active layers in organic photovoltaic devices. Such single component active layers can consist, among other materials, of block copolymers. Block copolymers facil-itates spatial and electronic control by changing molecular weight, volume fraction and intermolecular interactions. However, it is difficult to predict the influence of molecu-lar connectivity on interfacial charge dynamics in block copolymers. In this thesis, we study photon-to-charge conversion in the block copolymer Poly(3-decylthiophene-2,5-diyl-block-isocyanide)(P(3 DT-b-ICP 4)). The photon-to-charge conversion is studied by two experiments, namely time-correlated single photon counting and ultrafast tran-sient absorption spectroscopy. We observe faster fluorescence and absorption quenching of the P(3DT-b-ICP 4) relative to P3DT. Additionally, the global analysis shows efficient early time photon-to-singlet exciton conversion in P(3DT-b-ICP 4). Consid-ering the fast quenching along with efficient early time singlet exciton decay followed by a constant charge signal by the global analysis, makes us confident to conclude that efficient photon-to-charge conversion is not observed in P(3DT-b-ICP 4). However, photons are converted into other excited state pathways, which can be either CTS, triplet or polaron formation.

Title: Charge Carrier Dynamics in Block Copolymer Organic Semiconductors Author: Laura E. Schleeper BSc, laura.schleeper@student.uva.nl

UvA Student Number: 10587462 VU Student Number: 2532497 Supervisor: dr. E.L. von Hauff Second examiner: dr. R. Frese Size: 42 EC

Submission date: February 9, 2020

Conducted between September 1st, 2018 and January, 2020.

Research conducted in this thesis was performed at Imperial College London, Ultrafast Optoelectronics Group, Exhibition Road, London SW7 2AZ, United Kingdom.

Faculty of Science

Universiteit van Amsterdam

Science Park 904, 1098 XH Amsterdam http://www.uva.nl/fnwi/

Faculty of Science

Vrije Universiteit Amsterdam

De Boelelaan 1081, 1081 HV Amsterdam https://beta.vu.nl/

(3)

1 Introduction

Sustainable energy technologies are now more needed than ever [1]. The main energy resource of the past decades, i.e. fossil fuels, have caused climate change, of which we recently witnessed the consequences. For example, the current wildfires in Australia have destroyed an area three times the size of the Netherlands and have taken wildlife as well as human lives [2]. Among other sustainable energy resources, sunlight is the most abundant, safe and cleanest resource we have on this planet [1]. For this reason, photovoltaic technologies have been studied by scientist and companies for decades [3]. The third generation photovoltaic solar cell consists of, among other devices, organic photovoltaic (OPV) cells [4]. In contrast to silicon solar cells, OPV cells use carbon-based semiconducting molecules in π-conjugated systems to generate power. OPV cells have several advantages over silicon solar cells, i.e. they are dye-based, flexible, have low weight and low production costs [5]. Nevertheless, research is needed to improve the efficiency and stability of the active layers [6, 7]. The research conducted in this Master Project contributes to the efficiency improvements of OPVs.

Efficiency is, among other things, affected by the morphology of the OPV solar cell’s active layer [8, 9]. The active layer consist of semiconductors with electron-donating and accepting properties. The donor and acceptor modities have to blend on a nanometer length scale to facilitate efficient photon-to-charge conversion [10]. However, morphology control at such a small scale is challenging in current OPV [8, 10]. Therefore, scientists are looking for new types of active layer materials, such as materials that have shown efficient morphology control in other applications. Nevertheless, there is renewed inter-est in the type of single component photoactive material [11], due to significant increase in efficiency up to 5.58% in the past 3 years [12, 13]. The block copolymer (BCP ) used as active layer in OPV consist of a single photoactive material, which has shown a thermal stable morphology [14]. In general, BCP facilitates spatial and electronic control by changing molecular weight, volume fraction and intermolecular interactions [6]. However, the conjugated version of BCP needed for OPV comes with synthetic challenges. It is difficult to predict the influence of molecular connectivity on interfacial charge dynamics [6, 7, 14]. Therefore, we need further contributions to enhance synthe-sis protocols.

The contribution of this report to enhance synthesis protocols is an optical investigation of photon-to-charge conversion in the conjugated BCP P(3DT-b-ICP 4). The charge carrier dynamics are studied for P(3DT-b-ICP 4) with different donor and acceptor ratios. The kinds of measurements for this project consist of obtaining time-resolved

(6)

emission and absorption spectra from transient fluorescence and transient absorption spectroscopic experiments. This thesis aims to answer the question ’Do conjugated BCP P(3DT-b-ICP 4) enable photon-to-charge conversion?’.

(7)

2 Theory

Efficiency in organic photovoltaic devices is, among other things, determined by the rate of photon-to-charge conversion, i.e. Internal Quantum Efficiency (IQE). Every other path an electron in the excited state takes contributes to the losses to the performance of the solar cell. The active layer of photovoltaic devices consist of organic semiconductors. The first subsection introduces organic semiconductors. Next, optical transitions in organic semiconductors are discussed. Finally, I discuss how single component BCP could improve the rate of the favoured path, namely photon-to-charge conversion in organic semiconductor (OSC).

2.1 Organic semiconductors

Organic semiconductors used in photovoltaic devices obtain their semiconducting prop-erties from carbon based conjugated molecules. The electronic structure of the carbon atoms in the backbone of molecules allows to form hybrid orbitals. The electronic struc-ture of carbon atoms is given by 1s2_2s2_p1

xp1ypz. In particular, sp2 hybrid orbitals are

formed, i.e. bond between a s and p orbital of the same carbon atom. Parts of the sp2 hybrid orbitals in carbon atoms are occupied by electrons.

The electron configuration of a sp2 hybridised ethane molecule in the ground state is depicted in fig. 2.1A [15]. The carbon atoms form three σ-bonds with direct surround-ing atoms. Additionally, the pz orbital forms delocalized π-orbitals [16]. The highest

occupied molecular orbital (HOMO), along with the lowest unoccupied molecular or-bital (LUMO), of an ethane molecule in the ground state are indicated in the fig. 2.1A. Hence, the electron configuration depicts negative values below vacum for the σ-orbitals and π-orbitals, due orbital- nuclei binding. Nevertheless, the electron configuration is unable to show effects on the energy levels as a result of electron-electron interaction. For instance, different energy levels for singlet and triplet states can not be depicted in the electron configuration. Therefore, we need an extra diagram to depict the effects of electron-electron interactions on energy levels of the molecule, i.e. state diagram. The state diagram is introduced in the section 2.2 to explain optical transitions in organic semiconductors.

Fig. 2.1B is a simplified picture of 2.1A, which shows only the frontier orbitals. Electrons can be removed from or added to the HOMO or LUMO of the molecule, respectively, referred to as ionisation of the molecule. The energy required to remove an electron from

(8)

Figure 2.1: (A) Electron configuration ethane molecule in the ground state. Ethane in the ground state is sp2 hybridised. The lowest unocupied (HOMO) and highst occupied molecular orbitals (LUMO) are indicated. (B) LUMO and HOMO energy level of a organic semiconductor with corresponding electron affinity (EA) and ionsisation potential (IP).

the HOMO of a molecule in the ground state to vacuum is called ionisation potential (IP) (see fig. 2.1B) [17]. The energy released by adding an electron from vacuum into the LUMO of the molecule in ground state is called the electron affinity (AE) (see fig. 2.1B).

