ACTAUNIVERSITATIS

UPSALIENSISUPPSALA

2017

Digital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology 1604

Conducting Redox Polymers forElectrode Materials

Synthetic Strategies and Electrochemical Properties

XIAO HUANG

ISSN 1651-6214ISBN 978-91-513-0168-6urn:nbn:se:uu:diva-334562

Dissertation presented at Uppsala University to be publicly examined in B41, BMC,Husargatan, Uppsala, Friday, 19 January 2018 at 09:15 for the degree of Doctor ofPhilosophy. The examination will be conducted in English. Faculty examiner: Prof. DavidMecerreyes (University of the Basque Country).Berit Olofsson (Stockholm University).Dr.Fredrik Björefors (Strukturkemi).Docent Tom Lindfors (Åbo Akademi University).

AbstractHuang, X. 2017. Conducting Redox Polymers for Electrode Materials. Synthetic Strategiesand Electrochemical Properties. Digital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology 1604. 83 pp. Uppsala: Acta UniversitatisUpsaliensis. ISBN 978-91-513-0168-6.

Organic electrode materials represent an intriguing alternative to their inorganic counterpartsdue to their sustainable and environmental-friendly properties. Their plastic character allows forthe realization of light-weight, versatile and disposable devices for energy storage. Conductingredox polymers (CRPs) are one type of the organic electrode materials involved, which consistof a π-conjugated polymer backbone and covalently attached redox units, the so-called pendant.The polymer backbone can provide conductivity while it is oxidized or reduced (i. e., p- orn-doped) and the concurrent redox chemistry of the pendant provides charge capacity. Thecombination of these two components enables CRPs to provide both high charge capacity andhigh power capability. This dyad polymeric framework provides a solution to the two mainproblems associated with organic electrode materials based on small molecules: the dissolutionof the active material in the electrolyte, and the sluggish charge transport within the material.This thesis introduces a general synthetic strategy to obtain the monomeric CRPs buildingblocks, followed by electrochemical polymerization to afford the active CRPs material. Thechoice of pendant and of polymer backbone depends on the potential match between these twocomponents, i.e. the redox reaction of the pendant and the doping of backbone occurring withinthe same potential region. In the thesis, terephthalate and polythiophene were selected as thependant and polymer backbone respectively, to get access to low potential CRPs. It was foundthat the presence of a non-conjugated linker between polymer backbone and pendant is essentialfor the polymerizability of the monomers as well as for the preservation of individual redoxactivities. The resulting CRPs exhibited fast charge transport within the polymer film and lowactivation barriers for charge propagation. These low potential CRPs were designed as the anodematerials for energy storage applications. The combination of redox active pendant as chargecarrier and a conductive polymer backbone reveals new insights into the requirements of organicmatter based electrical energy storage materials.

Keywords: Organic electrode material, Energy storage, Conducting redox polymer,Polythiophene, Terephthalate, PEDOT

Xiao Huang, Department of Chemistry - BMC, Organic Chemistry, Box 576, UppsalaUniversity, SE-75123 Uppsala, Sweden.

© Xiao Huang 2017

ISSN 1651-6214ISBN 978-91-513-0168-6urn:nbn:se:uu:diva-334562 (http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-334562)

We could choose to live like a candle, flame from the beginning till the end, and bring the warmth

and light to our surroundings.

List of Papers

This thesis is based on the following papers, which are referred to in the text by their Roman numerals.

I Yang, L., Huang, X., Gogoll, A., Strømme, M., Sjödin, M.

(2015) Matching Diethyl Terephthalate with n-doped Conduct-ing Polymers. The Journal of Physical Chemistry C, 119(33):18956–18963.

II Huang, X., Yang, L., Bergquist, J., Strømme, M., Gogoll, A., Sjödin, M. (2015). Synthesis and Redox Properties of Thio-phene Terephthalate Building Blocks for Low-Potential Con-ducting Redox Polymers. The Journal of Physical Chemistry C, 119(49):27247-27254.

III Yang, L., Huang X., Gogoll, A., Strømme, M., Sjödin, M. (2016). Conducting Redox Polymer based Anode Materials for High Power Electrical Energy Storage. Electrochimica Acta, 204:270-275.

IV Yang, L., Huang X., Gogoll, A., Strømme, M., Sjödin, M. (2016). Effect of the Linker in Terephthalate-Functionalized Conducting Redox Polymers. Electrochimica Acta, 222:149-155.

V Huang, X., Yang, L., Emanuelsson, R., Bergquist, J., Strømme, M., Gogoll, A., Sjödin, M. (2016). A Versatile Route to Poly-thiophenes with Functional Pendant Groups Using Alkyne Chemistry. Beilstein Journal of Organic Chemistry, 12:2682-2688.

VI Yang, L., Huang X., Gogoll, A., Strømme, M., Sjödin, M. (2017). Conducting Redox Polymers with non-Activated Charge Transport properties. Physical Chemistry Chemical Physics, 19:25052-25058.

Reprints were made with permission from the respective publishers.

My contribution to the included papers

Paper I: Participated in the planning of the study and discussion in the data analysis. I provided all the synthetic work and contributed to part of the writ-ing process.

Paper II: Participated in the planning of the study and performed all the synthetic work and computations. I wrote the initial manuscript and contrib-uted to the continued writing process. Paper III: Participated in the planning of the study and discussion in the data analysis. I provided all the synthetic work and contributed to part of the writing process. Paper IV: Participated in the planning of the study and discussion in the data analysis. I provided all the synthetic work and contributed to part of the writing process. Paper V: Participated in the design and planning of the study and performed all the synthetic work. I wrote the initial manuscript and contributed to the continued writing process. Paper VI: Participated in the planning of the study and discussion in the data analysis. I provided all the synthetic work and contributed to part of the writing process.

Also published

Sterby, M., Emanuelsson, R., Huang, X., Gogoll, A., Strømme, M., Sjödin, M. (2017). Characterization of PEDOT-Quinone Conducting Redox Poly-mers for Water Based Secondary Batteries. Electrochimica Acta, 235:356-364. Huang, X., Yang, Li., Strømme, M., Sjödin M., Gogoll, A. (2016). C-C bond forming and C-N bond forming on 3,4-ethylenedioxythiophene (EDOT): the new synthetic way of building blocks for functionalized PEDOTs. 6th EuCheMS Chemistry Congress, Seville, Spain. Huang, X., Yang, Li., Strømme, M., Sjödin M., Gogoll, A. (2016). 3-(3,4-ethylenedioxythiophene)prop-1-yne (pyEDOT): A new versatile building block for functionalized PEDOTs. 25th Organikerdagarna, Umeå, Sweden. Huang, X., Yang, Li., Strømme, M., Sjödin M., Gogoll, A. (2015). Synthesis and Redox Properties of Thiophene-Terephthalate Building Blocks for Low Potential Conducting Redox Polymers. 66th Annual Meeting of the International Society of Electrochemistry, Taipei, Taiwan. Yang, Li., Huang, X., Strømme, M., Sjödin M., Gogoll, A. (2014). ”Greener” Energy Storage Material – Thiophene-based Terephthalate Redox Polymer. 65th Annual Meeting of the International Society of Electrochemistry, Lausanne, Switzerland. Huang, X., Yang, Li., Strømme, M., Sjödin M., Gogoll, A. (2014) Novel n-type thiophene-based terephthalate redox polymer for energy storage. 248th American Chemical Society National Meeting, San Francisco, USA. Huang, X., Yang, Li., Strømme, M., Sjödin M., Gogoll, A. (2014) Low potential Thiophene-based Terephthalate Redox Polymer for Energy Storage. Gordon Research Conference: Electronic Processes in Organic Materials, Lucca, Italy.

Contents

1. General introduction ................................................................................. 17

2. Aims of the thesis...................................................................................... 18

3. Electrical Energy Storage ......................................................................... 19 3.1 Organic Energy Storage ..................................................................... 21 3.2 Polymers as active electrode materials ............................................... 23

3.2.1 Organic Conducting Polymers .................................................... 24 3.2.2 Organic Redox Polymers ............................................................ 28 3.2.3 Conducting Redox Polymers ...................................................... 29

4. Monomer Design and Synthetic Strategies for Low Potential Conducting Redox Polymers ........................................................................ 30

4.1 Potential match ................................................................................... 30 4.2 Synthetic Tools ................................................................................... 32

4.2.1 Formation of carbon-heteroatom bonds ...................................... 33 4.2.2 Carbon-carbon bond formation ................................................... 37 4.2.3 Synthetic route design ................................................................. 44

4.3 Effect of the linker and computational tools ...................................... 47 4.3.1 DFT calculation .......................................................................... 49 4.3.2 The redox reaction described in aprotic organic solvent ............ 50 4.3.3 Electrochemical polymerizability ............................................... 55

5. Conducting Redox Polymers based on Polythiophene ............................. 57 5.1 Electrochemistry ................................................................................. 57 5.2 Cycling stability ................................................................................. 59 5.3 Mass effect ......................................................................................... 60 5.4 Kinetics............................................................................................... 62 5.5 Conductance ....................................................................................... 64 5.6 Spectroelectrochemistry ..................................................................... 66

6. Concluding Remarks ................................................................................. 70

7. Svensk sammanfattning ............................................................................ 72

8. Acknowledgements ................................................................................... 75

9. References ................................................................................................. 78

Abbreviations

ATR-FTIR Attenuated Total Reflectance-Fourier Transform Infrared CB Conduction Band CI Configuration Interaction CP Conducting Polymer CRP Conducting Redox Polymer CSM Continuum Solvation Model CV Cyclic Voltammetry DET Diethyl Terephthalate DFT Density Functional Theory EES Electrical Energy Storage EDOT 3,4-Ethylenedioxythiophene EPR Electron Paramagnetic Resonance EQCM Electrochemical Quartz Crystal Microbalance Fc Ferrocene FES Free Energy Surface GC Glass Carbon HOMO Highest Occupied Molecular Orbital HF Hartree-Fock IDA Interdigitated Array ITO Indium Tin Oxide LB Lewis Base LIB Lithium-ion Battery LUMO Lowest Unoccupied Molecular Orbital MeCN Acetonitrile MO Molecular Orbital MP Møller-Plesset perturbation theory ORM Organic Redox Molecule pyEDOT 3-(3,4-Ethylenedioxythiophene)prop-1-yne PEDOT Poly(3,4-ethylenedioxythiophene) PG Pendant Group PMF Potential of Mean Force PPT Polyphenylthiophene PT Polythiophene QCM Quartz Crystal Microbalance RP Redox Polymer SCRF Self-Consistent Reaction Field

SEM Scanning Electron Microscope SMD Solvation Model Density TBAPF6 Tetrabutylammonium Hexafluorophosphate TEAPF6 Tetraethylammonium Hexafluorophosphate TMAPF6 Tetramethylammonium Hexafluorophosphate UV-vis Ultraviolet-visible VB Valence Band XRD X-ray Diffraction

Major symbols

Symbol Name SI unit E Potential V E0 Standard Potential V E0

calc Calculated standard potential V Eg Band gap energy J Eonset Onset doping potential V Eox Oxidation potential V Ep Peak potential V E0

1r The first redox potential V F Faraday constant C mol-1 FWHM Full width at half maximum V G Conductance S = Ω-1 G0 Gibbs free energy J mol-1 Gcon-rovib Conformational vibrational-rotational free energy J mol-1 Gp Polarization free energy J mol-1 I Current A = C s-1 k0 Rate constant s-1 kB Boltzmann constant J K-1 Q Charge C R Molar gas constant J mol-1 K-1 T Temperature K t Time s v Scan Rate V s-1 Δm Mass change per molar charge g mol-1 α Transfer coefficient -

17

1. General introduction

Conducting polymers (CPs) exhibit exclusive electrical, optical and mechan-ical properties, making them attractive for not only energy storage but also solar cells,1 displays,2 biosensors,3 etc. The electronic conductivity together with the possibility to tune ionic transport of CPs makes them appealing for electrical energy storage (EES) applications.4 Apart from highly conductive CPs,5 there is another type of electroactive polymers for EES, classified as redox polymers (RPs).6-7 In this thesis we define RPs consisting of redox active moieties immobilized on a non-conductive polymeric network. These redox centers provide the high capacity in these polymers. By combining the advantage offered by CPs and RPs, the redox pendant groups (PGs) can be designed to attach to a conductive polymer. Such polymers are referred to as conducting redox polymers (CRPs). Conventional organic redox active PGs are for example carbonyls,8-9 radicals,10-11 and organic sulfur compounds.12 The combination of CPs and RPs into CRPs helps to solve the two major problems of organic EES materials: dissolution of the active material and low electronic conductivity.12 The CPs are not conducting at all potentials, so in the design of CRPs it is important for the redox matching of PGs and CPs, i.e. the PGs redox potential is in the range where the CPs conduct (dop-ing potentials). It is also important that the individual redox properties from the two moieties are preserved in the CRPs, i.e. the high conductivity of the CP and the high charge storage capacity and well-defined, stable and facile redox reaction of the pendant.13

The work presented in this thesis involves the synthesis and DFT calcula-tion of thiophene-based CRPs building blocks using diethyl terephthalate (DET) as the capacity carrier in the low potential region. The corresponding CRPs have been investigated regarding their electrochemistry, kinetics, con-duction mechanism, cycling stability, mass effect and spectroelectrochemis-try. A thiophene derivative has been designed for fast synthetic access to poly(3,4-ethylenedioxythiophene) (PEDOT) based CRPs, which then has been employed for the synthesis of PEDOT related energy storage materials.

18

2. Aims of the thesis

The aim of this work is to design synthetic routes for CRPs and to under-stand the fundamental characteristics of these materials. As the first attempt on CRPs for organic anode materials, this work introduces a general guide-line to the CRPs rationalization for EES materials. The work is described in the included papers as follows: Potential match between CPs and PGs under various experimental con-

ditions. Design and synthesis of CP monomers attaching PGs via vari-ous linkers and DFT calculations on their redox properties. (Paper I, II & V)

Investigate the electrochemistry of DET-functionalized thiophene mon-omers to understand the effect of linker on their electrochemical proper-ties and polymerizabilities. (Paper II & IV)

Study the electrochemical and spectroscopic properties of CRPs, to explore their kinetics, conduction mechanism, cycling stability, mass ef-fect and doping-induced optical properties. (Paper III, IV, VI)

19

3. Electrical Energy Storage

Electrical energy supports the modern life of mankind. Its power stretches into nearly every single activity in the society. Thus it is important that this power source is stored after electricity generation and supplied uninterrupt-edly. Electrochemical energy storage devices are a key component in nu-merous areas of technological relevance ranging from transportation to cus-tomer electronics. They can be divided into two main categories: Electro-chemical capacitors (ECs) and batteries. ECs store energy directly as charge between the electrode and electrolyte and they display significantly faster charge-discharge rate than the batteries, which makes them suitable for high power applications.14 However, ECs energy density is much less compared to batteries which store energy in the form of chemical energy by faradaic reactions. In the modern art of rechargeable battery technology commercial-ly available examples include valve-regulated lead acid (Pb-Acid), nickel-cadmium (Ni-Cd), nickel-metal-hydride (Ni-MH) and lithium-ion devices (Figure 1). Lithium-ion batteries (LIBs) stand out from these masterpieces due to their high energy densities (per unit volume or per unit mass). A bat-tery consists of a negative electrode (anode), a positive electrode (cathode) and an electrolyte. A rigid separator is often placed between the two elec-trodes to prevent a short circuit. Figure 2 shows a schematic overview of a Li-ion battery. The two electrode materials decide the specific energy of the batteries. The most extensively used commercial cathode materials are based on metal oxides, e.g. LiCoO2 and LiFePO4. Capacities of these cathodes are, however, limited by the number of Li ions that can be intercalated while retaining structural stabilities. Thus the specific capacities are usually in the range of 120-160 mAh g-1. Limited capacity together with limited electronic conductivity require large-area electrodes to reach sufficient capacity. Com-mercially available anodes are made from carbonaceous materials such as graphite, which is capable of intercalating one Li atom per six carbon atoms (LiC6) with specific capacities of 372 mAh g-1.15

Organic materials have drawn attentions from researchers and they are proposed as a feasible solution to decrease the cost and keep environmental friendliness. These materials can also bring new properties, such as the flex-ibility, improved processability and they provide compatibility with other cycling ions than Li, for instance Na or ammonium based ions. The main building blocks for organic compounds (e.g. carbon, nitrogen, sulfur, oxygen) are all abundant and inexpensive. Modern synthetic organic chemistry can

20

provide vast chemical tunability of these compounds by rational synthetic design, which makes them even more attractive. These organic molecules can be designed to optimize capacity, redox potential, or charge-discharge cycle. Moreover, the amorphous organic materials preclude the problem of structural changes during charge and discharge cycles. Thus it offers the possibility to make green and plastic batteries.16-19

Figure 1. a) Batteries store more energy per unit weight than ECs. Thus batteries are suitable for long-term energy supply, while ECs can provide high power in a short time period. Other systems are included for comparison. b) Volumetric and gravi-metric energy density for the major small-scale rechargeable battery systems.20 Lith-ium based batteries store the most energy.