2.1.1 Free charge carriers

The difference between IP and AE is the energy needed to create free charge carriers (Ef), i.e. the energy to overcome the Coulomb attraction between the electron-hole pair.

The Ef can be described by the terms of IP and the EA namely

Ef inal = IP + EA. (2.1)

The Ef exceeds the energy of the absorbed photon that excites the molecule. As a result,

absorption of a photon by the molecule does not directly induces free charge carriers. In order to create free charge, an additional mechanism is required a part from photon absorption to dissociate the electron-hole pair.

A mechanism to enhance electron-hole dissociation is the addition of an OSC. The dissociation is enhanced if one OSCs has a lower IP and EA relative to the other OSC. Although incorrect, most literature uses the simplified one-electron picture of one OSC with lower HOMO and LUMO levels than the other OSC (see fig. 2.2). The OSCs with

(9)

Figure 2.2: Type-II heteronjunction to enhance electron-hole dissociation in OSCs. OSC with the highest HOMO and LUMO levels is referred to as electron donor, whereas the OSC with the lowest HOMO and LUMO levels is called electron acceptor.

the highest HOMO and LUMO levels is called the electron donor, whereas the other OSC is called the electron acceptor. This aligned arrangement of the donor and acceptor OSC energy levels is called type-II heterojunction [17]. The type-II heterojunction at the interface of the donor and acceptor OSCs is crucial for enhancement of electron-hole dissociation, which is explained in more detail in section 2.3.

Once the electron and hole are separated, the internal electric field transports electrons to the electrode [18]. Although the fundamental mechanism of this process is unsolved, experimental research indicates that efficient singlet exciton dissociation is established for OPV if the donor and acceptor moieties blend at a nanometer length scale [6]. The events involved by generation of free charge carriers in OSC can cause polarisation of its surrounding, distort the molecular arrangement and lower the over all energy [16]. The interaction of these three events is considered as a quasi particle, namely polaron [16, 17]. Consequently, absorption and emission spectra of an excited molecule differ from neutral molecules.

2.2 Optical Transitions in Organic Semiconductors

Optical transitions that occur between the absorption and emission of a photon by an organic molecule are schematically shown in a state diagram (see fig. 2.3) [19]. The

(10)

state diagram is a different method to represent energetic states of the molecule, than electron configuration of fig. 2.1. This diagram depicts the ground (S0), lowest singlet

(S1), and lowest triplet (T1) states of a molecule. Additionally, the vibrational energy

levels of the electronic states are indicated by 0, 1, 2 and 3. The S0 is used as a reference

and set to zero energy. Contrary to fig. 2.1, excited state energies, i.e. S1 and T1, have

positive energies.

The state diagram schematically shows how the molecule can shift between the ener-getic states by photon absorption, photon emission or non-radiative processes. Optical transitions are depicted as vertical lines in the energy diagram. These transitions occur at different timescales and different wavelengths.

2.2.1 Absorption to Excited States

Molecules in singlet states follow from the absorption of a photon. Two distinguishable absorption events can occur, namely ground state and excited state absorption. Ground state absorption occurs if a photon is absorbed by the molecule in the ground state, whereas excited state absorption occurs due to photon absorption by the molecule in an excited state. Hence, absorption only results in excited singlet states, which are optically allowed. Molecules in the singlet state follow typically three pathways, namely fluorescence, intersystem crossing followed by phosphorescence and non-radiative decay [20, 21]. Absorption events happen on 10−15s timescale.

2.2.2 Nonradiative processes: Internal Conversion

Excitation of molecules from the ground state mostly results in a transition to higher vibrational levels of the excited state. As a result, the molecules relax to energetically lower vibrational levels emitting heat to its surrounding. The relaxation of molecules is called internal conversion. Internal conversion appear within 10−12 s or less [19].

2.2.3 Radiative Decay: Fluorescence

Emission occurs when the molecule in the excited state falls back into the ground state. We distinguish two emission types based on the excited state they decay from, namely fluorescence and phosphorescence.

Fluorescence follows from the molecule in the singlet state. The electron involved in the singlet state of the molecule have opposite spin, i.e. S = 0. Therefore, electrons in the singlet state of the molecule are allowed to degenerate back to the ground state of the molecule [22]. Hence, the fluorescence spectrum shows strong resemblance to ground state absorption spectra due to the same type of transition. However, from the energy diagram follows that involved singlet states of fluorescence have lower energy than ground

(11)

Figure 2.3: Optical transitions of excited electrons with corresponding timescales. (1) ground state absorption to the first or second excited state. (2) Internal conversion of an electron towards the energetically lowest vibrational mode of that state. (3) Fluorescence of an electron falling back to the ground state while emitting a photon. (4) intersystem crossing of an electron from the first excited state to the triplet state. (5) Phosphorescence of electron in the first triplet state falling to the ground state. (6) charge generation as explained section 2.3.

state absorption. Therefore, fluorescence spectra contain longer wavelengths, referred to as Stokes shift. The fluorescence lifetime is near 10−8 s.

2.2.4 Radiative Decay: Phosphorescence

Phosphorescence is a forbidden optical transition, which follows from decay to the ground state of the molecule in triplet states, i.e. T1 7→ S0. The electrons involved in the T1

state of the molecule have parallel spins, i.e. S = 1. The spin-spin coupling between the electrons results in lower energy level for the triplet state relative to singlet state [20]. However, the parallel spins of the electrons do not allow the molecule to transition from the ground state to the triplet state [22]. The triplet state can only be reach by the non-radiative intersystem crossing of the from the singlet excited state to the energetically lower triplet states. Consequently, phosphorescence has a rate constant several orders smaller than fluorescence, namely around 10−1− 100_s−1 _[19].

(12)

2.3 Excited State Pathways

This section outlines the current understanding of photon-to-charge conversion at the donor/acceptor-interface in OSC (see fig. 2.4) [23]. The first step in this model is the creation of singlet exciton by absorption of a photon. This singlet exciton can follow many paths, of which separation of charge is of the most interest.

2.3.1 Exciton Formation

Singlet excitons are formed by molecular absorption of a photon, which intramolecularly excites an electron from the HOMO to the energetically higher LUMO state. Singlet excitons are mostly excited to higher vibrational modes and relaxed down to the lowest vibrational mode. Electrons in the lowest vibrational excited state mode are bound to the hole in the HOMO, i.e. singlet excitons. Excitons are bound by the Coulomb force, which is given by

Fcoulomb =

e2 4πr0r2

, (2.2)

where e is the electron charge, r dielectric constant of the surrounding medium, 0

the permittivity of vacuum and r the distance between the electron and hole [21]. In silicon-based solar cells, the r = 12, which is compared to the OSC r = 3 − 4 larger.

Hence, the Fcoulomb is inversely proportional to r2. In other words, the larger r and r2,

the weaker the Fcoulomb of the exciton.

2.3.2 Charge Transfer

To create separate charges, the photon-induced singlet excitons have to overcome the Fcoulomb. This is facilitated by singlet exciton diffusion towards the type-II heterojunciton

(13)

Figure 2.4: Photon-to-charge conversion at the interface of electron donor and acceptor molecules [23]. Absorption of a photon by the donor excites an electron from the HOMO to the LUMO. The electron is bound to the resulting hole by the Coulomb force, i.e. singlet exciton. The exciton diffuses towards the donor/acceptor-interface and the exciton undergoes ultrafast charge transfer. As a result, the electron moves to the acceptor into the charge transfer state (CTS). The electron and hole are inter-molecularly bound by the Coulomb force. The larger spatial separation distances between the electron-hole in CTS weakens the Coulomb force. Separated charges are generated if the Coulomb attraction is overcome.

There is firm evidence that, at the D/A-interface, the singlet exciton undergoes ultrafast quasi-adiabatic charge transfer [20, 21]. During this process, the intra-molecular Fcoulomb

is overcome due to larger offset energy levels between the donor and acceptor [21]. As a result, the hole remains on the donor molecule and the electron moves to the acceptor. The electron-hole pair is still bound inter-molecularly by a weaker Fcoulomb as a result of

larger separation distance r. This new state is referred to as charge transfer state (CTS) [23].

The singlet exciton in the CTS causes an external electric field. This electric field induces a quadratics Stark shift on the energy levels of the surrounding molecules [20]. Changes in energy levels result in changing emission and absorption spectra of the surrounding molecules. Mostly, the quadratic Stark effect results in redshift of the absorption spec-trum, due to lowering the optical absorption gap of surrounding molecules [20]. Since charge transfers occurs in order of a few ps, the induced changes in absorption spec-tra are expected to be observed within the same time-scale [20, 23]. In general, it is unpredictable which excited state pathways excitons take.

(14)

2.4 Block copolymers

Photon-to-charge conversion appears in the active layer of OPV devices. The active layer consists of blend donor and acceptor OSC, which result in multiple D/A-interfaces (bulk heterojunction). Well controlled morphology of the active layer in OPV is a critical prerequisite for efficiency improvements of devices [8, 9]. The morphology has to balance the following constraints: (1) exciton diffusion length of 5-10 nm, (2) film thinkness should be larger than 100 nm to provide significant photon absorption and (3) straight continuous paths are required to transport charges towards the electrods [24]. To meet constraint (1) donor and acceptor domains have to phase separate at nm scale. Contrary, constraint (3) requires alignment and continuous paths, which results in a decrease of phase separation.

Scientist are studying materials that allow to control both parameters to optimal con-ditions. In the past three years, single-component active layer materials have shown significant increase in efficiency up to 5.58% [12, 13].

Among other materials, single-component active layers consist of BCP . BCP are a class of polymers. A homopolymer is a macromolecule consisting of 100 or more monomer un-tis, i.e. single conjugated repeating units. Copolymers consist of two or more conjugated repeating units. The arrangement of the conjugated repeating units determines the class of copolymers. When the repeat units of the copolymer form a chain of alternating con-jugated units, i.e. AAAABBBBAAAABBBB, we referred to block copolymers. The blocks in copolymers are covalently bound to one and other.

BCP used as active layer in OPV devices, consists of two alternating conjugated repeat-ing units with respectively electron donor an acceptor properties, i.e. DDDDAAAAD-DDDAAA [11]. Consequently, phase-separation between the donor and acceptor is ob-tained at nm scale. The phase-separation can be controlled by volume fraction, molecular weight and intermolecular interaction [6].

Current research of BCP OSC is focuses on donor material poly(3-alkylthiophene) derivatives with different acceptor molecules [7]. In particular, the donor poly(3-hexyl thiophene) (P3HT) has showed excellent optoelectronic properties, environmental sta-bility and well synthesis [25]. Studies on other donor molecules used in BCP has been limited due to challenges in synthesis [7].1 _{The BCP based on poly(3-alkylthiophene)}

have reported efficiency of 2.3-3.0 % for BCP P3HT-b-PBIT2 by Verduzo et al. [26]. Additionally, Lee et al. measured a PCE of 3.87% for the BCP with benzodithio-phenecarboxy late-based donor copolymer and a naphtalenedicarboximide-selenophene-based acceptor copolymer [7].

An acceptor molecule that has been studied since 1970s are polyisocyanides (PIC), see fig. 2.5. The advantage of PIC is that the backbone contains continuous sp2-hybridised carbon atoms [27]. Nonetheless, the backbone carbon atoms are connected by single

1_{Unfortunately it is beyond the scope of this master thesis to discuss the chemistry synthetic}

(15)

bonds, i.e. cross-conjugated. This results in poor electronic properties for PIC. Schraff et. al have shown that the introduction of functional groups allow to vary the optical bandgap [27]. Additionally, Schraff et. all found a LUMO level of -3.8eV for PIC2. Furthermore, PIC have shown successful synthesis as co-block in block copolymers [28].

Figure 2.5: Chemical structure of polyisocyanide [27].

In this thesis, I will study excited state dynamics of the BCP consisting of the donor poly(3-dithiophene)-block (P3DT) along with a co-block (PIC) acceptor. Zong-Quan Wu et al. reported successful syntheses of block copolymer P3HT with a PIC group [25]. This type of BCP has shown tunable molecular weights and compositions resulting in microphase separation [25]. Moreover, Foster et al. has shown that OSC P3HT blend with PIC positively influence the performance of devices relative to a perylene acceptor. However, Foster et al. reports that future research should focus on optimising geometries for charge collection, by using techniques to obtain ordered PIC strands on the mesoscale [29]. The OSC proprieties along with the ability to synthesis BCP have contributed to investigating the photon-to-charge conversion in the BCP P(3DT-b-IPC4). Next chapter discusses the experiments that allow us to investigate the photon-to-charge conversion in P(3DT-b-IPC4).

(16)

3 Method

To study photon-to-charge conversion in P(3DT-b-ICP 4) I selected two time-resolved spectroscopy techniques. These are Time-Corrolated Single Photon Couting (T CSP C) and transient absorption spectroscopy (T AS). T CSP C measurements investigate time-resolved fluorescence spectra. On the other hand, T AS studies absorption spectra, which allows us to also study time-dependent changes in excited states. The next subsections will discuss the above techniques along with the sample properties.

3.1 Sample properties

Photon-to-charge conversion is studied for the polythiphene homopolyer P3DT and block copolymer organic semiconductor P(3DT-b-ICP 4) provided by partners from the University of Ulm. Additionally, the influence of donor acceptor ratio on excited state dynamics are obtained by varying the thiophene to acceptor ratios of P(3DT-b-ICP 4). The chemical structure of P(3DT-b-ICP 4) is depicted in fig. 3.1.

Table 3.1 shows the different sample types, number average molecular weight (Mn), polydispersity index (PDI) and film preparation conditions. All samples are doctor-bladed on glass from chloroform solution used. No annealing is used. The last column contains the m/n ratios, referring to thiophene to isocyanide ratio. T AS measurements are performed with samples fixed in a quartz cuvette under dry nitrogen atmosphere.

Figure 3.1: Chemical structure of block copolymer organic semiconductor P(3DT-b-ICNP 4).

(17)

Table 3.1

Sample type Mn [kDa] PDI m/n Ratio

Polythiophene Homopolymer P3DT 24.5 1.10 – Polyisocyanide Homopolymer PICNP4 132 2.0 – DA-Blockcopolymer P(3DT-b-ICP4) 23.5 1.03 1:1 DA-Blockcopolymer P(3DT-b-ICP4) 24.5 1.03 2:1 DA-Blockcopolymer P(3DT-b-ICP4) 21.4 1.03 3:1

Figure 3.2 depicts the absorption spectra obtained by JASCO V-670 UV-VIS-NIR Spec-trophotometer of the samples listed in table 3.1. Baseline and background corrections are carried out to compensate for substrate signals. The spectra are normalised to the highest absorption peak, i.e. P3DT. The donor P3DT shows strong absorption for 450-600 nm, whereas the acceptor unit PICNP4 shows overall weak absorption. Excita-tion wavelengths between 425-575 nm are selected in the optical experiments considering sufficient absorption of all samples within this range.