Figure 2. A schematic representation of the process of a) charging and b) discharg-ing of a Li-ion battery. It stores energy by lithium ion intercalation.

21

3.1 Organic Energy Storage Applying organic redox molecules (ORMs) to batteries is not a new concept and can be traced back to the 1960s.21 These organic materials are normally featured by light weight, easy preparation and various redox chemistries, which make it possible to tailor batteries with high energy density, fast elec-tron transfer and economical waste recycling.6,17 Elementary oxygen and sulfur offer huge theoretical capacities of 3350 and 1672 mAh g-1 respective-ly, and they are fascinating building blocks for functional redox moieties that provide energy storage by conversion reaction. These are divided into three major classes: organic sulfur compounds, free radical compounds and car-bonyl compounds.12 In organic sulfur compounds a disulfide bond is em-ployed to provide charge storage through the S-S bond cleav-age/dimerization redox reaction. However, this reaction experiences poor kinetics with large anodic and cathodic peak separation.22 Moreover, the dissolution of these organosulfur compounds into organic electrolyte limits its use in batteries. As an improvement polymeric organosulfur compounds have been developed with S-S bonds in the polymer backbone or as PG of the backbone.23-24 Free radical compounds are mostly centered on oxygen, providing nitroxyls (NO·) and phenyloxyls (PhO·) as examples of this redox site. The reaction of such radical oxides shows fast charge transport through charge propagation within the material. These molecules are also soluble in common organic electrolyte, so a polymeric frame is normally applied on these radical oxides to solve dissolution problems.25

The carbonyl group is a common organic chemical structural moiety composed of a sp2 carbon and oxygen connected by a double bond. Besides their structural simplicity carbonyl-containing compounds can reach theoret-ical capacities of 120 ~ 600 mAh g-1 depending on the molecular structure.26 The carbonyl moiety undergoes one electron reductive reversible reaction. The reduced anionic carbonyl group can be stabilized by different functional moieties. Depending on the stabilizing mechanism of this anion the state-of-the-art materials can be grouped into three categories shown in Figure 3a.26 The molecules in category I employ vicinal carbonyls, for example 1,2-dione, and the structure can be stabilized by forming enolate upon reduction. Category II are aromatic compounds consisting of the carbonyl group direct-ly connected to an aromatic core which allows the negative charge to dis-perse by delocalization, for example conjugated carboxylate and imide. In the last category III the compounds are quinone derivatives. The reduction product is stabilized by forming an aromatic ring. The different structures, thus, result in their distinct electrochemical properties, especially in the re-dox potential. Figure 3b shows a systematic overview of the electrochemical properties of these carbonyl compounds. Quinone derivatives normally work in relatively high potential range as cathode materials27 and conjugated car-boxylates in lower potential range as anode materials.28-32 Terephthalate is

22

composed of dicarboxylate bridged by a π-conjugated phenyl ring. Its lithi-um salt (di-lithium terephthalate) performs robust electrochemical reactivity below 1V (vs. Li/Li+),28 therefore making it attractive as anode material in LIBs. Beside its electroactive response to lithium ion, terephthalate has also been observed to be active towards sodium ions33-35 and alkylammonium ions as presented in our work.

However, there are disadvantages of these small organic molecules which limits them for battery applications, for example the dissolution problem and low electronic conductivity. The dissolution takes place due to their low molecular weight that favors the solubility in the common organic solvents used in batteries (e.g. acetonitrile, propylene carbonate). Dissolved active materials can shuttle between the electrodes, which then lead to the self-discharge of the battery and affects the cycling stability. This turns to be no problem for flow batteries36-37 but for the traditional batteries with the active materials in solid phase. There are several methods to solve the dissolution problem, including immobilizing the molecules on the surface of an insolu-ble substrate,38 increasing the molecular weight to reduce solubility,39 using charged structures to reduce solubility40 or using solid electrolyte/ionic liq-uid.41-42 Traditionally conductive additives (i.e. carbon black) can increase the conductivity by blending with the electroactive materials.43 However, they are just dead weight, contributing no capacity to the material, and sev-eral research groups reported less than 100% of the theoretical capacity indi-cating hidden regions, i.e. part of the active materials is hidden due to the lack of conductivity.

Figure 3. a) Representative molecular structures for the three categories of electrode active materials based on carbonyl compounds. b) Overview of the cell discharge potential and the theoretical capacities of the three different carbonyl categories.

23

3.2 Polymers as active electrode materials The conventional inorganic Li-ion batteries reach their limits when used in mobile applications such as mobile sensor systems, smart clothes, or smart packaging. The demand for such thin-film applications clearly differs from the conventional batteries. The essential requirements for such applications are for batteries to be flexible, free of toxic and harmful metals, and being produced from abundant and, ideally, renewable resources. Organic materi-als are probably a good option to circumvent the shortcomings of inorganic material (e.g. lithium cobalt oxide, lithium nickel cobalt manganese oxide). As previously mentioned, the solubility in electrolyte represents the chal-lenge in the course of developing organic electrode materials. The incorpora-tion of redox-active units to a polymeric backbone turns out to be a feasible synthetic approach to overcome this issue. Organic polymers are known to be flexible and can be produced and processed on an industrial scale. While most redox active organic polymers are synthesized from fossil carbon sources there have already been several attempts and achievements in using renewable resources.44 In contrast to the inorganic electrodes whose redox reactions are based on complex intercalation mechanisms, organic polymers perform redox reactions which results in high rate performance and long cycling life.7 In contrast to metal based electrodes, usually, the uncharged form of organic polymers is processed, which means the cathode is oxidized initially by applying an external current while the anode is reduced. The electrode charges are then compensated by the counter ions from electrolyte solution, as shown in Figure 4.

Figure 4. A schematic representation of the process of a) charging and b) discharg-ing of an all-organic battery. It stores energy by simple redox reactions of the elec-trode materials.

24

3.2.1 Organic Conducting Polymers The organic battery story starts with the discovery of the electronic conduc-tivity of doped conjugated polymers, polyacetylene, in 1977 by Alan J. Hee-ger, Alan G. MacDiarmid and Hideki Shirakawa.45 An early attempt was then made to use conducting polymers as active material in an organic bat-tery.46-47 After that the main focus shifted to cyclic conjugated polymers such as poly(aniline), poly(pyrrole), poly(thiophene), poly(p-phenylene), and poly(carbazole).6 In the late 1980s, commercial batteries were then launched by Bridgestone/Seiko and VARTA/BASF, based on poly(aniline) and poly(pyrrole). However, they were discontinued after a short period of time mainly due to their inferior charge/discharge cycling behavior.48

CPs store charge upon doping either by oxidation (p-doping) or reduction (n-doping). Charge carriers are delocalized within the extended π-system. The band structure in CPs is analogous to inorganic semiconductors, provid-ing CPs their intrinsic property – conductivity.49 To understand the conduc-tion mechanism it is useful to consider their electronic band structure. The electrons in an atom can be described by the atomic orbitals, which have discrete sets of energies. When atoms are combined into a molecule, the atomic orbitals interact with each other and form the molecular orbitals (MOs) that have different energies compared to atomic orbitals. The combi-nation of more and more atoms into the molecule will eventually cause these discrete energy levels to merge into a continuum, a so-called band. The va-lence band is made up of electron-occupied orbitals and the conduction band is made up of unoccupied orbitals. The gap between the valence band and the conduction band, i.e the range of energies not covered by any available states, are referred to as the band-gap. Polyacetylene is the classic model polymer to illustrate the band structure of organic semiconducting materials (Figure 5). Polyacetylene with equal bond lengths would have no band gap (Figure 5a).50 The absence of a band gap would lead to a metallic conduction in the neutral (undoped) material, but this has not been observed experimen-tally.45 The reason why its neutral form behaves like an insulator is the oc-currence of Peierls distortion that shortens every second bond into a double bond in the chain (Figure 5b).51 This results in a decreased energy of occu-pied π orbitals and simultaneously increased energy of occupied σ orbitals. The overall energy changes from these bonding orbitals destabilize the cor-responding anti-bonding orbitals. It determines this distortion and thus the band gap. Since the width of the gap energy is larger than the thermal energy available at room temperature, polyacetylene in its neutral state is an insula-tor. Upon oxidation or reduction of the chain, new thermally accessible states are formed within the gap, which allows conductivity at room temper-ature (Figure 5c).50

25

Figure 5. Polyacetylene as the model to understand the band structure in a conduct-ing polymer: Conducting band (CB), valence band (VB), band gap, mid-gap states).

CP can be prepared by either chemical polymerization52 or electrochemical polymerization.5 In this thesis the monomeric substrates are polymerized electrochemically due to the convenience for electrochemical studies. The afforded polymeric materials are prepared via anodic (oxidative) polymeri-zation, which includes several steps: radical formation, dimerization, depro-tonation, oligomerization, nucleation, deposition, growth and solid state pro-cess. This method is commonly seen in the polymerization of PTs,53 polypyrroles54 and polyanilines.55 As illustrated by the thiophene electro-chemical polymerization, the initial step involves oxidation of a thiophene monomer which produces a radical cation. The radical can then couple with a second radical cation to form a dicationic dimer, or with another monomer to produce a radical cation dimer. After elimination of two protons a two-unit coupling is complete. Prolongation of these oligomers by dimerization yields a precipitation of CP film on the electrode surface. (Scheme 1)

Scheme 1. Anodic polymerization of thiophene.

Doping processes typically lead to the charging of CPs. The doping can be induced chemically or electrochemically by either oxidation (p-doping) or reduction (n-doping). The chemical doping reagents are either electron ac-ceptors, such as iodine and FeCl3, or electron donors, such as potassium naphthalide. The polymers can also be doped in an electrochemical cell. Similar to the redox reaction in solution, the polymeric film on an electrode can be oxidized (p-doped) or reduced (n-doped) by applying potentials. To

26

maintain the electroneutrality, counter-ions flow in and out during the pro-cess of charging and discharging. Figure 6 illustrates the process of p-doping in PT. The doping and dedoping process causes the local geometrical distor-tion shuttling between a quinoid-like and a benzenoid-like structure. The neutral state of PT has a band gap Eg of about 2 eV56 and only one transition is seen between VB and CB corresponding to the band gap transition. One-electron oxidation (or lightly doped) generates one single electron filled and one unfilled in-gap state. These localized electronic states form the charge carriers in the polymer chain, referred to as polaron with a spin ½.57 The transitions from VB to in-gap states are much lower in energy compared to the Eg, which contribute the optical absorptions related to transitions involv-ing in polaron states.58 When the doping increases, i.e. more polarons are generated within the polymer, there are chances that the polarons are recom-bining to form the spinless bipolaron.57-58 When comparing the creation en-ergy of a bipolaron relative to that of two polarons, the calculation results from polyacetylene, polyparaphenylene and polypyrrole indicate that the distortion energy to form one bipolaron is roughly equal to that to form two polarons. However, the energy gain by pairing to the counter ions is greater in the bipolaron case, which makes it thermodynamically more stable than two polarons in these systems despite the Coulomb repulsion between two same charges.57 Theoretical calculations also suggest three absorption bands for the polaron and two for the bipolaron state.59

Figure 6. Schematic view of the PT p-doping. Band structure in the conjugated polymer upon doping from polaron to bipolaron.

On the other hand n-doping of CPs usually occurs at extreme conditions, i.e. at quite negative potentials and an oxygen- and water-free environment is required. This difficulty results in n-doping being less explored. The most studied CPs with n-type capacity doping are PT and its derivatives.60-63 The-se materials show capacitive doping of both p- and n-type as illustrated in Figure 7. N-doping of PT becomes more reversible when using smaller al-

27

kylammonium cations and it shows different cycling stabilities when varying the size of cations (details in Paper I). Thus the choice of electrolyte and solvent results in various n-doping levels of CPs at different experimental conditions.64-67

Figure 7. Cyclic voltammograms of PT (black curve) in MeCN containing different electrolytes (0.1 M). TB/TE/TMAPF6 are tetrabutyl/ tetraethyl/ tetramethylammoni-um hexafluorophosphate. Red curves show the individual scans at only negative potential region. Adapted from Paper I with permission from the publisher. Copy-right 2015 American Chemical Society.