(18)

Figure 3.2: Absorption spectra for neat P3DT, neat PICNP4 and P(3DT-b-ICNP4) at the different ratios as specified in table 3.1. The spectra are normalised to the highest absorption peak, i.e. P3DT.

3.2 Fluorescence measurements

By performing T CSP C, we optically excite the block copolymers and monitor radiative emission. Time-resolved fluorescence spectra provide us information about the charge dynamics in OSC.

3.2.1 Set-Up

Fluorescence spectra are obtained by counting single photons for different time delays relative to the excitation pulse. The time-resolved fluorescence spectra are conducted using DeltaFlex TCSPC system (Horiba Scientific). We excite the sample by a nanoLED at 467nm. The photons were detected at 660 nm with a picosecond photon detector (PPD). The time-resolution is 2.743 × 10−2ns. This time-resolution can not detect CTS.

3.2.2 Data Analysis

To provide statistically reliable data, a minimum limit of 60,000 counts is set. The instrument response function is obtained by the prompt measurement. Furthermore,

(19)

baseline and background corrections are carried out in the data analysis to compensate for the substrate signals.

3.3 Transient Absorption Spectroscopy

Ultrafast excited state dynamics in the OSCs are investigated by T AS measurements (see fig. 3.3) [30]. The high temporal resolution of T AS spectroscopy allows us to study the ultrafast charge transport processes.

3.3.1 Set-Up

T AS measurements require two ultrafast pulses, namely the pump and probe pulse. These two pulses together result in a time-resolved absorption spectrum. The pump pulse excites the sample and the probe pulse traces the excited state pathways. The near-infrared, i.e. 950-1300 nm, probe pulse obtains a time sequence of absorption spectra for different delays (∆τ ) relative to the pump pulse. Varying the τ provides

Figure 3.3: Schematic representation of transient absorption spectroscopy set-up [30]. Charge carrier dynamics are studied using two pulses. First, the pump pulse excites the sample. Then, the second probe pulse traces the photo-induced changes in the OSC. The probe pulse is detected by the spectrometer. The delay-stage regulates the delay time of the probe pulse relative to the pump pulse.

time-resolved absorption spectra of the sample. With a time-resolution in the 100f s time range, the TAS data contains information about excitons, charge transfer and charge formation.

(20)

3.3.2 Absorption Spectra Characteristics

Transient absorption experiments obtain two types of absorption spectra, namely ground state absorption (Aground) and excited state absorption spectra (Aexcited). The Aground

is obtained from BCP s in the ground state, i.e. without the pump pulse. The Aexcited

is obtained from the excite state of the BCP created by the pump and traced by the probe pulse. The difference between the two absorption spectra is calculated by

∆A

A =

Aexcited− Aground

Aground

. (3.1)

The ∆A(λ, τ ) includes contributions from three events, namely ground state bleach (GSB), stimulated emission (SE) and photo-induced absorption (PIA) (see fig. 3.4) [16]. GSB appears due to a change in ground state population. Excited samples absorb less photons in the ground state than the non-excited samples, i.e. the Aexcited < Aground,

which results in ∆A(λ, τ ) < 1. Consequently, GSB has a negative contribution to the ∆A(λ, τ ). SE is established by electrons shifting from the excited state to the ground state. SE only occurs for optical allowed transitions. Since the probe pulse intensity is weak, SE will not affect the excited state population. As a result of the additional SE photons, the SE results in a negative addition to the ∆A(λ, τ ). Hence, SE shows resemblance to the fluorescence spectrum. SE can also emerge from CTS. PIA is a result of absorption by pump excited electrons. Therefore, it has a positive contribution to ∆A(λ, τ ), i.e. Aexcited > Aground [16].

(21)

Figure 3.4: Characteristics of transient absorption spectra [16]. The absorption spectra has contributions of ground state bleach, stimulated emission and photo-induced absorption. The optical transitions, along with the corresponding spectral absorption contributions, are illustrated for these three processes.

3.3.3 Global Analysis

The time-resolved spectra obtained from T AS measurements are two dimensional data-sets. The first dimension is the wavelength, which is an independent experimental spec-tral parameter. The other dimension is an independent experimental parameter to trace spectral change, i.e. delay-time of probe pulse with respect to the pump pulse [31]. The aim of global analysis (GA) is to reveal the dynamics of processes underlying charge carrier dynamics in observed absorption spectral changes [32]. The process underlying the charge carrier dynamics in the ps time domain consists of two components, namely charge and exciton formation.

The GA is carried out by the Genetic Algorithm for the Reconstruction of Spectra written by Simon G´elinas University of Cambridge. In general, this algorithm starts with a population of non-identical species (components), breeding the fittest together. Over time, small mutations are implemented for the fittest species. Repeating this many

(22)

times results in the strongest species (or components). In case of TAS measurements, the spectra obtained by the experiment consist of different excited states: singlet exciton, triplet exciton or polarons.

Data obtained by the TAS is used as input data. The input could be seen as an initial guess of what the spectra of the exciton and charge may look like. To create a popula-tion of spectrum replicas, the input data is perturbed by adding a Gaussian distribupopula-tion in order to differentiate them from the input [31]. Next, the kinetics that best fit the spectra are generated and used to create the next population of spectra. In total 50 popu-lations are generated to identify the underlying exciton and charge absorption spectrum. To reduce the computational time of this algorithm, a spectrum of a pump-probe delay-time after 1 ns is used as a reference for the underlying charge component. A late delay-time reference has been used because no other components are expected to appear after this delay-time [23].

(23)

4 Results and Discussion

To investigate the photon-to-charge conversion in P(3DT-b-ICP 4) we performed time-correlated single photon counting (TCSPC), transient absorption spectroscopy (TAS) experiments and a global analysis (GA). Additionally, we used the electron donor P3DT as a reference to compare the P(3DT-b-ICP 4) results. This chapter presents the results of P(3DT-b-ICP 4) and P3DT of the previous experiments and provides our interpretation of these results.

4.1 Time-correlated single photon counting

Fig. 4.1 compares the T CSP C results for thin films with different ratios of P(3DT-b-ICP 4) and the donor polymer P3DT.

Fluorescence kinetics of P(3DT-b-ICP 4) 1:1, 2:1, 3:1 and P3DT show decay in photon counts over time, i.e. fluorescence quenching. P(3DT-b-ICP 4) 1:1, 2:1, 3:1 show faster fluorescence quenching than P3DT.

Faster fluorescence quenching indicate that less singlet excitons decay to the ground state while excitons diffuse in P(3DT-b-ICP 4) 1:1, 2:1, 3:1 with respect to P3DT [33]. In other words, a larger amount of singlet excitons in the excited states of P(3DT-b-ICP 4) 1:1, 2:1, 3:1 do not radiatively recombine. The larger amount of singlet excitons of that do not radiatively recombine in P(3DT-b-ICP 4) 1:1, 2:1, 3:1 suggest that more sinlget excitons have taken other excited state pathways than singlet excitons of P3DT [33, 34].