Even though the conductivity represents one of the major advantages, their inherent connectivity of the redox centers to the current collector yields a drawback as well. Since the charged centers are not separated electronically and instead strongly affecting each other, it gives rise to the redox potentials of conjugated polymers depending on the doping levels and they gradually change during the charging/discharging cycles (Figure 8). The resulting sloppy cell voltage limits the range of applications. Another issue is the lim-ited achievable degree of doping. Ideally, one charge could be stored per one repeat unit in the polymer chain, but conjugated polymers typically give reversible doping levels of at most 0.3 to 0.5, which results in significant decrease in capacity.4-5 A much higher level of doping would destabilize the polymer matrix through undesirable inter-chain interactions or cause a de-struction of the polymer chains.68-69

Figure 8. Representation of the comparison of the cell discharge curves between CPs and RPs (e.g. radical polymers).7

-2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0

-0.04

-0.02

0.00

0.02

0.04

0.06

-2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0

-0.08

-0.06

-0.04

-0.02

0.00

0.02

0.04

0.06

-2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0

-0.10

-0.05

0.00

0.05

0.10

E (V vs. Fc+/Fc0)

PT in TBAPF6

E (V vs. Fc+/Fc0)

PT in TEAPF6

I (m

A)

PT in TMAPF6

E (V vs. Fc+/Fc0)

28

3.2.2 Organic Redox Polymers Compared to CPs, RPs have flatter plateau voltages for the charg-ing/discharging curve and accommodate more charges. RPs are based on a nonconductive backbone which either bears electroactive PGs70-72 or incor-porates the redox active groups into the backbone.39,73-74 These electroactive groups contribute to the distinct redox potential due to the localized redox site and are also dedicated to accept one or more electrons per repeating unit, such as carbonyls, organosulfur, etc. (discussed in section 3.1). Organic radi-cal polymers represent a prominent subclass of RPs for EES due to their superior redox chemistry, in particular the favorable kinetics.75-77 The pen-dant organic radicals attached to the polymeric matrix possess an unpaired electron in the uncharged state. Because the redox process of radicals does not involve any bond-forming and -breaking, it enables fast electron transfer, i.e. high rate performance.78 The stability problem of unpaired electrons (i.e. high reactivity, tendency to dimerize) can be overcome through rational de-sign by introducing a combination of resonance effect and steric hindrance. For example, 2,2,6,6-tetramethylpiperidinyl-N-oxyl (TEMPO) as the classi-cal representative of p-type nitroxyl radical species bears four methyl groups surrounding the active site as hindrance and nitrogen is serving as the reso-nance counterpart during redox reaction.75 Phenoxyls, such as galvinoxyl, have two tert-butyl groups as hindrance and phenyl as resonance stabilizer, serving as n-type radical material.79 More detail on the radical polymer ac-tive materials can be found in the Schubert review.7

Because the charges of RPs reside on the localized redox sites on the non-conjugated polymer chain, its conductivity involves a hopping mechanism.25 The charge transport in such materials is through an electron self-exchange process along the redox gradient formed among these close radical units and also minor structure changes, different from the materials with conjugated π-systems (Figure 9). This charge gradient-driven electron transfer is suffi-ciently fast in thin layer polymer films, but it becomes inadequate when the film gets thicker and carbon conductive additives are thus needed to provide enough electron transfer pathways.7 These carbon additives are non-active electrode components. Therefore the energy density of such actual radical devices is reduced.

Figure 9. Graphical illustration of the structure characteristics of CPs vs. RPs. The charge transfer in CPs involves intra-and inter-chain electron transfers, whereas in RPs it involves electron hopping between the redox sites.

29

3.2.3 Conducting Redox Polymers CRPs combine features from CPs and RPs. A CP backbone is decorated with redox pendants such as carbonyls,80-81 dithiols,12 radicals,10-11 metal com-plexes66,82 and viologens.83-84 Ideally this design would bring CRPs high charge capacity and fast charge transport. This concept is not new. Polypyr-role with several functional groups (quinone derivatives, phenothiazine, an-thracene) attached at the N position was published in the 90s.85-86 Later, in a synthetically similar fashion, polypyrrol was hosting a ferrocene redox cou-ple and this CRP was firstly used as electrode material in LIBs.82 The charge capacity was increased compared to the pristine polymer because of the combination of redox sites with the polypyrrol backbone. Most of the efforts have been devoted to the research on cathode materials resulting from the readily accessible p-doping of most CPs. In the low potential region, anode materials based on n-doping CPs are much less explored and none of the studies83 on such materials realizes synergetic effects between the redox properties of pendant unit and the CP polymeric frame due to potential mis-match between the two active components.13 Thus, potential matching be-tween the pendant redox site and polymer doping becomes the first step to explore CRP based electrode material. Beside the fundamental realizations of CRPs, the synthetic effort to make such polymers is constantly the bottle neck for new materials production. This thesis will introduce the synthetic methodologies to implement the polymer design and explore CRP based anode materials.

30

4. Monomer Design and Synthetic Strategies for Low Potential Conducting Redox Polymers

The rational design of CRPs requires: potential match between the pendant groups and the polymer backbone, a linker in between these two moieties that allows the individual properties of them to be preserved, and a low mass of the repeating units. Potential match means that the redox reaction of the redox pendant group lies in the potential region imposed by the CP backbone, where the CP becomes electronically conducting but at sufficiently low doping level to ensure that the polymeric frame is stable.87 This is the prerequisite to achieve fast charge transfer, stable redox cycling and to avoid the charge ca-pacity losses resulting from the inactive region in the final material. The pre-sent work shows that the linker is another important factor that would give a great impact on the electrochemical properties of the dyad product by affecting the polymerizability of the monomers and modulating the interaction between polymer backbone and redox pendants. In this part I will discuss the potential match between CPs and ORMs, how to achieve the molecular fragments inte-gration via organic synthetic tools and the impact of the linker on the electro-chemical and optical properties of the dyad monomers.

4.1 Potential match Figure 10 illustrates a collection of polythiophene derivatives under both n- and p-doping together with a series of ORMs of redox potentials and theoret-ical capacities that have been reviewed in the recent literature.12,26 The ex-amples of PT, PEDOT and polyphenylthiophene (PPT) show conductance increase in both doping regions. However, this is not significant in the n-doping region compared to p-doping. This is presumably due to the unstable n-doped state resulting from its electron rich feature which is sensitive to oxidative and proton sources. The ORMs shown at the bottom of the figure cover a wide range of potentials, -2.4 ~ 0.4 V vs Fc+/Fc0, and capacities. A clear trend can be seen with the carboxylates fitting into the low potential region matching well with the n-doping of PT derivatives. Quinones and dithiols cover an interzone across both n- and p-doping regions. The redox reactions of radicals occur at higher potential regions covered by the p-doping of PT derivatives. The various matching between ORMs and CPs

31

provides a better insight for designing further CRPs. For instance, quinone derivatives have been combined to PEDOT and produce CRPs based elec-trode materials.19,88 For anode materials, terephthalates have attracted inter-est.28,34 Upon reduction, such terephthalates are believed to be transformed into species with reduced aromaticity as a result of delocalization of the ad-ditional electron into the aromatic ring, resulting in low reduction potentials for this transformation.28 To design a low potential CRP, the carboxylate derivative DET has been chosen as the pendant in conjunction with PT backbone in this thesis (Figure 10).

Figure 10. Potential match of thiophene-based CPs and ORMs in different potential regions. The upper part shows the conductance of CPs at both p- and n-doping re-gions (the blue inset shows the DET redox potential matching PT n-doping); the lower part shows a collection of the theoretical charge capacity and redox potential of ORMs suggested in the literature reviews. Conductance G was evaluated from the in situ conductance measurements with a CV process on an IDA electrode.

32

4.2 Synthetic Tools In all energy storage applications of organic materials it is essential to be able to synthesize the intricate molecular structures that are required for the optimal functionality. Organic chemistry has a long history and countless research efforts have been made to explore new methods to produce different chemical compounds. In this section, I will discuss the main organic synthet-ic strategies that are applied in this thesis work targeting the molecules in Scheme 2, to depict the synthetic routes to CRP monomers.

Scheme 2. Selected molecules from this PhD thesis discussed to illustrate synthetic strategies for CRP monomers. Blue color indicates PG and Red color monomer unit for the polymer backbone.

M1-14 consist of two molecular fragments, which are connected by different linkers. The two fragments serve as polymer building block (thiophene or pyridine) and redox charge carrier, respectively. They may be connected by all-carbon or carbon/heteroatom linkers. Usually, different reactions are re-quired to create carbon-carbon or carbon-heteroatom bonds.

33

4.2.1 Formation of carbon-heteroatom bonds

Nucleophilic substitution Nucleophilic substitution is a type of polar reactions occurring between spe-cies with increased or diminished electron density at certain atoms, so-called nucleophiles or electrophiles, respectively.

A nucleophile is a chemical compound that has a relatively high energy pair of electrons capable of making a new chemical bond. A nucleophilic atom may be neutral or negatively charged. For example, alcohols (ROH), thiols (RSH), sulfides (R2S), amines (R3N), phosphines (R3P), and alkoxides (RO-) have atoms oxygen, sulfur, nitrogen, and phosphorus (III) with elec-tron lone pairs serving as nucleophilic site. In Paper V, we encountered sev-eral situations where the compounds contain an electrophile-halogen (E-X) bond. The electrophilic atom E has an octet, but it is attached to one or more groups of X atoms by a σ bond, constituting leaving groups. It means that when the nucleophile attacks the E atom to form a new bond, the E-X bond would undergo heterolytic cleavage, with the leaving group including the electron pair dissociating to form an independent anionic species. The syn-thesis of viologen M10 from the bipyridine precursor applied the nitrogen with its lone pair as nucleophile which was able to attack the electrophilic carbon in methyl iodide, resulting in dissociation of iodide as anionic leaving group (Scheme 3).

Scheme 3. The nucleophilic substitution to synthesize viologen M10.

The electrophilicity of carbon in methyl iodide stems from the electronega-tivity of the iodine atom. When the nitrogen lone pair is localized in an sp2 orbital on an aromatic ring, such as pyridine or imidazole (N-3), it is highly reactive towards silyl chloride, forming a very reactive silylation reagent for the purpose of protecting alcohols (Scheme 4, step 1).

34

Scheme 4. The synthesis of pyEDOT.

The first step of synthesizing pyEDOT involves the alcohol protection of glycidol by silylation using imidazole as the catalyst, so as to conduct the following nucleophilic reaction on an epoxide group. In step 2, the lithium trimethylsilylacetylene contains a bond between a nonmetal and a metal, constituting a sigma-bond nucleophile. The electrons in this bond are used to form a new bond between the nonmetal part and the electrophile part. It at-tacks on the less hindered carbon of the epoxide and the reaction is acceler-ated by the electrophilic Lewis acid BF3 coordinating with the oxygen atom (Scheme 4, step 2). The boron atom lacks an octet and has a low energy non-bonding orbital, usually a p-orbital. The electrons from the oxygen lone pair are used to form a new bond with boron, giving it its octet. The formal charge of oxygen increases by 1. Then the electrons in the C-O bond become likely to reduce the O+ which makes the adjacent carbon more electrophilic. After addition of the nucleophile the oxygen formal charge decreases back. Dissolving of K2CO3 in methanol provides methoxide,89 which can attack the silicon centre to cause desilylation. The nucleophilic attack by Lewis base (LB) leads to a pentavalent silicon centre which is permitted due to hybridi-zation with the vacant 3d-orbital of silicon. The entropy gain by forming a bond with methoxide is the driving force for a fast cleavage of the substrate. In a similar fashion the desilylation can also be induced by fluoride anion F-. The transetherification in step 4 is catalyzed by acid. The protonation of a methoxyl group makes the thiophene β-carbon amenable to nucleophilic attack and it leaves as a methanol after substitution. Finally the deprotona-tion by a Lewis base (methanol) gives pyEDOT.

35

The carbonyl group is a π bond electrophile. In this type of electrophile, the electrophilic atom E is attached by a π bond to an atom that can accept a pair of electrons. Examples are C=O, C=N or C≡N groups, in which the less electronegative atom C is the electrophilic one. The attack from a nucleo-phile would break the π bond and decrease the formal charge of electron-accepting atom by 1. The nitrogen from amine attacks the electrophilic car-bonyl carbon in acyl chloride to form a hemiaminal intermediate. Chloride, as a leaving group, causes the hemiaminal to reconvert to the carbonyl group, producing HCl as the byproduct (Scheme 5). This reaction forms an amide bond as the linker to afford the M13 dyad.

Scheme 5. The synthesis of M13.

Under acidic conditions, the protonation of the carbonyl oxygen turns car-boxylic acid into a stronger electrophile. After addition of ethanol the tetra-hedral structure quickly collapses back to ester by eliminating one molecule of water. The terephthalic acid can be protected by ethanol under acidic con-dition (Scheme 6). This protection removes the acidic proton from molecule which allows the further functionalization on terephthalate in basic environ-ment.

Scheme 6. The synthesis of DET.

The deprotonation of alcohol or carboxylic acid from base gives alkoxide. The same fashion is also applicable for thiols. This type of ionic nucleophile is quite useful for making a linker between two chemical units, such as ester, ether or thioether. Depending on the basicity of different alkoxides, it would influence the synthetic strategy for certain target molecules. In paper IV the monomers were designed with different linkers bridging terephthalate moie-ty and EDOT backbone. M6 was made from the nucleophilic substitution of EDOT-CH2-Br by phenoxide (Scheme 7a). However, benzyloxide, as anoth-er nucleophile, would not substitute the bromide on the same molecule, in-

36

stead it removed a proton from EDOT-CH2-Br and left EDOT-alkene and benzyl alcohol as products (Scheme 7b). Due to the resonance from the ben-zene ring phenoxide gets less electron-density on the oxygen atom, reducing the basicity on this position but still retaining certain nucleophilicity. The same effect is observed for the carboxylate anion. In benzyloxide a meth-ylene group isolates the oxygen from the aromatic ring, thus retaining the original features of alkoxide, i. e. being both a good nucleophile and a good base. The basicity of alkoxide promoted the β-elimination on EDOT-CH2-Br and produced alkene as the product. Thus the route towards M7 required adjustment to switch the roles of nucleophile and electrophile. EDOT-CH2-alkoxide was used as the nucleophile to substitute benzyl bromide (Scheme 7c). Because there is no available proton at the β-position to allow elimina-tion of HBr or, seen from another perspective, phenyl on the α-position is a very poor leaving group, the β-elimination would not happen to benzyl bromide.

Scheme 7. a) Aryl- and b,c) alk-oxides carry out substitution reactions. As the basici-ty increases, the alkoxide nucleophile could trigger β-elimination reaction for EDOT-CH2Br.

Cycloaddition The azide-alkyne cycloaddition provides another way of making a carbon-heteroatom bond. It is given the name “click” chemistry, where “click” indi-cates its synthetic easiness and the use of readily available starting materials. This makes it attractive to materials chemistry. The reaction between azide and alkyne belongs to the Huisgen 1,3-dipolar cycloaddition. However, it is not a “click” reaction because this thermal cycloaddition requires much ele-vated temperature and it often produces mixtures of two regioisomers when

OO

S

BrO

O O

O

O

O

OS

O

O O

OO

Br

O

O O

O

OOO

S

Br

H

O

O O

O

OH

OO

S

a)

b)

EDOT-CH2-Br

OO

S

OO

O O

O

Br

O

OS

O O

OO

O

Br

EDOT-CH2-Oxide

M6

c)

M7

37

using asymmetric alkynes. A copper-catalyzed variant that follows a differ-ent mechanism can be conducted under aqueous condition and even at room temperature.90 It gives 1,4-disubstituted triazole regioisomers selectively and can thus be considered being a “click” reaction. As an alternative, a rutheni-um-catalyzed option gives the opposite regioselectivity with the formation of 1,5-disubstitution products.91 In Paper V, the Cu-catalyzed azide-alkyne cycloaddition was tested on bridging the EDOT unit with a phthalimide functional unit (Scheme 8). DFT calculation indicates that a σ-bound copper acetylide bearing a π-bound copper coordinates the azide, resulting in de-creased energy of the transition state. It forms an unusual six-membered copper metallacycle. The second Cu acts as both a stabilizing electron donor and as an acceptor ligand, forming bonds with N and C. The reductive elimi-nation of the second Cu induces the ring contraction and gives a triazolyl-copper derivative. Then protonation produces the triazole ring and completes the catalytic cycle.92

Scheme 8. The proposed mechanism of Cu-catalyzed azide-alkyne cycloaddition.

4.2.2 Carbon-carbon bond formation The carbon-heteroatom bond involving N, O, S, P and halogens is polarized due to the different electronegativities of the bound atoms. Thus, these bonds intrinsically have limited tolerance to acidic or basic conditions, therefore promoting hydrolysis. When incorporated into the linkers between redox PG and polymer backbone, they reduce the stability of the compounds, some of them even being electrochemically redox active. The stability of the linker

38

would thus constitute an additional parameter to consider when designing functional polymers. In contrast, the C-C bond is chemically inert and there-fore more suitable for durable linkers.