Considering the timescale of the TCSPC measurements, we indicate the most likely alternative pathways electrons took in P(3DT-b-ICP 4) 1:1, 2:1, 3:1 that of charge transfer [23]. The electrons of P(3DT-b-ICP 4) 1:1, 2:1, 3:1 transfer from the donor monomer P3DT to the acceptor monomer P(3DT-b-ICP 4) 1:1 [33]. An other path-way may be polaron formation [20, 33].

Furthermore, P(3DT-b-ICP 4) 2:1, 3:1 show faster fluorescence decay, compared to the block copolymer P(3DT-b-ICP 4) 1:1. The P(3DT-b-ICP 4) 2:1, 3:1 films reach half of the photon counts around 0.125 ns, while the block copolymer P(3DT-b-ICP 4) 1:1 reaches half of the photon counts at 1.5 ns. The faster fluorescence decay of P(3DT-b-ICP 4) 2:1, 3:1 may be a results of a higher crystallinity of the acceptor block P3DT-phase in the P(3DT-b-ICP 4) 2:1, 3:1 films. Higher crystallinity of

(24)

the P3DT-phase can result in more self-quenching of P(3DT-b-ICP 4) 2:1, 3:1 than P(3DT-b-ICP 4) 1:1.

Figure 4.1: The time-correlated single photon counting results show faster fluorescence quenching for the block copolymer films P(3DT-b-ICP 4) 1:1, 2:1, 3:1 than the neat P3DT. The faster quenching of the block copolymers suggest singlet exctions take other excited state pathways, i.e. charge transfer, triplet formation or charge separation.

Although TCSPC indicates that singlet excitons in P(3DT-b-ICP 4) 1:1, 2:1, 3:1 films take other excited state pathways than singlet excitons in P3DT, further exper-iments are required to identify which other excited state pathway the singlet excitons in P(3DT-b-ICP 4) 1:1, 2:1, 3:1 may take. In consideration of the decay time-scale, the alternative pathway of electrons in the excited state of the block copolymer can be charge transfer from the electron donor block towards the electron acceptor block [33]. However, additional experiments with a resolution in ps domain are needed to support the indication of electron transfer in P(3DT-b-ICP 4) 1:1, 2:1, 3:1.

Besides, extra experiments are required to determine if photon absorption yields to free charge generation. Evidence for this hypothesis cannot be provided by TCSPC because of three limiting properties. First, TCSPC counts photons at only one wavelength,

(25)

namely 660 nm. This one wavelength provides us information on only a small part emission processes occurring in the films. Second, not all processes as a result of photon excitation are followed by emission. Therefore, we might miss events occurring in the dark. Lastly, the time resolution of the TCSPC is not small enough to confirm charge transfer in block copolymers, nor does the time resolution of the TCSPC provide infor-mation on the end state of the electrons that took an alternative excited state pathway in P(3DT-b-ICP 4) 1:1, 2:1, 3:1.

4.2 Transient Absorption

Transient absorption measurements are preformed for the films P(3DT-b-ICP 4) 1:1, 2:1, 3:1 and P3DT. The next subsections will separately present the results of the transient absorption spectra and kinetics. These results will be discussed in section 4.3 along with the global analysis results.

4.2.1 Spectra

The transient absorption spectra of P(3DT-b-ICP 4) 1:1 and P3DT are depicted in fig. 4.2 (see fig. 4.2). Temporal changes of the spectra are indicated by the color gradient. The P(3DT-b-ICP 4) 1:1 spectra show rapid redshift for early points of ∆τ from 0.2 - 10 ps. Additionally, at ∆τ ≤ 100 ps the normalised spectra show a blue shift. Furthermore, P(3DT-b-ICP 4) 1:1 TAS spectra show low intensities close to zero at ∆τ ≤ 100 ps. Lastly, negative values at wavelengths ≥ 1075 nm are measured for spectra obtained at a later point in time ∆τ ≥ 100 ps.

The P3DT transient absorption spectra showed a blueshift over time. The blueshift indicates increase in absorption energy of the molecule [20]. Furthermore, the intensity decreased gradually at earlier points in time, whereas we observe smaller decreases in intensity for spectra at ∆τ ≥ 100 ps. Further, we observe negative values for the P3DT spectra at ∆τ = 1-5 ns.

(26)

Figure 4.2: Transient absorption spectra of P(3DT-b-ICP 4) 1:1 (red) and P3DT (green) for different pump probe delay times. The bottom transient absorp-tion spectra are normalised with respect to the 0.2-0.5 ps spectrum.

(27)

4.2.2 Kinetics

From the spectra in fig. 4.2 we selected 1200nm and studied the decay in intensity over ∆τ for P(3DT-b-ICP 4) 1:1, 2:1, 3:1 and P3DT (see fig. 4.3). The kinetics in fig. 4.3 confirm the faster decay at earlier points in time observed in the spectra of fig. 4.2.

Figure 4.3: Normalised transient absorption kinetics at 1200 nm of P3DT and for dif-ferent thiophene to acceptor ratios of P(3DT-b-ICP 4).

The P(3DT-b-ICP 4) 1:1, 2:1, 3:1 reached half the intensity within a few ps. Ad-ditionally, the intensity for P(3DT-b-ICP 4) 1:1, 2:1, 3:1 is approximately constant around zero for ∆τ ≥ 300 ps. The P3DT kinetics reach half of the intensity at 1200nm within 50 ps. Moreover, the intensity of P3DT does not stabilise within 900 ps.

In comparison, the kinetics at 1200 nm provides clear evidence that P(3DT-b-ICP 4) 1:1 2:1, 3:1 show faster decay in absorption relative to P3DT. The P(3DT-b-ICP 4) 1:1, 2:1, 3:1 films reach half the absorption intensity within 2 ps, while P3DT reaches half its intensity after 50 ps. This corresponds to the faster decrease in intensity P(3DT-b-ICP 4) 1:1 observed for all wavelengths relative to spectra of P3DT, which are shown in fig. 4.2.

(28)

4.3 Global Analysis Transient Absorption data

A global analysis is performed on the transient absorption data of P(3DT-b-ICP 4) 1:1 and P3DT. The global analysis deconvolutes the transient absorption spectra and ki-netics into the underlying events occurring during photon-to-charge conversion [31]. The indicators to identify these species are the evolution in time and shift in peak position of the spectra, see fig. 4.4. In order to provide an objective description of the global anal-ysis results, the two species are first addressed by their colour, i.e. blue or orange. Once we interpreted these results, I will address the species by the names of our interpretation, i.e. singlet state, triplet state, charge transfer state, or polaron formation.

Figure 4.4: Summary of the global analysis applied to the transient absorption re-sults of the donor P3DT and block copolymer organic semiconductor P(3DT-b-ICP 4). (A) and (B) are normalised absorption spectra of P3DT and P(3DT-b-ICP 4). These spectra indicate two components, namely exciton and charge. (C) and (D) are the normalised kinetics obtained by a global analysis of P3DT and P(3DT-b-ICP 4).

(29)

4.3.1 Global Analysis of P(3DT-b-ICP 4) 1:1

The global analysis produced a deconvoluted spectrum of P(3DT-b-ICP 4) 1:1 (see fig. 4.4B). The two components underlying the transient absorption spectra are indicated in blue and orange. Changes in intensity of the spectra P(3DT-b-ICP 4) 1:1 over time ∆τ are given in fig. 4.4 D. The kinetics of P(3DT-b-ICP 4) 1:1 show decay for both components.