Nucleophilic substitution The use of metal-carbon sigma-bond nucleophiles can provide access to carbon-carbon bond formation, as being exemplified by organolithium rea-gent RLi in Scheme 4, step 2. Similarly, Grignard reagent RMgBr, and Gil-man reagent R2CuLi are also useful synthons in making C-C bonds. In Paper II, the methylterephthalic acid was formed from the starting material para-dibromotoluene. The activation of the sp2-carbon-bromine bond by magnesi-um metal leads to a Grignard reagent which is reactive towards the electro-philic CO2, producing a terephthalate di-anion after nucleophilic attack (Scheme 9a). The acid form is simply produced by protonation. In the early stage of this PhD work the project was aimed at a pyridine derivative M14. One synthesis was applying Gilman reagent for the nucleophilic substitution selectively with the sp3 C-Br bond (Scheme 9b), producing a methylene bridge between the two pyridine units. The pyridine Gilman reagent was obtained by lithium halogen exchange, followed by adding CuI to the metal complex.

Scheme 9. a) Grignard reagent and b) Gilman reagent are used for molecular func-tionalization via C-C bond formation.

Cross-coupling reaction The C-C bond forming coupling synthesis has an important part in making polymer building blocks described in this thesis. The well-known syntheses include Suzuki-Miyaura coupling,93 Stille coupling,94 Negishi coupling,95 Kumada coupling,96 Sonogashira coupling,97 and Heck coupling (Scheme 10).98 In these reactions, an aryl halide or pseudohalide undergoes a Pd-catalyzed substitution reaction with a “nucleophilic” alkyl-metal compound such as R-B(OR’)2 (Suzuki coupling), R-≡-Cu (Sonogashira coupling), or others. The Heck reaction is different in that an electrophilic alkene reacts directly with the aryl halide to give a more substituted alkene.

39

Scheme 10. The common Pd-catalyzed cross-coupling name reactions applied in C-C bond-forming synthesis.

The catalytic cycle consists of three important steps: Oxidative addition, transmetallation, and reductive elimination (Scheme 11). The oxidative addi-tion on C(sp2)-X bonds (i.e. aryl and vinyl halide) occurs by coordinating the Pd to the π bond to form a metallacyclopropane complex resonance structure. Then an electrocyclic ring opening takes place, followed by eliminating the anionic halide to give a new cationic Pd-vinyl complex. When I‾ associates with the Pd+, the oxidative addition is finished. The transmetallation is driv-en by the metal electronegativity. Fragment X, i. e. halogen or pseudo-halogen, prefers to combine with the highly electropositive metals. If M is more electropositive than Pd, it is thermodynamically favorable for the ex-change of R (i.e. aryl, vinyl or alkyl) and X to occur. The steric ligands co-ordinated on Pd push the two fragments close to allow a thermodynamically more stable σ bond formed in between by a reductive elimination on Pd.

Scheme 11. The mechanism illustration for each step in the cross-coupling catalytic cycle. a) oxidative addition; b) transmetallation; c) reductive elimination.

40

In the Suzuki coupling an aqueous base is needed to activate the boronic acid or ester. This activation of boron atom enhances the polarisation of the or-ganic ligand to facilitate the transmetallation process. In the Sonogashira coupling, however, a base and a Cu(I) salt (e.g. CuI) are combined to facili-tate the transmetallation. The alkyne is more acidic than most hydrocarbons (pKa ≈ 25), but it is not sufficiently acidic to be deprotonated by amines (pKb ≈ 10). To solve this problem CuI is found to be able to co-catalyze the reac-tion. CuI lowers the thermal energy required for the Sonogashira coupling by converting the alkyne (R-≡-H) to a Cu(I) acetylide (R-≡-Cu), which under-goes the transmetallation with Pd(II). The answer to this conversion might be that the alkyne forms a π complex with Cu, and this complex could be deprotonated (similar to E2 elimination) to give the Cu acetylide.

Scheme 12. The reaction mechanism of the Suzuki and the Sonogashira couplings.

In paper II the molecules M1 and M4 were synthesized by the Suzuki cou-pling method. The acidic carboxylic acid needs to be protected prior to the coupling reaction occurring under basic condition. The benzylic position is also capable of performing oxidative addition. The coupling between 2-bromoterephthalate and thiophenyl boronic acid gives the corresponding directly linked monomer M1 and the methylene-linked monomer M4 (Scheme 13a). The mechanism involving benzyl bromide is slightly different from the one involving aryl or vinyl bromide in that the ring opening gives elimination on the methylene carbon. The double-bond is then shifted to the thermodynamically more stable position in the ring, driven by the [1,3] sig-matropic rearrangement (Scheme 13b).

41

Scheme 13. a) The Suzuki coupling reaction to afford the terephthalate-thiophene dyad linked in different fashion. b) The proposed mechanism for the Pd oxidative addition on benzylic halide.

In paper II and IV the Sonogashira coupling was applied to bridge the thio-phene unit and terephthalate pendants with acetylene. The obvious ad-vantage of the Sonogashira coupling is that the organometallic starting mate-rials are avoided. The organometallic structure makes such types of com-pounds unstable to the ambient environment. Normally a pre-preparation and careful preservation are needed. Compared to the organometallic compound, the acetylene-attached organic molecule is much more stable. This feature becomes much more important when it comes to the production of functional materials in larger quantities. It is one of the reasons for designing the com-pounds discussed in Paper V, where the multi-functional synthetic precursor pyEDOT bears an acetylene tail.

In practice, one advantage of cross coupling reactions is the readily avail-able starting materials. This shortens the synthetic route to the targeted mol-ecules. Another advantage is the provided “less-aggressive” environment, being free from acid, harsh base and nucleophiles, especially pronounced in the Sonogashira coupling condition. It helps to preserve the sensitive redox-active functional groups (e.g. carbonyl, esters). Thus the cross coupling reac-tions provide attractive options of forming C-C bonds, which may be used to build up a heteroatom-free linker for the dyad monomers.

Reduction reactions One of the first metal-catalyzed reactions learned in the organic chemistry course is the Pd-catalyzed hydrogenation of alkenes and alkynes. This reac-tion may convert the alkene or alkyne linker formed in the cross coupling product to a saturated hydrocarbon chain. It was applied for the synthesis of monomers M5 and M9 in Paper IV (Scheme 14a). The two valence electrons from the d orbital of Pd can back-bond to the hydrogen σ* orbital, lengthen-

42

ing and weaking the H-H bond until the complex is better described as a dihydride complex H-Pd-H. Then the complex proceeds to the insertion by coordinating the π bond to the metal. In the insertion, a new σ bond is formed at the expense of a π bond. The nature of the reaction requires that the new C-Pd and C-H bond form to the same face of the π bond. Finally, the reductive elimination reduces Pd(II) back to Pd(0) and forms a new C-H σ bond.

Scheme 14. a) The orbital illustration of Pd oxidative addition to H2, and the hydro-genation of an alkyne to produce M5 and M9. b) The proposed Pd-catalyzed hydro-genation mechanism on alkene and alkyne.

Cycloaddition From the names we can see the cycloadditions seem to make the molecules bigger and form rings. The cycloadditions are further classified as [m + n] according to the number of atoms or π-electrons in each component. The six electron [4+2] cycloaddition, namely the Diels-Alder reaction, is one of the useful synthetic methods to combine two different molecular structures. In the Diels-Alder reaction a 1,3-diene reacts with the π bond of a dienophile to give a six membered ring with one π bond. Two new σ bonds are formed at the expense of two π bonds. Because of this the Diels-Alder reaction is nor-mally exothermic. The retro-Diels-Alder reaction can happen when one of the products is N2, CO2, or an aromatic ring. Most Diels-Alder reactions occur with the normal electron demand, which means an electron-rich (nu-cleophilic) diene reacts with an electron-poor (electrophilic) dienophile. Thus, the dienes substituted with electron donating groups such as RO or

43

R2N and the dienophile substituted with electron withdrawing groups such as C=O, CN or NO2, make the reaction become easier. I will pick up one mole-cule M12 in my “failed-functioning library” as an example to demonstrate this type of synthesis (Scheme 15). The transetherification of 3,4-dimethoxylthiophene with 1,4-dibromo-2,3-hydroxylbutane gives the mole-cule EDOT-(CH2Br)2, which can then be transformed to the diene biexo-methylene-EDOT by β-elimination under basic conditions. The biexometh-ylene-EDOT bears two ether groups which behave as electron donating to the diene; benzoquinone has two keto groups serving as the electron-withdrawing group for the double bond dienophile. The electron donation from the ether raises the HOMO energy of the diene and the electron with-drawing from the keto groups lowers the LUMO energy of the dienophile. The smaller the difference in energy between the two MOs, the stronger is their interaction, the lower is the energy of the transition state, and the faster is the reaction. The Diels-Alder reaction is allowed due to the symmetry match in the frontier MOs interaction according to Woodward-Hoffmann rules. During the reaction it conserves one symmetry mirror plane through the two reactants. If for both reactants the HOMO of the diene and the LU-MO of the dienophile are antisymmetric to the mirror plane, then they are matching for further interaction under thermal conditions. The reaction of these two molecules generates two σ bonds to form the two methylene bridges and one π bond connecting the two ether fragments. This cyclohex-enedione can exhibit keto-enol tautomerism under acidic or basic conditions to afford hydroquinone product M12. Why this molecule is not functioning to polymerize even though it bears a thiophene ring, will be discussed in the computational section.

44

Scheme 15. Synthesis of M12 by Diels-Alder reaction.

4.2.3 Synthetic route design Assembling molecular fragments to a certain molecule might be achieved in several ways. But choosing the most efficient and economical way is the essence of synthetic organic chemistry. So when designing a synthetic route there should be always a question in mind: “Is this making sense and is this to be optimal?” Sometimes your best proposal will fail to the cold reality, but this is the beauty of organic chemistry to keep organic chemists trying. I will present a few stories that I encountered within the project development and hopefully can illustrate for the audience.

In the beginning of the organic anode project we were looking for a con-ducting-redox derivative of polypyridine due to its low potential n-dopability.99 The target molecule is a bipyridine dyad with two bromines in 2,5- positions in one of the rings, serving as the polymerization spots. The added pyridine ring to the polymer backbone is expected to provide extra capacity in the polypyridine n-doping region. There are two proposed routes

45

to reach the target (Scheme 16). Route I involves a four step synthesis, and it looks synthetically reasonable. But more steps mean more time expense, less yield and higher risk of failure. Additionally, the practice told us the oxida-tion of picoline to nicotinic acid turned to be difficult, and of very low yield by using the traditional methods. By contrast, route II having only two steps and a nuleophilic substitution on benzylic bromide might be a solution. A free radical bromination and Gilman-type nucleophilic substitution finally produces the target molecule. However, efforts sometimes do not pay off. The pyridine dyad did not polymerize by using the same method100 as pro-ducing polypyridine, due to some unclarified reasons. The project was then shifted to the thiophene-based polymers.

Scheme 16. The retro-synthesis a) and the optimal route b) for the dibromo-pyridine dyad M14.

The synthesis of functionalized EDOT derivatives could be achieved by two routes as shown in Scheme 17. Both have two steps, but the question is which one is more efficient. As we have learnt previously EDOT is synthe-sized by the transetherification of 3,4-dimethoxylthiophene from ethylene diol. Route I shows the way of functionalizing ethylene diol prior to the transetherification. Then this leaves two problems to tackle: the protection of diol is needed if the synthesis of attaching PG allows for side reaction with alcohol, which occurs quite often under acidic or basic conditions; the acidic environment in transetherification gives a quite unstable factor to certain PGs. Moreover, it was found out in Paper V that different ethylene diol de-

46

rivatives gave considerably different product yields, therefore the outcome would be different for each ethylene diol. In route II the synthesis is con-ducted the another way around, thus, the transetherification is prior to the attachment of PG, leading to a EDOT precursor bearing functional group Y. Examples are the hydroxylmethyl derivative EDOT-CH2OH,101 aminomethyl derivative EDOT-CH2NH2

102 and the methylenethiol derivative EDOT-CH2SH,103 as nucleophiles, as well as the halomethyl derivative EDOT-CH2-Cl/Br104-105 and the exomethylene-EDOT as electrophiles.106 This EDOT derivative precursor is then used to attach PGs. Obviously, this route allows the pendant to avoid one more step in the synthesis and to perform faster library synthesis. Thus an EDOT derivative precursor bearing an acetylene tail, pyEDOT, was developed in Paper V, to facilitate the demanding library synthesis of EDOT derivatives. The alkyne group is a versatile synthetic building block which can be functionalized in a number of fashions.107-110 These include, e.g., well-developed cross-coupling reactions, cycloaddition reactions, radical reactions and reductive addition reactions. So pyEDOT is believed to be a useful precursor for the efficient synthesis of EDOT deriva-tives (Scheme 18).

Scheme 17. Retro-synthesis of EDOT-based dyad.

Scheme 18. Previous and present EDOT functionalization route.

47

4.3 Effect of the linker and computational tools By using the synthetic tool from previous section a series of thiophene-DET dyad monomers M1-9 have been synthesized with variant linkers, as shown in Scheme 2. M1-3 have linkers that bridge the dyad via π-conjugation. The others have no -π-conjugation between the thiophene ring and DET.

The electrochemistry of these monomers was investigated by cyclic volt-ammetry and the resulting voltammograms are shown in Figure 11. They indicate three notable redox reactions: two reductions that are DET-centered (peaks below -2 V vs. Fc+/Fc0) and one thiophene-centered oxidation (peaks above 1V vs. Fc+/Fc0). For all compounds except M7, the first reduction is a chemically fully reversible reaction and then followed by a second irreversi-ble reaction. Comparing all the first redox potentials (E0

1r), they are similar to that of the DET molecule except for M2, M3 and M8 which show a high-er E0

1r (Table 1). This is probably due to the extended conjugation from un-saturated linkers. Oxidation of the thiophene ring occurs in a much more positive potential and the oxidation potential (Eox) varies slightly with differ-ent linkers. In M6-9, the ethylenedioxyl group in EDOT plays a role as elec-tron donating group to the thiophene ring, so their oxidations take place at slightly lower potential.

Understanding the intrinsic behavior of molecular electronics by simple electrochemical tools is not easy. To illustrate the chemistry behind people normally use molecular modeling to assist the in-depth study. Later we will use computational chemistry as a tool to try to understand how the redox potential can be predicted and the effect of linkers on the polymerizability and electronic property of the monomers.

48

Figure 11. Cyclic voltammograms of the DET molecule and thiophene-DET dyad monomers in MeCN containing 0.1 M TEAPF6, showing two subsequent reductions of the DET pendant and one oxidation of the thiophene ring. Blue curves show the scanning of only the first reduction. Adapted from Paper IV with permission from the publisher. Copyright 2016 Elsevier.

Table 1. Experimental redox potential (E01r) and calculated potential (M1-5) (Calc.

E01r) of the first reduction from DET pendant and oxidation potential (Eox) from

thiophene ring in M1-8.

Monomer M1 M2 M3 M4 M5 M6 M7 M8 M9

E01r

(V vs. Fc+/Fc0) -2.15 -2.11 -2.00 -2.18 -2.20 -2.24 -2.18 -2.05 -2.23

Calc. E01r

(V vs. Fc+/Fc0) -2.19 -2.10 -2.03 -2.22 -2.23 × × × ×

Eox (V vs. Fc+/Fc0) 1.31 0.93 1.24 1.16 1.14 0.81 0.80 0.81 0.86

-3 -2 -1 0 1 2 -3 -2 -1 0 1 2 -3 -2 -1 0 1 2

M1 M2 M3

M4 M5 M6

M7 M8

E (V vs. Fc+/Fc0)

M9

DET

49

4.3.1 DFT calculation Computational science is a rapidly growing area that uses advanced compu-ting capabilities to understand and solve a complex system. Especially in the recent decade the hardware and software are developing in a tremendous speed. Chemistry is a complex system in the microcosm of molecules. Com-putational chemistry seeks to explain this microcosm using technological simulations. The methods used can be based purely on quantum mechanics and physical constants, called ab initio methods, in which molecular struc-tures are calculated using nothing but Schrödinger equations. Other methods are called empirical or semi-empirical that use experimental data to provide the input to the mathematical model. Solving the full Schrödinger equation is a calculation costly project, especially for the larger atoms or molecules with many electrons. Thus the Born-Oppenheimer approximation is always ap-plied to simplify the equation. Some important types of modern calculations are Hartree-Fock (HF), Møller-Plesset perturbation theory (MP) and config-uration interaction (CI), and density functional theory (DFT) calculations.111 When solving the Schrödinger equation, the main practical difficulty is the proper treatment of electron-electron interactions in the atoms or molecules that contain more than two electrons. HF, MP and CI rely on the computa-tion of wave function, whereas DFT requires computation of the total elec-tron density and technically no requirement of a wave function.112 The total electron density and the normalized wave function can be interrelated be-cause the total electron density is the absolute value of the square of the normalized wave function divided by the number of electrons. HF is not appropriate for describing electronic energies, so MP and CI built upon HF provide the accurate values by using very large basis sets. However, they are limited to small atoms or molecules due to the expensive computational load. Modern implementation of DFT can provide much higher accuracy than HF at a much lower computational cost. Thus the low cost has driven a steady increase in the use of DFT for the study of electronic structure of larger molecules.

The calculations in this thesis are based on DFT unless otherwise noted. It is a useful tool to predict molecular structure, free energy at finite tem-peratures, and time-dependent phenomena and excited states, etc.

50

4.3.2 The redox reaction described in aprotic organic solvent To aid in the compound design, accurate prediction of the terephthalate re-dox potential was desired (Paper II). A thermodynamic cycle (Born-Haber cycle) was used in this study where the standard Gibbs free energy of redox half reaction, ∆G0

solv (in solution), consists of the free energy change in the gas phase ∆G0

gas and the solvation free energy of the reactant (A) and re-duced (An-) species. These values are then used to calculate the overall reac-tion of the standard Gibbs free energy, ∆G0

solv (eq. 1, Scheme 19). The free energy and solvation energy of electron will be cancelled when referenced against e.g. ferrocene later and is therefore neglected when computing the cell potential.

∆ = ∆ + ∆ ( ) − ∆ ( ) (1)

Scheme 19. Born-Haber cycle describing the thermodynamic process between gas and solution phases, and the equation obtaining the standard Gibbs free energy in solution.

From ∆G0solv the half-cell potential can be computed through eq 2. ∆ = − (2)

The cell potential E0calc was then referenced to the experimentally known

reduction potential of Fc+/Fc0 in acetonitrile, which is 4.980 V,113 to afford the reduction potential Ecalc by ferrocene as the reference.

Solvation Model A continuum solvation model (CSM) called SMD114 was applied in the solv-ation energy calculation. It was based on the quantum mechanical charge density of a solute molecule interacting with a continuum description of the solvent. We let Rg and Rl denote the equilibrium geometries of the solute in the gas phase and in the liquid phase. The standard-state free energies in these two phases are represented by equations (3) and (4) in the quantum calculation package.

51

( ⁄ ) = ⁄ + ( ⁄ ) + ( ⁄ ) (3) ( ⁄ ) = ( ) + ( ⁄ ) + ( ⁄ ) (4)

The G0trans is the translational free energy, Gcon-rovib is the conformational-

vibrational-rotational free energy, and the G0lib is the liberational115 free en-

ergy. The W(R) represents the quantity of the potential of mean force (PMF) or free energy surface (FES) in the condensed phase. PMF is the solution-phase analogue of a gas-phase potential energy surface.116 The expression of W(Rl) is given by equation (5): ( ) = ( ⁄ ) + ( ) + ( ) (5) The E(Rl/liq) is the liquid-phase expectation value of the gas-phase Hamilto-nian. GP is the polarization free energy of the solute–solvent system when the solute is inserted. GCDS is called the non-bulk-electrostatic effect. The CDS subscript is nominally associated with the free energy change related to the solvent cavitation (C), changes in dispersion energy (D), and possible changes in local solvent structure (S).

If we choose a standard state (indicated by an asterisk) where the concen-tration of the solute is fixed in the gas phase (p0 = 1 bar) to be the same as in solution (V0 = 1 L mol-1), the translational energy G*

trans(S/gas) in the gas phase equals to the liquid-phase liberational energy G*

lib(S/liq).115 Then the fixed-concentration free energy of solvation of solute ΔG*

s is defined by equation (6). ∆ ∗( ) = ∗( ⁄ ) − ∗( ⁄ ) (6)

Combining equations (3), (4), (5) and (6) yields: ∆ ∗( ) = ( ⁄ ) − ⁄ + ( ) + ( ) + ∆ ( ) (7) ( ⁄ ) − ⁄ + ( ) = ∆ ( )

The ∆GEP is literally named bulk electrostatic component, which is usually obtained from a self-consistent reaction field (SCRF) quantum-mechanical calculation at a fixed geometry, in another way saying, choosing a constant geometry for both gas-phase and liquid-phase electronic energies.114 Then this bulk electrostatic component is refined to equation (8), where the solute geometry optimized in solution is chosen for the solvation energy calculation with regard to a recent refinement article by Truhlar:117 ∆ ( ) = ( ⁄ ) − ( ⁄ ) + ( ) (8)

∆Gcon-rovib refers to the change in solvation free energy due to the shift of nuclear coordinates when transferring from the gas phase to the liquid con-densed phase, also named nuclear relaxation. In many cases, the ∆Gcon-rovib is

52

comparatively small to the mean error intrinsic to the continuum solvation model. This justifies neglecting it during parametrization in the implicit solvation model. It will be considered only if it becomes significant for some arbitrary applications.118 The fixed-concentration solvation free energy is then refined to: ∆ ∗( ) = ( ⁄ ) − ( ⁄ ) + ( ) + ( ) (9)

The standard-state solvation free energy ∆G0s (eq. 10) is usually expressed

through the fixed-concentration solvation free energy plus an energy correc-tion based on 1 mol of an idea gas upon compression from 1 bar pressure (p0 = 105 Pa) to 1M concentration (V0 = 10-3 m3 mol-1), which equals ~ 8 kJ mol-1 at T = 298 K: ∆ ( ) = ∆ ∗( ) + ( ) (10)

It is worth to note that calculation of the solute free energy in solution direct-ly from equation (4) at high level of theory would involve intrinsic bias due to the parametrization nature of the CSMs. It should be recognized that CSMs have been parametrized at relatively low levels of theory (e.g. HF and DFT with small basis sets) to reproduce experimental solvation energies, not total free energies in solution. The SCF error is also expected to be canceled from term E(Rl/liq) – E(Rl/gas) in equation (9), but this cancellation is not possible in equation (4). The resulting solution phase free energies, calculat-ed at such low level of theory, may not be accurate enough for quantitative studies.119 Thus, the preferred approach for solution phase free energy would be to use CSMs to calculate the quantity for which they are parametrized (the solvation energy), and be then combined with the gas phase free ener-gies that calculated at much higher levels of theory by equation (11): ( ⁄ ) = ( ⁄ ) + ∆ ( ) (11)

53

In Practice The calculations was performed by the Gaussian 09 software120 and exempli-fied by M1-5 (Table 1). For the first reduction calculation (n = 1, in Scheme 19), all solutes (A and A-) geometries were initially optimized in the gas phase (Rg) and the gas-phase free energies G0(S/gas) were obtained at the theory level of DFT (B3LYP, 6311++Gdp). Then the solution phase opti-mized geometries (Rl) were obtained by loading the integral equation formal-ism model (IEFPCM) at the same theory level. The fixed-concentration solvation energy ∆G*

s was calculated at the theory level of DFT (B3LYP, 631Gd). The E(Rl/liq), Gp and GCDS terms in equation (9) were obtained as a sum from the SCRF calculation by using the universal solvation model SMD at the fixed geometries Rl. The gas phase term E(Rl/gas) was obtained from SCRF calculation in vacuum. For the second reduction calculation (n = 2, in Scheme 19), all the procedures were identical to the first reduction calcula-tion, but the solutes (A and A2-) were optimized and free energy obtained at the theory level of DFT (M062X, 6311++Gdp), since the highly electron rich species A2- could be more correctly described by involving electron ex-change-correlation terms such as M062X functional.121 The solvation energy was calculated by SMD at the theory level of DFT (M062X, 631+Gdp).

The results showed quite great agreement with the experimental value, yielding a predicted mean absolute error (MAE) of 30 mV for the first reduc-tion and of 40 mV for the second reduction (Table 1 and Paper II). A com-parison of the experimentally determined redox potentials with those pre-dicted from computations was shown in Figure 12.

Figure 12. Comparison of calculated and experimental redox potentials in nonaqueous solution (MeCN). The diagonal black line is a guide for the eye representing equivalence (calculated = experimental). Blue squares (■) represent the first electron reduction data, and red circles (●) represent the second electron reduction data. Adapted with permission from Paper II. Copyright (2015) American Chemical Society.

54

The first reduction potentials followed the general trend M5 = M4 < M1 < M2 < M3. This correlated well with the degree of thiophene based orbitals in the composition of the total LUMO, which was obtained by the orbital composition analysis module of Multiwfn122 by Becke method (Figure 13a).123 In the orbital composition analysis, the MOs were decomposed to basis function, shell and atom compositions. Fragments could be defined by the chosen basis functions, shells, atoms or even mixtures of them. In our case the thiophene unit was defined as a fragment and the thiophene related orbital contribution to the whole MO was defined. As the LUMO is general-ly a terephthalate-centered orbital (Figure 13b) in all monomers, the inclusi-on of thiophene assigned functions in the LUMO composition is a measure of the extent of conjugation between the two ring systems and hence that the observed increase in reduction potential is related to the extent of conjugati-on between the DET and thiophene unit. As expected the second reduction potentials also followed the same trend.

Figure 13. a) Correlation between the orbital contribution from the thiophene unit to the total LUMO (x, %) in the neutral state and the reduction potential (y, V) for the first reduction process in M1-5. Open and filled circles correspond to experimentally and computationally derived reduction potentials respectively. b) Representation of LUMO orbitals of M1-5 at DFT level of B3LYP, 6-311++Gdp. Adapted with per-mission from Paper II. Copyright (2015) American Chemical Society.

From both experimental and computational results we can see that the redox reaction of pendant is effected by the property of linker. The π conjugated linker could disperse the electron density among the pendant and backbone, rendering the redox chemistry of both units into a hybrid. In the next section we will see how the linker would effect the polymerizability of the mono-mers.

55

4.3.3 Electrochemical polymerizability The success of electrochemical polymerization relies on a sufficiently low oxidation potential of the monomers, a sufficient radical density on the polymerization site at the oxidation state and inconspicuous steric hindrance of the radical site. The Eox of all monomers M1-8 can be reached before the electrolyte starts decomposing, i.e. within the electrolyte stability window. (e.g. a MeCN solution with electrolyte TEAPF6). However, not all mono-mers can be polymerized electrochemically, i.e. M1-3 were not able to form polymer films on the electrode surface. The radical spin density of M1-5 was simulated by DFT calculation in Figure 14. The density was delocalized through the π conjugation in M1-3, which caused the reduced radical activity on the thiophene ring. By contrast, the radical density was much localized in M4 and M5 due to the non-conjugated linker. This indicates that sufficient radical density on the dimerization site (α position) would facilitate the polymerization (Scheme 1).

Figure 14. Cationic radical spin density surface of M1-5 at the DFT level: B3LYP/6-311Gd, IEFPCM (MeCN).

Figure 15. Cationic radical spin density surface of HQMeVDOT (M12) and HQMeEDOT at the DFT level: B3LYP/6-311Gd, IEFPCM (MeCN).

56

As discussed the synthesis of HQMeVDOT (M12) in section 4.2.2, this mol-ecule failed to polymerize electrochemically. The reason was then uncovered by the computational modeling of its cationic radical species. In Figure 15, the results were showing that HQMeVDOT cationic radical had reduced spin density on the thiophene ring spreading out to the middle unsaturated linker. A similar calculation result was obtained in a previous article on molecule 3,4-vinylenedioxythiophene (VDOT).124 On the contrary, the density could be more localized in the ring by saturating the linker as illustrated by HQMeEDOT. Then, the synthesis of molecule HQMeEDOT was attempted experimentally by hydrogen reduction on HQMeVDOT, but no success has been achieved so far.

57

5. Conducting Redox Polymers based on Polythiophene

All the CRPs in this thesis were prepared by a CV experiment (unless other-wise noticed) under the potential where the oxidation of thiophene occurred. The monomers M4-9 were polymerized smoothly as their pristine derivative, suggesting the insignificant influence from the DET pendant. In this chapter I will discuss the electrochemical properties of these polymers, followed by a discourse on the mass effect, kinetics, conductance and spectroelectro-chemistry of the polymer P5 as a model compound for CRP.

5.1 Electrochemistry As demonstrated in Figure 11 the pendant DET has two subsequent reduc-tions, one reversible reaction followed by a partially reversible reaction, in the potential range from -2 to -3 V vs. Fc+/Fc0. Thus we chose to cycle the CRPs cutting from the first reversible reaction to the polymer p-doping re-gion as shown in Figure 16.

The electrochemical properties of both pendant and polymer backbone are retained in all the CRPs. A pair of sharp redox peaks was observed at low potential from -2.5 to -2 V vs. Fc+/Fc0, as comparable to that of the monomers (i.e. E0

p1r ≈ E01r, Table 2). The n-doping of the polymer backbone

was taking place at the same potential region, but it was hardly observable because of the overlap from the pendant redox reaction. This potential over-lapping projected that the redox reaction from pendant occurred while the polymer was conducting in the n-doped state. The p-doping of the polymers started at more positive potentials (E0

p,onset in Table 2). Compared to the po-tential Eox of respective monomer, the lowered oxidation potential Ep,onset of the polymer was due to the increased conjugation length from coupled thio-phene rings. In the p-doping region, some foreign oxidation or reduction peaks were observed and distort the p-doping behavior, such as the oxidation peaks at around -0.74 V vs. Fc+/Fc0 in both P7 and P8. The extent of distor-tion varied between different materials. These foreign current responses re-sulting from n-doping processes are called charge trapping peaks in the con-ducting polymer literature.66,125

58

Figure 16. Cyclic voltammograms of P4-P9. All polymers were deposited on a glassy carbon working electrode and cycled in a freshly prepared MeCN solution containing 0.1 M TEAPF6 as electrolyte. Adapted from Paper IV with permission from the publisher. Copyright 2016 Elsevier.

Table 2. Pendant redox potentials (E0p1r) and p-doping onset potentials (E0

p,onset) in the prepared CRPs.