The kinetics of the blue component P(3DT-b-ICP 4) 1:1 reach half their intensity in 2 ps. For ∆τ ≥ 200 ps, the blue component kinetics stabilise close to zero. Singlet exction formation followed by charge transfer appears within a few picoseconds [23]. Therefore, we interpreter the blue data in fig. 4.4 D as singlet exciton formation.

The peak position at high wavelengths of the blue spectrum confirms singlet exciton formation in P(3DT-b-ICP 4) 1:1 [23, 34, 35]. The P(3DT-b-ICP 4) 1:1 spectrum in blue shows a peak at high wavelengths, i.e. around 1275 nm (fig. 4.4 B). The peak position at high wavelength is in agreement with the observed peak positions in fig. 4.2 at short time scales.

The second component of P(3DT-b-ICP 4) 1:1 in orange shows a spectrum with a peak around 975 nm. Peak positions around 975 nm is in line with the observed P(3DT-b-ICP 4) 1:1 peak position for TAS spectra at later ∆τ (fig. 4.2). Since I have selected the orange spectra as input for the global analysis myself, such that the orange spectrum corresponds to results for ∆τ 1-3 ns, I am confident to say that the orange data shows a derivative product of a singlet exciton [20].

This is further supported by the kinetics of the orange P(3DT-b-ICP 4) 1:1 dynamics. The orange component P(3DT-b-ICP 4) 1:1 reaches half its intensity in less than 1 ps, followed by a constant intensity at 0.2 ∆OD. The orange constant signal suggest a derivative product of the singlet exciton. Nevertheless, the identity of this derivative product is debatable. It can be an intermolecularly bound singlet exciton in the CTS, triplet exction or polaron formation [20, 35].

Overall, the singlet exciton derivative product spectrum (orange) is blueshifted relative to the singlet exciton spectrum (blue) of the P(3DT-b-ICP 4) 1:1, which indicates that the absorption energy has increased at later points in time [20]. Increase in ab-sorption energy suggest that P(3DT-b-ICP 4) 1:1 is in an energetically higher excited state than the singlet excited state.

The redshift followed by blueshift is in line with temporal changes in peak position observed in fig. 4.2. Jakowetz et al. (2017) have found a similar pattern of transient absorption peak shift for PCDTBT/mPCBM, i.e. rapid redshift at earlier points in time followed by blue shift after 100 ps [20]. The authors suggest that this corresponds to charge generation in disorder regions (redshift) followed by charges moving towards more ordered regions (blueshift) [20]. Furthermore, Jakowetz et al. (2017) say the early time motion, i.e. redshift, allows the electron-hole pair to efficiently overcome the coulomb attraction within sub-200 fs [20].

(30)

However, our samples consist of a block copolymer rather than a blend of two separate organic semiconductors like Jakowetz et al. (2017). Therefore, we are reluctant to identify the derivative product of the exciton as charge. In our case, the derivative product can also be a polaron rather than separate charges. An additional reason to be reluctant with identification of the exciton derivative product is that intensities of the P(3DT-b-ICP 4) 1:1 spectra are low at ∆τ > 100 ps, where we obtained blue shift. Lastly, the orange spectrum shows negative values at high wavelengths, just like the spectra of fig. 4.2 at ∆τ ≤ 100 ps. The negative values might be a result of artefact in the delay stages. This is an other reason to be careful with the identification of the derivative product, especially for results at later point of ∆τ .

In summary, the blue results show a peak position around 1200 nm along with ultrafast decay, which I assign to ultrafast singlet exciton formation in P(3DT-b-ICP 4) 1:1. The orange spectrum shows a peak around 975 nm in addition to a decrease in intensity followed by a constant signal, which is ascribed to formation of an exciton derivative product. This derivative product can either be a charge transfer state, triplet state or polaron formation. Further experiments are required to identify which of the 3 possible singlet excited state pathways correspond to the orange signal.

4.3.2 Global Analysis of P3DT

Observed changes in intensity of the blue and orange spectra of P3DT are depicted in fig.4.4 C. The blue data reaches half the intensity at 50 ps and stabilises around zero after 1000 ps. Additionally, the blue spectrum of P3DT shows a peak around 1200 nm. The blue signal of the P3DT GA is ascribed to exciton formation, as a results of the relatively fast decay in conjunction with a peak around 1200 nm [34, 36].

The orange P3DT spectrum shows a peak around 1100 nm, which suggests that the derivative product of the singlet exciton increases the energy of the excited state slightly [33]. Furthermore, the orange dynamics of P3DT starts with a constant signal around 0.35 ∆OD untill 5 ps. The constant signal observed at ∆τ = 0 ps suggest that charge is generated upon laser radiation [36]. After 5 ps, the orange P3DT data shows an increase in intensity. After 500 ps, the orange signal decreases. The rapid rise and decrease of the orange signal is interpreted as charge that recombines quickly. The rise in signal appearing at relatively late times along with the followed rapid decay strongly indicates geminate charge recombination [36]. Further, the orange spectrum of P3DT has negative values at high wavelengths. As mentioned before, the negative values might be a result of a artefact in the delay stage.

In summary, the blue GA of P3DT is assigned to singlet exciton formation, due to the peak position around 1200 nm along with fast decay in intensity. The orange GA results of P3DT are interpreted as singlet exciton recombination, as a result of the rapid rise and decay of the orange dynamics. However, the orange results show negative values which may be caused by a artefact in the delay stage. Therefore, we are careful with

(31)

the interpretation of the orange data.

4.3.3 Comparison of P(3DT-b-ICP 4) 1:1 and P3DT

In comparison, both blue spectra of P(3DT-b-ICP 4) 1:1 and P3DT show singlet exciton formation, corresponding to spectrum with a peak around 1200 nm [23, 34, 35]. However, P(3DT-b-ICP 4) 1:1 dynamics show efficient singlet exciton transfer in to other excited pathways than P3DT does, due to faster singlet exciton decay observed for P(3DT-b-ICP 4) 1:1 GA kinetics compared to P3DT. Efficient singlet exciton formation in P(3DT-b-ICP 4) 1:1 is further support by the fast fluorescence quenching. In other words, the presents of the co-clock PIC4 positively contributes to singlet exciton transfer into other excited state pathways.

Moreover, the orange GA results show that singlet excitons take different pathways in P(3DT-b-ICP 4) 1:1 and P3DT. The orange spectra show different peak posi-tions for both films. The peak of the orange P(3DT-b-ICP 4) 1:1 is around 950 nm, whereas the orange spectra of P3DT has a peak at 1100 nm. The different peak positions of the orange GA spectra, is in agreement with the distinct tempo-ral changes of the P(3DT-b-ICP 4) 1:1 and P3DT TAS spectra. In other words, P(3DT-b-ICP 4) 1:1 increases the absorption energy more than neat P3DT com-pared to the exciton peak.