CRP P4 P5 P6 P7 P8 P9

E0p1r

(V vs. Fc+/Fc0) -2.19 -2.19 -2.21 -2.18 -2.08 -2.23

E0p,onset

(V vs. Fc+/Fc0) 0.05 0.12 -0.72 -0.91 -0.91 -0.83

-4

-3

-2

-1

0

1

2

-2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0

-3

-2

-1

0

1

2

-4

-3

-2

-1

0

1

2

-2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5

-6

-4

-2

0

2

4

6

-10

-5

0

5

10

-2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5

-4

-2

0

2

4

P5

P4

1 scan 2 scan 3 scan 4 scan 5 scan

P6

1 scan 2 scan 3 scan 4 scan 5 scan

E vs. (Fc+/Fc0)/V

1 scan 2 scan 3 scan 4 scan 5 scan

E vs. (Fc+/Fc0)/V

P8

E vs. (Fc+/Fc0)/V

1 scan 2 scan 3 scan 4 scan 5 scan

P7

1 scan 2 scan 3 scan 4 scan 5 scan

P9

1 scan 2 scan 3 scan 4 scan 5 scan

59

5.2 Cycling stability The doping process of polymers is accompanied by uptake and expulsion of the counter-ions and solvent molecules. The cycling in the p-doping region retains great stability for these CRPs, but the polymers begin to show variant degradations as indicated by the current decay when cycling to the n-doping potential region (Figure 16). To represent the decay of the materials, the retained charge was evaluated from the redox process of the pendant and the CP backbone by integrating the reduction peak of the pendant and the p-dedoping peak of the polymer backbone (Figure 17). The results showed that the charge decays from P5 were on similar magnitudes on both pendant and backbone. The cycling of the polymer in the n-doping region turned the solu-tion into a red color and precipitated some particles onto the bottom after several cycles. The characterization of these particles indicated similar elec-trochemical and optical properties as compared to the intact polymers (Paper IV). The cycling of pristine PT showed similar charge decay under the same experimental condition (Paper I). Thus, we concluded that the charge loss of P5 was due to polymer dissolution when cycling in the n-doping region. Moreover, in situ IR spectroscopy showed an increased amount of solvent molecules entering the electrode surface upon reduction of P5, again sug-gesting the detachment of polymer film from the electrode surface (Paper VI). P6 and P7 showed stable electrochemical response from the PEDOT backbone but degradation from the side chains during cycling, while the pristine PEDOT is known to be stable by cycling under the same experi-mental conditions. Therefore, we suggested that the degradation of P6 and P7 was probably due to the break of ether linker under such reductive poten-tial. Especially the instability was pronounced in P6, suggesting the phenol was not favorable serving as part of the linker. P8 gave the best charge reten-tion, while P9 showed significant decay just by switching the linker from an acetylene to an alkyl chain. The reason behind this observation is still not clear. P5 and P7 became more stable after several cycles. Hence, the poly-mer film was stabilized by pre-cycling prior to the studies of P5 in the fol-lowing section.

60

Figure 17. The charge retention of pendant group (PG) (black color) and polymer backbone (PB) (red color) of P5, P6, P7, P8, P9 at the 1st, 2nd and 5th scans during the electrochemical cycling. The data were extracted from the redox process of the pendant and the polymer backbone by integrating the reduction peak of the pendant and the p-dedoping peak of the polymer backbone.

5.3 Mass effect The uptake and removal of counter-ions and solvent can cause polymer mass changes. This can be followed by the EQCM measurement. Figure 18 showed the mass changes while cycling the polymer P5 in different potential intervals. When cycled in the p-doping region, the polymer increased mass during oxidation of a weight change per mole charge (∆M) of 259.8 g mol-1. It is interpreted as the uptake of one unit of anion (PF6

-) and three units of solvent molecules (MeCN). On the reverse reduction scan, the mass was decreased and changed back to the origin, resulting in the electrochemical reversible cycle of the material. When the polymer was scanned to the n-doping region, an initial mass increase was due to the uptake of cation (TEA+), and then it was followed by a huge mass drop and partial regain from the reverse scan. In the subsequent scans the mass decreased further and led to a heavily distorted curve. By contrast the current response did not show such a decay in the CV of P5 beyond -2.2 V vs. Fc+/Fc0. This result implied that the polymer film was detached from the surface. Further support forthis detachment explanation was also provided by in situ IR measure-ments (Paper VI). The partial recovery on the subsequent scans indicated

1 2 3 4 5

20

40

60

80

100

120

1 2 3 4 5 1 2 3 4 5

5PG 5PB

Cha

rge

Ret

entio

n %

Scan Number

6PG 6PB 7PG 7PB

8PG 8PB 9PG 9PB

61

that the detachment was to some extent reversible and that the polymer film was able to re-attach onto the surface.

Figure 18. a) The changes of mass and the corresponding CVs of P5 in the p-doping region. b) The changes of mass and the corresponding CVs of P5 in the region cov-ering p- and subsequent n-doping. The arrows indicate the initial scan direction. P5 was electrochemically deposited on a QCM electrode and cycled in a fresh MeCN solution containing 0.1 M TEAPF6. Adapted from Paper VI with permission from the publisher. Copyright 2017 PCCP Owner Societies.

62

5.4 Kinetics We investigated the kinetics of this type of CRP by using P5 as a model compound. Its thickness dependence and scan rate dependence were studied.

The polymer film of P5 was electrochemically deposited on a GC elec-trode with different thickness. The samples were then cycled at different scan rates within the potential region of the pendant redox reaction. The lg(Ip) (Ip is the absolute reduction peak current) was plotted as a function of lg(v) in Figure 19a, with different thickness of 2, 4.5 and 26 μm. The thinnest film gave a linear relation while increasing the scan rate, with slope of ~ 1. As the thickness increased, the slope decreased to ~ 0.8 while still keeping the line-arity. This surface-confined behavior of the polymeric film showed that there was no limitation of counter-ion movement or electron diffusion within the active material on the applied time scale, even for a 10 μm film. The fast ion diffusion was suggested due to the porous structure of the material, which was observed by the scanning electron microscope (SEM) analysis in Figure 19b. The accompanied rapid electron transport was likely facilitated by the conducting polymeric frame.

Figure 19. a) The logarithm of reduction peak current of P5 with different layer thickness as a function of the logarithm of scan rate. b) SEM images of P5 and PT. The polymer films were prepared electrochemically on an ITO glass under the same experimental conditions. Adapted from Paper III with permission from the publisher. Copyright 2016 Elsevier.

P5 with the thickness of 4.5 μm was chosen to study the rate constant (kox0

and kre0 are for the oxidation and reduction respectively) and transfer coeffi-

cient (α) for the redox reaction occurring at low potential by CV at different scan rates. They were estimated by eq. 12 and 13, where R is the molar gas constant, T is the temperature (295 K), F is the Faraday constant. From the

63

voltammogram were extracted the peak potential (Ep,a and Ep,c), peak separa-tion (ΔEp) and the full width at half maximum (FWHM). FWHMa and FWHMc were for the anodic and cathodic peaks respectively. The obtained parameters are listed in Table 3. : , = , + (12)

: , = , + (1 − ) ln (1 − ) (13)

Symmetric anodic and cathodic peaks were observed at scan rates lower than 320 mV s-1, with the smallest ΔEp of 1 mV for the lowest scan rate at 10 mV s-1. The FWHM was about 100 mV for both anodic and cathodic peaks, which was close to the value of 90 mV representing an ideal surface-confined reversible Nernstian redox reaction with non-interacting redox groups. These non-broadened peaks suggested that the redox active DET groups did not affect each other so that the distribution of the reaction poten-tials was quite narrow. Additionally, in a surface-confined reaction the num-ber of converted species was following the Boltzmann distribution manifest-ed in Nernst equation during the redox conversion. Thus the whole peak area was subjected to a uniform potential given by the potential applied to the current collector.

Table 3. Kinetic parameters of P5.

polymer ΔEp(mV) FWHMc(mV) FWHMa(mV) kre0 (s-1) kox

0 (s-1) P5 1 109 104 8.28 8.46

The redox peaks got broadened as the scan rate was increased, giving lower Ip and larger ΔEp. Above 1.5 V s-1 the peaks were separated even further and the ΔEp increased linearly with lg(v). The linear fit by Laviron treatment126 for the diffusionless system yielded α = 0.25, kre

0 = 8.28 s-1 for the reduction and 1 – α = 0.24, kox

0 = 8.64 s-1 for the oxidation at E0 (-2.2 vs. Fc+/Fc0). Such a high rate constant of both oxidation and reduction of P5 implied a fast electron transfer within the polymer film (details in Paper III).

64

Figure 20. a) CVs of P5 at different scan rates. Note: the current is normalized. b) the plots of peak potential against lg(v), dashed line shows the redox potential of P5, -2.2 V vs. Fc+/Fc0; solid line is the linear fit of the reduction or oxidation peak po-tentials at varying scan rates. Adapted from Paper III with permission from the pub-lisher. Copyright 2016 Elsevier.

5.5 Conductance The research on the electronic conductivity of CPs covers the efforts on studying the effects from doping levels,127-128 chain length and ordering,129-130 chemical structure,63 dopants and solvents,131-132 etc. Especially, the tempera-ture dependent conductivity measurements provide insight to the activation barriers and charge transport.133-135

We continued investigating P5 to disclose the conduction mechanism of redox pendant-functionalized CPs. The polymer film was prepared instead on an Inter Digitated Array (IDA) electrode. The current and conductance response were recorded in Figure 21. The voltammograms were obtained similarly as they were on GC electrodes (Paper III). The corresponding con-ductance G increased with the polymer film growth and a constant value was reached after several polymerization cycles. It indicated that the volume between two working electrodes was filled with sufficient thick polymer film, which made the outer layer inactive with regard to the charge transport. This ensured therefore a constant geometry for the conductance set-up and made the results comparable between different samples.

The conductance of P5 in the region covering the p-doping processes was quite stable but only one scan in the n-doping region caused dramatic decay of the conductance and the CVs were distorted in the n-doping region com-pared with the polymer on a disc electrode (Paper VI). The instability of the conductance response, which was found to be due to polymer detachment (Paper III and VI), prevented further investigation of the conductivity in the n-doping region. However, according to Coropceanu et al. the electron and hole mobilities are often comparable and governed by the same factors.136

-2.8 -2.6 -2.4 -2.2 -2.0 -1.8 -1.6

-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

-2.0 -1.6 -1.2 -0.8 -0.4 0.0 0.4 0.8 1.2-2.6

-2.5

-2.4

-2.3

-2.2

-2.1

-2.0

-1.9

-1.8

60mV s-1

100mV s-1

150mV s-1

320mV s-1

1.5V s-1

3.2V s-1

6V s-1

10V s-1

a) b)

E (V vs. Fc+/Fc0)

Nor

mal

ized

I (m

A s

V-1)

E p (V

vs.

Fc+ /F

c0 )

lg (ν /(V s-1))

reductionoxidation

65

The results presented herein on the conductance behavior in the p-doping region are therefore likely to hold also for the n-type conductivity.

Figure 21. (a) CVs and (b) the corresponding conductance during the polymeriza-tion of P5 on an IDA electrode by using an Ebias of 1 mV. The supporting electrolyte was a MeCN solution containing 0.1 M TEAPF6. Arrows indicate the direction of current/conductance increased during polymerization cycling. Adapted from Paper VI with permission from the publisher. Copyright 2017 PCCP Owner Societies.

The conductance was measured in situ during redox cycling of P5 at various temperatures (Figure 22a). The CVs response kept almost steady while changing the temperature, which suggested that the number of injected charge carriers was also constant with the temperature. The decrease of con-ductance with increasing temperature was therefore due to the decreased mobility of these charge carriers in the material. To visualize the effect of temperature on conductance, the G was plotted as a function of temperature at various chosen potentials (Figure 22b). At potentials below 0.7 V vs. Fc+/Fc0, G increased slightly with increasing temperature at the lowest tem-peratures under study and thereafter decreased when the temperature was further elevated, giving a maximum G at about 290 K. At potentials above 0.7 V vs. Fc+/Fc0 the G decreased steadily with increasing temperature. Ac-cording to the nonperturbative microscopic models,136 at high temperatures, when the reorganization energy for electron transfer is sufficiently less than the thermal energy, the thermal energy becomes large enough to dissociate the charge carriers (polarons/bipolarons) and the residual electrons are scat-tered by thermal phonons. As a result, the charge mobility becomes decrea-sed while temperature is increasing. The dependence of the conductance of P5 on temperature can thus be interpreted as resulting from the change from a temperature-activated mechanism to a residual scattering mechanism. Thus, the conductance in the temperature region under investigation in this thesis implied that the activation barrier for electron transfer in the material was

66

significantly less than the thermal energy kBT (which is about 0.025 eV at room temperature). This low electron transfer activation energy hence ex-plains the fast charge transport discussed previously in Paper III.

Figure 22. a) The potential dependence of conductance of P5 in the temperature interval between 276 to 337 K. The arrow indicates the decrease of conductance G with increasing temperature to high potential. The corresponding current response of P5 is shown as inset. The polymer was cycled in MeCN with 0.1 M TEAPF6 by using an Ebias of 10 mV. b) The plot of G vs. temperature at different doping poten-tials. Adapted from Paper VI with permission from the publisher. Copyright 2017 PCCP Owner Societies.

5.6 Spectroelectrochemistry The absorption spectra of CPs upon doping are dominated by the band-gap transitions and the pronounced transitions between the two in-gap states, which are usually observed in the visible and near infared region. These in-gap states are formed with the combination of appearing charge carriers i.e. polarons and bipolarons57 in the material. The spin concentration can be followed by Electron Paramagnetic Resonance (EPR) study at various poten-tials. Thus, the spectroscopic techniques are provided as the tools to monitor the doping induced electronic transitions and charge carriers in the CRPs.

-0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

0

5

10

15

20

25

280 300 320 340 360

0

5

10

15

20

0.0 0.2 0.4 0.6 0.8

-0.1

0.0

0.1

0.2 b)a)

G (m

S)

E (V vs. Fc+/Fc0)

276 K 284 K 293 K 301 K 310 K 319 K 328 K 337 K

G (m

S)T (K)

0.40 V 0.45 V 0.50 V 0.55 V 0.60 V 0.65 V 0.70 V 0.75 V 0.80 V 0.85 V

I (

mA

)

E (V vs. Fc+ /Fc0)

67

Figure 23. (a) Attenuated total reflection-Fourier transform IR spectra and (b) UV-vis spectra of P5 at different potentials. The dashed arrows indicate the change in absorbance with changing potentials. The numbers indicate the electronic transitions (solid arrows in panel c) between the VB and CB or the in-gap states in the polaron and bipolaron states. (c) Bands and in-gap states energy diagram. The energy levels (green) in the VB and CB correspond to the π and π* orbitals of the DET moieties. Their locations were estimated according to Paper I and II. Adapted from Paper VI with permission from the publisher. Copyright 2017 PCCP Owner Societies.

Figure 23 shows the absorption changes in ATR-FTIR and UV-vis spectra measured in situ during the redox conversion of P5. In the low energy region (0.05 ~ 0.2 eV), the absorptions were derived from intense doping-induced infrared-active vibrations enhanced by the strong electron-phonon cou-plings.137-138 These vibrations are associated with the p-doped polythiophene backbone and its formed charge carriers (polarons). In the infrared high energy region (0.2 ~ 0.5 eV), a broad peak showed increasing absorbance at potentials above the onset of polymer doping. This is related to the electronic transition of polarons between the VB and the in-gap state (transition 3 in Figure 23c). From the low energy edge an energy difference of 0.17 eV is estimated for the energy gap between the polaron state and the VB. Thus the

0.00

0.01

0.02

0.03

0.04

0.1 0.2 0.3 0.4 0.5-0.04

-0.03

-0.02

-0.01

0.00

1.2 1.6 2.0 2.4 2.8 3.2 3.6 4.0 4.4

0.0

0.2

0.4

0.6

0.8

Abs

Wavelength / nmΔA

bs 0.1 V 0.2 V 0.3 V 0.4 V 0.5 V 0.6 V

VB

CBDeT

DeT

3

2

In-gap

In-gap

1

VB

CBDeT

DeT

4

In-gap

In-gap

VB

CBDeT

DeT

1

Neutral state Polarons state Bipolarons state

E / eV

Wavenumber / cm-1

E / eV

3

2

1

(c)

(b) 0 V 0.4 V 0.6 V 0.8 V 1 V

(a)

4

500 1000 1500 2000 2500 3000 3500 4000 1000 800 600 400

68

energy barrier for promoting electrons from VB to the polaron in-gap state is 0.17 eV based on IR measurements. This spectroscopic band gap is reflected by the dominated amorphous regions as confirmed by the X-ray diffraction spectroscopy (more details in Paper VI). The low activation energy involved in conduction may be accounted for by the predominant conduction occur-ring via interconnected crystal domains in the polymeric matrix. In other words, the crystal domains promote electronic coupling between the orbitals of adjacent polymeric chains. The high energy IR peak extended to the UV-vis region, where specific absorption features of the polymer were observed (Figure 23b). The possible electronic transitions involve VB, CB and the in-gap states, which are related to polarons and bipolaron states (Figure 23c). At neutral state (0 V vs. Fc+/Fc0) P5 showed a strong peak absorption at around 2.9 eV, with the edge absorption at around 2.2 eV which was very close to the optical band gap of PT56 (2.2 eV). This was assigned to the band gap transition from the backbone (transition 1). When the potentials went higher, transforming from modest (below 0.8 V vs. Fc+/Fc0) to high doping level (above 0.8 V vs. Fc+/Fc0), the main peak was suppressed and showed a blue shift absorption, the polymer film color changing from light orange to blue. The new peaks are corresponding to the in-gap states transition 2 and VB-occupied in-gap state transition 3 at the polaron state, and only transiti-ons in the IR region (transition 4) remained at bipolaron state. A similar ab-sorption evolution has been reported for PT.139-140 This similarity implies that the PGs have negligable influence on the optical properties of the polymer backbone. Moreover, it suggests that the nature of the charge carriers and their dependence on the doping level follow the same general trend for the CRP and the unsubstituted polymer analogue.

The transition from polaron to bipolaron state was further confirmed by an EPR study (Figure 24). The spin concentration (peak intensity) of P5 increased with the potential increasing from the doping onset potential to 0.8 V vs. Fc+/Fc0, from which point the spin concentration started to decrease. This implies that the concentration of free radicals (polarons, spin ½) increa-sed at low to modest doping levels and decreased at higher doping levels. The increase is accompanied by a linwidth narrowing, coinciding with the increase of conductance. It supports a delocalized nature of the charge car-riers in the polymer. With the formation of bipolarons at around 0.8 V vs Fc+/Fc0, the linewidth became broadened, which could indicate that the re-maining unpaired electrons became more localized. The transformation of charge carriers from polaron at low doping level to bipolaron at high doping level was also observed in PT, and the bipolaron formation was accompa-nied with a line broadening in the EPR signal (more detail in Paper VI). The PT had bipolaron formation at lower potential 0.4 V vs. Fc+/Fc0, which sug-gested that the polarons in P5 were more stable. As polarons are less stable in the amorphous phase than in ordered regions141 the delayed bipolaron formation in P5 suggests that the DET, as PG, promotes polymer ordering in

69

agreement with the XRD results. In addition to introducing additional prop-erties to CPs the inclusion of the redox active PGs thus has the potential to enhance electron transport, rather than preventing it, by promoting self-organization.

Figure 24. (a) EPR spectra of P5 at various potentials. (b) The spin concentration of the free radicals increased at the beginning of the doping and then decreased with increasing potentials, giving maximum spin intensity of the polarons in doped P5. The inset depicts the CV of the polymer on the EPR working electrode. Adapted from Paper VI with permission from the publisher. Copyright 2017 PCCP Owner Societies.

-0.2 0.0 0.2 0.4 0.6 0.8 1.0

-0.2 0.0 0.2 0.4 0.6 0.8 1.0E / V vs. Fc+/Fc0

Peak

inte

nsity

E / V vs. Fc+/Fc03420 3440 3460 3480 3500 3520 3540

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

(b)

Inte

nsity

Field / G

E / V

vs.

Fc+ /F

c0

(a)

70

6. Concluding Remarks

This thesis describes the study of n-type CRPs from molecular design, syn-thetic fabrication of monomers to the electrochemical and spectroscopical investigation of the corresponding polymers. These polymers combine the conducting properties from the main chain PT backbone with the redox pro-perties from the DET pendant. They are designed for the anode materials for secondary battery applications. This type of CRPs can solve the two major problems for organic molecules in EES applications: the capacity decay caused by dissolution of active material and the limited conductivity of small molecules. Thus, the CRPs are provided as one possible solution to overco-me these problems. Some specific remarks with regard to the study are as follows:

To obtain the CRPs requires thoughtful design of the monomer synthesis. Different redox PG might tolerate different synthetic environments in wet chemistry labs. In Paper II the synthesis of DET-funtionalized thiophene monomers had been adjusted to tolerate ester funtional groups. Due to the interest in stability and easier n-doping or p-doping of the PEDOT CP back-bone, an alkyne-functionalized EDOT monomer precursor was designed and synthesized in Paper V. The advantage of such a thiophene synthetic precur-sor is that the synthesis allowes for generally mild conditions to tolerate a variety of functional redox pendants. Additionally, the resulting all-carbon linker between backbone and pendant is much more inert to the redox and acid-base environment than a linker involving heteroatoms would be.

The redox process of DET overlaps well with n-doping of PT as it has been investigated in Paper I. It was shown that the presence of DET pendant molecules in the electrolyte solution did not significantly affect the poly-merization of thiophene monomers or the conductivity and doping behavior of the afforded polymers. However, the charge capacity was limited due to the disproportional redox reaction of DET and the second electron reduction ocurred at lower potential causing irreversible reaction.

The nature of the linker between polymer backbone and the DET pendant affects the electrochemical properties of CRPs to a large extend as discussed in Papers II and IV. It affects the polymerizability, the redox potential and the electrochemical stability of the materials by controlling the interaction between the thiophene unit and the DET pendant. This study finds that a non-conjugated linker is prefered for the monomer design as it gives less

71

adverse affects on the polymerization position on the thiophene ring and preserves the individual electrochemical behavior of backbone and pendant.

Fast electron transfer was obtained from the CRPs in Paper III. No diffu-sion limitation of electrons or counterions was observed in polymer films of thickness up to ~ 10 μm. This poperty makes the CRPs interesting for high power application in EES. The study suggests that the fast charge transfer can be derived from i) a low electron transfer activation barrier involved in the charge propagation within the material as evidenced by the temperature dependent conductance measurements; ii) the potential match between the pendant redox reaction and the polymer conduction which allows for charge propagation through the CP backbone; iii) the open porous morphology of the material allowing for fast ion diffusion.

The electrochemical instability of such CRPs in the n-doping region is the main challenge to overcome before facilitating their application in EES. The study showed that the instability of P5 was due to the detachment of the polymer film from the electrode surface as evidenced by the EQCM studies in Paper VI. Further studies by ATR-FTIR spectroscopy indicated that the influx of solvent molecules onto the electrode surface contributed to this detachment during n-doping. Possible solutions to this problem could be adding binders or modifying the electrode surface to improve the adhesion to the surface, or using solid electrolyte. The study showed that the replacement of the PT backbone with the PEDOT backbone resulted in better stability for the CRPs. However, more design and synthetic work are still needed to be done to adjust the CRPs for a longer cycle life.

72

7. Svensk sammanfattning

En av den nuvarande tekniska utvecklingens mest utmanande frågorna är samhällets energibehov kan tillhandahållas på ett hållbart och ändå tillförlit-ligt sätt. Vissa viktiga hållbara energikällor är endast tillgängliga med stora temporära variationer som kräver intermediära energilagringsalternativ. Fos-sila energikällor ersätts av elektrisk energi som måste vara tillgänglig utan att det krävs permanent anslutning till elnätet. Nya smarta enheter kräver stora antal av mindre enheter för energilagring med låg kapacitet. Medan nuvarande batteriteknik i princip kan uppfylla dessa behov finns det en nackdel på hållbarhetssidan. Batteritekniken domineras för närvarande av oorganiska material som litium, kobolt och andra metaller för vilka den framtida tillgängligheten är osäker. Dessa material har också dåliga livscy-kelegenskaper, bland annat krävs stora mängder energi för både produktion och återvinning.

Ett alternativ kan vara batterimaterial baserade endast på organiska före-ningar. Sådana material skulle vara tillgängliga i stora mängder från bio-massa och kunde vid slutet av deras livstid brytas ner i enkla molekyler, såsom koldioxid, kväveoxider, svaveldioxid och vatten som kan vara källan till ny biomassa. Men innan detta scenario kan realiseras måste flera hinder övervinnas.

Organiska material är vanligen isolatorer, vilket hindrar flödet av elektrisk ström som är nödvändigt för användningen av ett batteri. Laddning och ur-laddning kräver att elektroner absorberas eller släpps ut av det organiska materialet, i combination med en lämplig kemisk redoxreaktion. Försök har gjorts att blanda redoxaktiva organiska föreningar med låg molekylvikt med ledande polymermaterial. Även om detta tillvägagångssätt enkelt ur ett ke-miskt perspektiv försämras de resulterande materialen lätt genom att de redoxaktiva föreningarna läcka ut ur blandningen, och kan även ha problem med låg elektrisk ledningsförmåga. Detta kan förebyggas genom att binda de redoxaktiva föreningarna till en matris av ledande polymerer via kovalenta bindningar. I praktiken syntetiseras en polymer som inkorporerar de redox-aktiva föreningarna som en del av dess molekylstruktur. Ett sådant tillväga-gångssätt är en större kemisk utmaning, vilket kräver noggrann utformning av det molekylära materialet. I synnerhet måste de två komponenterna vara väl matchade i deras elektrokemiska egenskaper, och monomererna måste kunna polymeriseras för att ge de erforderliga materialen. I avhandlingen beskrivs sätt att hantera dessa utmaningar.

73

Figur 1: Uppbyggnad av elektriskt ledande redoxpolymerer (CRP).

Kort sagt behövs så kallade elektriskt ledande redoxpolymerer (CRP). Dessa innehåller en elektriskt ledande polymerkedja till vilken redoxaktiva mole-kylenheter är fästa via en kovalent länk. Som polymerkedja har vi valt poly-tiofener, och mera specifikt en substituerad version där en dioxolanring är bunden sida vid sida till tiofenringen, vilket resulterar i en styvare polymer-ryggrad. Tidigare studier har visa tatt en flexible polymerryggrad kan leda till problem i samband med laddning och urladdning via de redoxaktiva gruppernas reaktioner. De erhållna polymererna blir ledande vid avlägsnande eller tillsats av elektroner, så kallad p- eller n-dopning. För de redoxaktiva molekylenheterna har flera redoxaktiva grupper undersökts, såsom kinoner och tereftalsyraestrar. En viktig del av projektet har varit att ta fram en mo-dulär syntetisk metod baserat på alkynkemi, vilket gör det enkelt att fästa ett stort urval av redoxaktiva grupper till samma tiofenderivat och därmed kunna optimera redoxpolymererna. Det har också visat sig att synteserna lätt kan skalas upp till tiotals gram. Dessa föreningar används i relaterade projekt inom forskargruppen. Länken, som i huvudsak ger en koppling mellan po-lymeren och den redoxaktiva gruppen, måste också uppfylla vissa elektro-niska och elektrokemiska egenskaper. Dessa har undersökts genom beräk-ningar, där relevansen av beräkningsresultaten har verifieras genom syntes följd av elektrokemisk karakterisering.

74

Figur 2: Funktion av ett batteri med elektriskt ledande redoxpolymerer

(CRP) som elektrodmaterial.

En serie CRP har sålunda designats och syntetiserats, och de har undersökts med avseende på deras elektrokemi, kinetik, ledningsmekanism, stabilitet vid upprepade laddning och urladdning, masseffekter och spektroelektro-kemi.

75

8. Acknowledgements

This thesis project was carried out between the Division of Organic Chemis-try, Department of Chemistry - BMC and the Division of Nanotechnology and Functional Materials, Department of Engineering Science at Uppsala University. It would not be possible without the collaboration among the people in the groups. Thank you all for the help and support throughout the-se enjoyable years. First of all, I would like to thank my two supervisors: Prof. Adolf Gogoll, for enrolling me in this challenging project and inspiring me the philosophy of science and training me how to become an independent scientist; Prof. Mar-tin Sjödin, for enlarging my knowledge on physical chemistry and inspiring me what attitude and personality should be needed to become a scientist as great as you are. I could not expect better supervisors than both you. No matter where I go in future, you are always my mentors both in the science and philosophy. Of course, the study would not be such wonderful without the support from my family. Especially thanking my wife Dr. Cuiyan Li, she is such a great lady who is able to bring the warmness to the man heart and to radiate the kindness to surroundings. The home you bring is the warmest and most en-joyable moments that I have ever experienced. I would also thank my early dead son Tieven. You gave me the feeling of being a father and teach me how life is unpredictable. Your suddenly leaving is the worst moment I have, but it trained me to be a stronger person and taught me the meaning of life. I would also like to express my gratitude to the following people: My lovely group AGHG: Dr. Adolf Gogoll, Dr. Helena Grennberg, Dr. Mi-chael Nordlund, Dr. Hao Huang, Dr. Magnus Blom, Dr. Anna Lundstedt, Ms. Sandra Olsson, Dr. Leili Tahershamsi. Thank you all for being the family in BMC and all the wonderful moments we have experienced together throughout the 5 years.

76

My second group HO: Dr. Henrik Ottosson, Mr. Kjell Jorner, Dr. Rabia Ayub, Ms. Aleksandra Denisova, Dr. Muhammad Rouf Alvi, Dr. Raffaello Papadakis. You are all lovely persons to be sharing an office with and pro-vided me a comfort working environment and a chance to know you. Dr. Li Yang, as my core collaborator, thank for your contribution on electro-chemistry in this project and nice knowledge sharing and discussion between us. Members of related projects, for many interesting and fruitful discussions: Rikard Emanuelsson, Hao Huang, Christoffer Karlsson, Petter Tammela, Christian Strietzel, Mia Sterby, Lisa Åkerlund, Chao Xu, Zhaohui Wang, Maxim Galkin, Christian Dahlstrand. My dear friends: Kjell Jorner, Rabia Ayub, Jiajie Yan, Jia-fei Poon, Dániel Kovács, Fu Le, Pengcheng Wang, Jing Lu, Chungang Feng, Qian Zhao, Lichuan Wu, Huan Wang, Mariia Pavliuk, Jonas Karlsson. Thank you for having the chance to meet you great people and sharing the treasurable life moments together. Also thank you for caring me when I encountered one of the most difficult time in my life.