Futhermore, the orange GA kinetics of P(3DT-b-ICP 4) 1:1 and P3DT show different shapes, which supports the distinct excited state pathways the singlet exctions took in the samples. P(3DT-b-ICP 4) 1:1 shows early decay, followed by a constant signal around 0.2 ∆OD at 5 ps. Contrary, P3DT starts with a constant signal around 0.35 ∆OD, followed by an increase from 5 - 500 ps followed by a decrease for ∆τ ≥ 500 ps. In summary, the times scales of the singlet exciton dynamics (blue), in combination with the different shapes of the singlet exciton derivative products (orange) dynam-ics, indicate that different processes occur after singlet exciton formation in the films P(3DT-b-ICP 4) 1:1 and P3DT [35, 36]. We assigned the orange P(3DT-b-ICP 4) 1:1 dynamics to an singlet exciton derivative product. The block copolymer monomer blocks P3DT and ICP 4 cause an exciton derivative product, which cannot be obtained by the P3DT alone.

Nevertheless, further experiments are required to investigate if the orange exciton deriva-tive product observed in P(3DT-b-ICP 4) 1:1 can result in free charges. Transient absorption pump-push-probe experiments can be used to observe the temporal changes of the singlet exciton derivative product. Research should investigate if the push pulse results in singlet exciton dissociation followed by fromation of free charges carriers. Besides, further research could investigate the influence of acceptor materials on the P(3DT-b-ICP 4) 1:1. The P(3DT-b-ICP 4) 1:1 results indicate good donor prop-erties. Therefore, the presence of an acceptor may influences the derivatives of exciton products.

(32)

5 Conclusion

Efficient photon-to-charge conversion is not observed in P(3DT-b-ICP 4) OSC. Nev-ertheless, photon-to-singlet-exciton conversion is observed in P(3DT-b-ICP 4). Sin-glet excitons efficiently diffuse into other excited state pathways. However, the nature of these other pathways is undetermined.

Singlet exciton formation and conversion towards excited state pathways are observed in the TCSPC results. The observed fluorescence quenching is assigned to decay of singlet excitons to the ground state. The faster fluorescence quenching of P(3DT-b-ICP 4) relative to the neat component P3DT indicates singlet exciton diffusion into excited state pathways other than radiative decay.

This is confirmed by the T AS kinetics at 1200 nm and by transient absorption spectra. The ps timescale of transient absorption measurements allows to investigate singlet ex-citon dynamics during diffusion, charge transfer and charge separation. The transient absorption spectra of P(3DT-b-ICP 4) show early time redshift, followed by blueshift at later times, which we ascribed to singlet exciton formation followed by a derivative product of the singlet exciton. This singlet derivative product can either be CTS, triplet formation or polaron formation.

The early time redshift followed by a blueshift is confirmed by the global analysis of transient absorption data. The underlying components of the transient absorption spec-tra confirm early time singlet exciton formation along with a derivative singlet exciton product. The time scales of singlet exciton formation along with the appearance of a constant derivative product signal confirms that the derivative product most likely undergoes charge transfer rather than triplet or polaron formation.

In summary, we conclude that efficient photon-to-charge conversion in P(3DT-b-ICP 4) is not observed by TCSPC and TAS experiments. P(3DT-b-ICP 4) shows efficient sin-glet exciton formation and diffusion into other excited state pathways. Nevertheless, the nature of this excited state pathway remains unrevealed.

Future research is required to investigate the singlet excited state pathway of the exciton derivative product, i.e. CTS, triplet or polaron formation. Experiments such as transient absorption pump-push-probe might be able to push the singlet exciton derivative product out of its current states and might result in free charges. Moreover, P(3DT-b-ICP 4) shows excellent donor properties. Therefore, further research could focus on adding an acceptor molecule to P(3DT-b-ICP 4) and investigate photon-to-charge conversion of this blend.

(33)

6 Acknowledgement

Initially, I would like to express my thanks to Dr. Elizabeth von Hauff for supporting me and by using here network to find me a research group outside the Netherlands. She has been an excellent supervisor for the master thesis by learning me a lot about organic semiconductors. Yet, here ability to guide and mentor students in their personal development towards a true scientist is what makes here stand out from others. I feel fortunate to work with and learn from here.

Also, I would like to thank Dr. Artem Bakulin for the opportunity to be part of his Ultrafast Optoelectronics Group at Imperial College London. He has always challenged me with the Ultrafast Spectroscopic experiments as well as the different data analysis methods he introduced to me. Additonally, I was invited to my first conference on sustainable technologies. Which is something I will never forget. Furthermore, I would like to thank Yifan Dong MSc who has been my daily supervisor at Imperial College London. She was always there to discuss my results and next steps in my research. Also, she has been a great teacher for improving my graph design skills.

Next, I would like to thank Dr. Frank Pammer for providing the samples and his contribution to the paper.

Additionally, I would like to thank the European Commission as well as Madeleine Julie Vervoort Fonds for funding my time and stay at Imperial College London.

Further, I would like to thank Simon G´elinas of the University of Cambridge for providing the Global Analysis tool in MATLAB.

Finally, I want to thank my family and friends for the support. It was great that you visit me during my time in London.

(34)

Bibliography

[1] P. K. Nayak, S. Mahesh, H. J. Snaith, and D. Cahen, “Photovoltaic solar cell technologies: analysing the state of the art,” Nature Reviews Materials, vol. 4, no. 4, p. 269, 2019.

[2] A. Gunia and T. Law. (2020) At least 24 people and millions of animals have been killed by australia’s bushfires. [Online]. Available: https://time.com/5758186/ australia-bushfire-size/

[3] D. M. Chapin, C. Fuller, and G. Pearson, “A new silicon p-n junction photocell for converting solar radiation into electrical power,” Journal of Applied Physics, vol. 25, no. 5, pp. 676–677, 1954.

[4] M. Neukom, S. Z¨ufle, S. Jenatsch, and B. Ruhstaller, “Opto-electronic characteriza-tion of third-generacharacteriza-tion solar cells,” Science and Technology of advanced MaTerialS, vol. 19, no. 1, pp. 291–316, 2018.

[5] G. Li, R. Zhu, and Y. Yang, “Polymer solar cells,” Nature photonics, vol. 6, no. 3, p. 153, 2012.

[6] F. Nubling, T. R. Hopper, B. Kuei, H. Komber, V. Untilova, S. B. Schmidt, M. Brinkmann, E. D. Gomez, A. A. Bakulin, and M. Sommer, “Block junction-functionalized all-conjugated donor–acceptor block copolymers,” ACS applied ma-terials & interfaces, vol. 11, no. 1, pp. 1143–1155, 2018.

[7] J. H. Lee, C. G. Park, A. Kim, H. J. Kim, Y. Kim, S. Park, M. J. Cho, and D. H. Choi, “High-performance polymer solar cell with single active material of fully conjugated block copolymer composed of wide-band gap donor and narrow-band gap acceptor blocks,” ACS applied materials & interfaces, vol. 10, no. 22, pp. 18 974–18 983, 2018.

[8] J. Hou, O. Ingan¨as, R. H. Friend, and F. Gao, “Organic solar cells based on non-fullerene acceptors,” Nature materials, vol. 17, no. 2, pp. 119–128, 2018.

[9] J. Zhang, H. S. Tan, X. Guo, A. Facchetti, and H. Yan, “Material insights and challenges for non-fullerene organic solar cells based on small molecular acceptors,” Nature Energy, vol. 3, no. 9, pp. 720–731, 2018.

[10] P. Hudhomme, “An overview of molecular acceptors for organic solar cells,” EPJ Photovoltaics, vol. 4, p. 40401, 2013.

(35)

[11] J. Roncali and I. Grosu, “The dawn of single material organic solar cells,” Advanced Science, vol. 6, no. 1, p. 1801026, 2019.

[12] Y. He, T. Heum¨uller, W. Lai, G. Feng, A. Classen, X. Du, C. Liu, W. Li, N. Li, and C. J. Brabec, “Evidencing excellent thermal-and photostability for single-component organic solar cells with inherently built-in microstructure,” Advanced Energy Materials, p. 1900409, 2019.

[13] F. Pierini, M. Lanzi, P. Nakielski, S. Paw lowska, O. Urbanek, K. Zembrzycki, and T. A. Kowalewski, “Single-material organic solar cells based on electrospun fullerene-grafted polythiophene nanofibers,” Macromolecules, vol. 50, no. 13, pp. 4972–4981, 2017.

[14] G. Feng, J. Li, F. J. Colberts, M. Li, J. Zhang, F. Yang, Y. Jin, F. Zhang, R. A. Janssen, C. Li et al., ““double-cable” conjugated polymers with linear backbone toward high quantum efficiencies in single-component polymer solar cells,” Journal of the American Chemical Society, vol. 139, no. 51, pp. 18 647–18 656, 2017.

[15] P. Atkins, J. De Paula, and J. Keeler, Atkins’ physical chemistry. Oxford university press, 2018.

[16] J. Zhang, “Ultrafast spectroscopic studies of solution-processing organic photo-voltaic materials and devices,” Ph.D. dissertation, University of Cambridge, 2019. [17] A. K¨ohler and H. B¨assler, Electronic processes in organic semiconductors: An

in-troduction. John Wiley & Sons, 2015.

[18] C. M. Proctor, M. Kuik, and T.-Q. Nguyen, “Charge carrier recombination in organic solar cells,” Progress in Polymer Science, vol. 38, no. 12, pp. 1941–1960, 2013.

[19] J. R. Lakowicz, Principles of fluorescence spectroscopy. Springer Science & Business Media, 2013.

[20] A. C. Jakowetz, M. L. B¨ohm, A. Sadhanala, S. Huettner, A. Rao, and R. H. Friend, “Visualizing excitations at buried heterojunctions in organic semiconductor blends,” Nature materials, vol. 16, no. 5, p. 551, 2017.

[21] J. Kurpiers, T. Ferron, S. Roland, M. Jakoby, T. Thiede, F. Jaiser, S. Albrecht, S. Janietz, B. A. Collins, I. A. Howard et al., “Probing the pathways of free charge generation in organic bulk heterojunction solar cells,” Nature communi-cations, vol. 9, no. 1, p. 2038, 2018.

[22] D. J. Griffiths and D. F. Schroeter, Introduction to quantum mechanics. Cambridge University Press, 2018.

[23] A. A. Bakulin, A. Rao, V. G. Pavelyev, P. H. van Loosdrecht, M. S. Pshenichnikov, D. Niedzialek, J. Cornil, D. Beljonne, and R. H. Friend, “The role of driving energy and delocalized states for charge separation in organic semiconductors,” Science, vol. 335, no. 6074, pp. 1340–1344, 2012.

(36)

[24] M. Sommer, S. Huettner, and M. Thelakkat, “Donor–acceptor block copolymers for photovoltaic applications,” Journal of Materials Chemistry, vol. 20, no. 48, pp. 10 788–10 797, 2010.

[25] Z.-Q. Wu, D.-F. Liu, Y. Wang, N. Liu, J. Yin, Y.-Y. Zhu, L.-Z. Qiu, and Y.-S. Ding, “One pot synthesis of a poly (hexylthiophene)-b-poly (quinoxaline-2, 3-diyl) rod–rod diblock copolymer and its tunable light emission properties,” Polymer Chemistry, vol. 4, no. 17, pp. 4588–4595, 2013.

[26] C. Guo, Y.-H. Lin, M. D. Witman, K. A. Smith, C. Wang, A. Hexemer, J. Strzalka, E. D. Gomez, and R. Verduzco, “Conjugated block copolymer photovoltaics with near 3% efficiency through microphase separation,” Nano letters, vol. 13, no. 6, pp. 2957–2963, 2013.

[27] S. Schraff, Y. Sun, and F. Pammer, “Fulvenyl-functionalized polyisocyanides: Cross-conjugated electrochromic polymers with variable optical and electrochemical properties,” Macromolecules, vol. 51, no. 14, pp. 5323–5335, 2018.

[28] R. J. Ono, A. D. Todd, Z. Hu, D. A. Vanden Bout, and C. W. Bielawski, “Synthesis of a donor–acceptor diblock copolymer via two mechanistically distinct, sequential polymerizations using a single catalyst,” Macromolecular rapid communications, vol. 35, no. 2, pp. 204–209, 2014.

[29] S. Foster, C. E. Finlayson, P. E. Keivanidis, Y.-S. Huang, I. Hwang, R. H. Friend, M. B. Otten, L.-P. Lu, E. Schwartz, R. J. Nolte et al., “Improved performance of perylene-based photovoltaic cells using polyisocyanopeptide arrays,” Macro-molecules, vol. 42, no. 6, pp. 2023–2030, 2009.

[30] A. Bakulin, “Materials characterisation: Introduction to the spectroscopy of elec-tronic matirals (lecture 3+4),” 2018.

[31] I. H. Van Stokkum, D. S. Larsen, and R. Van Grondelle, “Global and target analysis of time-resolved spectra,” Biochimica et Biophysica Acta (BBA)-Bioenergetics, vol. 1657, no. 2-3, pp. 82–104, 2004.

[32] Y. Dong, H. Cha, J. Zhang, E. Pastor, P. S. Tuladhar, I. McCulloch, J. R. Durrant, and A. A. Bakulin, “The binding energy and dynamics of charge-transfer states in organic photovoltaics with low driving force for charge separation,” The Journal of chemical physics, vol. 150, no. 10, p. 104704, 2019.

[33] H. Cha, S. Wheeler, S. Holliday, S. D. Dimitrov, A. Wadsworth, H. H. Lee, D. Baran, I. McCulloch, and J. R. Durrant, “Influence of blend morphology and energetics on charge separation and recombination dynamics in organic solar cells incorporating a nonfullerene acceptor,” Advanced Functional Materials, vol. 28, no. 3, p. 1704389, 2018.

[34] H. Cha, G. Fish, J. Luke, A. Alraddadi, H. H. Lee, W. Zhang, Y. Dong, S. Limbu, A. Wadsworth, I. P. Maria et al., “Suppression of recombination losses in polymer:

(37)

Nonfullerene acceptor organic solar cells due to aggregation dependence of acceptor electron affinity,” Advanced Energy Materials, p. 1901254, 2019.

[35] B. Kan, J. Zhang, F. Liu, X. Wan, C. Li, X. Ke, Y. Wang, H. Feng, Y. Zhang, G. Long et al., “Fine-tuning the energy levels of a nonfullerene small-molecule acceptor to achieve a high short-circuit current and a power conversion efficiency over 12% in organic solar cells,” Advanced Materials, vol. 30, no. 3, p. 1704904, 2018.

[36] K. Kawashima, Y. Tamai, H. Ohkita, I. Osaka, and K. Takimiya, “High-efficiency polymer solar cells with small photon energy loss,” Nature communications, vol. 6, p. 10085, 2015.