Kjell, you are the greatest theoretical chemist that I have ever met. I learnt a lot of knowledge from you, not only chemistry but also the way of organizing work and life. As a friend I am appreciated on your personality and wisdom of treating surroundings.

Rabia, your kindness and careness are radiating surroundings. As three years office neighbor your strength on treating difficult research is touching and it approves you are a strong and responsible lady. As a friend your con-sideration of others is grateful and wish Allah be always with you and bless you the whole life.

Jiajie and Jia-fei, your names are too similar in English letters and so con-fusing. Both of you are great chemists and I learnt a lot from you on synthet-ic chemistry. As friends you are kind and enjoyable to be together. Especial-ly I am appreciated on your memory Jiajie, your hippocampus is the best one I have ever met.

Dániel, the happy and great Hungarian that rush into my life. I am appre-ciated on your responsibility to the work and your attitude on making things the best and seriousness to the knowledge. My working experience with you as lab teachers makes me feel you will be a great chemist in future.

Fu Le, you are a reliable and thoughtful man, full of ambience of litera-ture and arts. Being friend with you I feel the elegancy and politeness when treating things. Your simplicity on cooking food could still make the meal such enjoyable, which inspires me and evidence me the real cooking master.

Pengcheng and Jing, we meet each other late but time cannot deny such great and kind couple engaging into my life. It is so fan and funny to talk with you and to share the wonderful life moments with you. Your happiness

77

and careness make us forget the annoyance. Wish the superpower from uni-verse be with you and bless you.

Chungang and qian, we met each other under the roof of holy Christian chapel. It seems to be doomed I will learn the humble and wisdom from you. Thank you also for sharing the knowledge of biology and agriculture.

Lichuan, we are from the same hometown and talking beside you gives me the feeling of comfort. Your research on meteorology gets me to know a new area.

Huan, thank you for taking care and being beside us when Cuiyan and I were experiencing the darkest moments. Your kindness and support take us though and I will be grateful for my whole life.

Masha, thank for knowing you in Uppsala. You evidence me the true heart of kindness and teach me the angle of seeing things, being with you the whole world is sunshine and colorful. Your broad heart is inspiring and leans me to a positive view of life.

Jonas, thank you for being my mentor on Swedish language. As friend you are always helpful and humble. Your kind heart is of glory and pure. You evidence me how the constancy can be when insisting on learning things. The last but not the least, I would thank my parents and my mother-in-law, your great support and love are the endless power of my life.

78

9. References

1. Li, G.; Zhu, R.; Yang, Y. Nat. Photonics 2012, 6, 153-161. 2. Dyer, A. L.; Thompson, E. J.; Reynolds, J. R. ACS Appl. Mater. Inter. 2011, 3,

1787-1795. 3. Xia, L.; Wei, Z. X.; Wan, M. X. J. Colloid. Interf. Sci. 2010, 341, 1-11. 4. Snook, G. A.; Kao, P.; Best, A. S. J. Power Sources 2011, 196, 1-12. 5. Heinze, J.; Frontana-Uribe, B. A.; Ludwigs, S. Chem. Rev. 2010, 110, 4724-

4771. 6. Novak, P.; Muller, K.; Santhanam, K. S. V.; Haas, O. Chem. Rev. 1997, 97,

207-281. 7. Janoschka, T.; Hager, M. D.; Schubert, U. S. Adv. Mater. 2012, 24, 6397-6409. 8. Rosciano, F.; Salamone, M. M.; Ruffo, R.; Sassi, M.; Beverina, L. J.

Electrochem. Soc. 2013, 160, A1094-A1098. 9. Karlsson, C.; Huang, H.; Stromme, M.; Gogoll, A.; Sjodin, M. RSC Adv. 2015,

5, 11309-11316. 10. Aydin, M.; Esat, B.; Kilic, C.; Kose, M. E.; Ata, A.; Yilmaz, F. Eur. Polym. J.

2011, 47, 2283-2294. 11. Casado, N.; Hernandez, G.; Veloso, A.; Devaraj, S.; Mecerreyes, D.; Armand,

M. Acs Macro Lett. 2016, 5, 64-69. 12. Liang, Y. L.; Tao, Z. L.; Chen, J. Adv. Energy Mater. 2012, 2, 742-769. 13. Holliday, B. J.; Swager, T. M. Chem. Commun. 2005, 23-36. 14. Conway, B. E., Electrochemical supercapacitors : scientific fundamentals and

technological applications. Kluwer Acad. [u.a.]: New York [u.a.], 2009. 15. Dahn, J. R.; Zheng, T.; Liu, Y. H.; Xue, J. S. Science 1995, 270, 590-593. 16. Armand, M.; Tarascon, J. M. Nature 2008, 451, 652-657. 17. Song, Z. P.; Zhou, H. S. Energ. Environ. Sci. 2013, 6, 2280-2301. 18. Larcher, D.; Tarascon, J. M. Nat. Chem. 2015, 7, 19-29. 19. Emanuelsson, R.; Sterby, M.; Stromme, M.; Sjodin, M. J. Am. Chem. Soc. 2017,

139, 4828-4834. 20. Yoshio, M.; Brodd, R. J.; Kozawa, A., Lithium-ion batteries science and

technologies. Springer: New York; London, 2009. 21. Williams, D. L.; Byrne, J. J.; Driscoll, J. S. J. Electrochem. Soc. 1969, 116, 2-&. 22. Liu, M. L.; Visco, S. J.; Dejonghe, L. C. J. Electrochem. Soc. 1989, 136, 2570-

2575. 23. Visco, S. J.; Liu, M.; Armand, M. B.; Dejonghe, L. C. Mol. Cryst. Liq. Cryst.

1990, 190, 185-195. 24. Tsutsumi, H.; Oyari, Y.; Onimura, K.; Oishi, T. J. Power Sources 2001, 92,

228-233. 25. Oyaizu, K.; Nishide, H. Adv. Mater. 2009, 21, 2339-2344. 26. Haupler, B.; Wild, A.; Schubert, U. S. Adv. Energy Mater. 2015, 5, 1402034. 27. Yokoji, T.; Matsubara, H.; Satoh, M. J. Mater. Chem. A 2014, 2, 19347-19354. 28. Armand, M.; Grugeon, S.; Vezin, H.; Laruelle, S.; Ribiere, P.; Poizot, P.;

Tarascon, J. M. Nat. Mater. 2009, 8, 120-125.

79

29. Walker, W.; Grugeon, S.; Vezin, H.; Laruelle, S.; Armand, M.; Tarascon, J. M.; Wudl, F. Electrochem. Commun. 2010, 12, 1348-1351.

30. Walker, W.; Grugeon, S.; Mentre, O.; Laruelle, S.; Tarascon, J. M.; Wudl, F. J. Am. Chem. Soc. 2010, 132, 6517-6523.

31. Walker, W.; Grugeon, S.; Vezin, H.; Laruelle, S.; Armand, M.; Wudl, F.; Tarascon, J. M. J. Mater. Chem. 2011, 21, 1615-1620.

32. Fedele, L.; Sauvage, F.; Becuwe, M. J. Mater. Chem. A 2014, 2, 18225-18228. 33. Park, Y.; Shin, D. S.; Woo, S. H.; Choi, N. S.; Shin, K. H.; Oh, S. M.; Lee, K.

T.; Hong, S. Y. Adv. Mater. 2012, 24, 3562-3567. 34. Zhao, L.; Zhao, J. M.; Hu, Y. S.; Li, H.; Zhou, Z. B.; Armand, M.; Chen, L. Q.

Adv. Energy Mater. 2012, 2, 962-965. 35. Wang, C. L.; Xu, Y.; Fang, Y. G.; Zhou, M.; Liang, L. Y.; Singh, S.; Zhao, H.

P.; Schober, A.; Lei, Y. J. Am. Chem. Soc. 2015, 137, 3124-3130. 36. Huskinson, B.; Marshak, M. P.; Suh, C.; Er, S.; Gerhardt, M. R.; Galvin, C. J.;

Chen, X. D.; Aspuru-Guzik, A.; Gordon, R. G.; Aziz, M. J. Nature 2014, 505, 195-198.

37. Winsberg, J.; Hagemann, T.; Janoschka, T.; Hager, M. D.; Schubert, U. S. Angew. Chem. Int. Ed. 2017, 56, 686-711.

38. Genorio, B.; Pirnat, K.; Cerc-Korosec, R.; Dominko, R.; Gaberscek, M. Angew. Chem. Int. Ed. 2010, 49, 7222-7224.

39. Geng, J. Q.; Bonnet, J. P.; Renault, S.; Dolhem, F.; Poizot, P. Energ. Environ. Sci. 2010, 3, 1929-1933.

40. Renault, S.; Geng, J. Q.; Dolhem, F.; Poizot, P. Chem. Commun. 2011, 47, 2414-2416.

41. Hanyu, Y.; Honma, I. Sci. Rep-Uk 2012, 2. 42. Huang, W. W.; Zhu, Z. Q.; Wang, L. J.; Wang, S. W.; Li, H.; Tao, Z. L.; Shi, J.

F.; Guan, L. H.; Chen, J. Angew. Chem. Int. Ed. 2013, 52, 9162-9166. 43. Muench, S.; Wild, A.; Friebe, C.; Haupler, B.; Janoschka, T.; Schubert, U. S.

Chem. Rev. 2016, 116, 9438-9484. 44. Gandini, A.; Lacerda, T. M. Progress in Polymer Science 2015, 48, 1-39. 45. Shirakawa, H.; Louis, E. J.; Macdiarmid, A. G.; Chiang, C. K.; Heeger, A. J. J.

Chem. Soc. Chem. Comm. 1977, 578-580. 46. Chiang, C. K. Polymer 1981, 22, 1454-1456. 47. Macinnes, D.; Druy, M. A.; Nigrey, P. J.; Nairns, D. P.; Macdiarmid, A. G.;

Heeger, A. J. J. Chem. Soc. Chem. Comm. 1981, 317-319. 48. Miller, J. S. Adv. Mater. 1993, 5, 587-589. 49. Moliton, A.; Hiorns, R. C. Polym. Int. 2004, 53, 1397-1412. 50. Tolbert, L. M. Accounts Chem. Res. 1992, 25, 561-568. 51. Andre, J. M.; Delhalle, J. Chem. Rev. 1991, 91, 843-865. 52. Yamamoto, T.; Koizumi, T. Polymer 2007, 48, 5449-5472. 53. Yen, W.; Chan, C. C.; Jing, T.; Jang, G. W.; Hsueh, K. F. Chem. Mater. 1991, 3,

888-897. 54. Vernitskaya, T. V.; Efimov, O. N. Uspekhi Khimii 1997, 66, 489-505. 55. Ciric-Marjanovic, G. Synthetic Met. 2013, 177, 1-47. 56. Kaneto, K.; Kohno, Y.; Yoshino, K. Solid State Commun. 1984, 51, 267-269. 57. Bredas, J. L.; Street, G. B. Accounts Chem. Res. 1985, 18, 309-315. 58. Patil, A. O.; Heeger, A. J.; Wudl, F. Chem. Rev. 1988, 88, 183-200. 59. Fesser, K.; Bishop, A. R.; Campbell, D. K. Phys. Rev. B 1983, 27, 4804-4825. 60. Sato, M. A.; Tanaka, S.; Kaeriyama, K. Synthetic Met. 1987, 18, 229-232. 61. Ding, H.; Pan, Z.; Pigani, L.; Seeber, R.; Zanardi, C. Electrochim. Acta 2001,

46, 2721-2732. 62. Laforgue, A.; Simon, P.; Fauvarque, J. F. Synthetic Met. 2001, 123, 311-319.

80

63. Skompska, M.; Mieczkowski, J.; Holze, R.; Heinze, J. J. Electroanal. Chem. 2005, 577, 9-17.

64. Mastragostino, M.; Soddu, L. Electrochim. Acta 1990, 35, 463-466. 65. Pokhodenko, V. D.; Krylov, V. A. Synthetic Met. 1991, 41, 533-536. 66. Zotti, G.; Schiavon, G.; Zecchin, S. Synthetic Met. 1995, 72, 275-281. 67. Pokhodenko, V. D.; Krylov, V. A.; Konoshchuk, N. V. Synthetic Met. 1999, 99,

91-95. 68. Krische, B.; Zagorska, M. Synthetic Met. 1989, 28, C257-C262. 69. Levi, M. D.; Aurbach, D. J. Power Sources 2008, 180, 902-908. 70. Nokami, T.; Matsuo, T.; Inatomi, Y.; Hojo, N.; Tsukagoshi, T.; Yoshizawa, H.;

Shimizu, A.; Kuramoto, H.; Komae, K.; Tsuyama, H.; Yoshida, J. J. Am. Chem. Soc. 2012, 134, 19694-19700.

71. Haupler, B.; Burges, R.; Friebe, C.; Janoschka, T.; Schmidt, D.; Wild, A.; Schubert, U. S. Macromol Rapid Comm. 2014, 35, 1367-1371.

72. Haupler, B.; Hagemann, T.; Friebe, C.; Wild, A.; Schubert, U. S. ACS Appl. Mater. Inter. 2015, 7, 3473-3479.

73. Kassam, A.; Burnell, D. J.; Dahn, J. R. Electrochem. Solid. St. Lett. 2011, 14, A22-A23.

74. Sharma, P.; Damien, D.; Nagarajan, K.; Shaijumon, M. M.; Hariharan, M. J. Phys. Chem. Lett. 2013, 4, 3192-3197.

75. Nakahara, K.; Iwasa, S.; Satoh, M.; Morioka, Y.; Iriyama, J.; Suguro, M.; Hasegawa, E. Chem. Phys. Lett. 2002, 359, 351-354.

76. Nishide, H.; Oyaizu, K. Science 2008, 319, 737-738. 77. Nishide, H.; Suga, T. Electrochem. Soc. Interface 2005, 14, 32-36. 78. Suga, T.; Pu, Y. J.; Oyaizu, K.; Nishide, H. B. Chem. Soc. Jpn. 2004, 77, 2203-

2204. 79. Suga, T.; Sugita, S.; Ohshiro, H.; Oyaizu, K.; Nishide, H. Adv. Mater. 2011, 23,

751-754. 80. Blanco, R.; Gomez, R.; Seoane, C.; Segura, J. L.; Mena-Osteritz, E.; Bauerle, P.

Org. Lett. 2007, 9, 2171-2174. 81. Arias-Pardilla, J.; Otero, T. F.; Blanco, R.; Segura, J. L. Electrochim. Acta 2010,

55, 1535-1542. 82. Park, K. S.; Schougaard, S. B.; Goodenough, J. B. Adv. Mater. 2007, 19, 848-

851. 83. Sen, S.; Saraidaridis, J.; Kim, S. Y.; Palmore, G. T. R. ACS Appl. Mater. Inter.

2013, 5, 7825-7830. 84. Gadgil, B.; Damlin, P.; Aaritalo, T.; Kvarnstrom, C. Electrochim. Acta 2014,

133, 268-274. 85. Andrieux, C. P.; Audebert, P. J. Electroanal. Chem. 1990, 285, 163-175. 86. Grimshaw, J.; Perara, S. D. J. Electroanal. Chem. 1990, 281, 125-132. 87. Karlsson, C.; Huang, H.; Stromme, M.; Gogoll, A.; Sjodin, M. Electrochim.

Acta 2015, 179, 336-342. 88. Sterby, M.; Emanuelsson, R.; Huang, X.; Gogoll, A.; Stromme, M.; Sjodin, M.

Electrochim. Acta 2017, 235, 356-364. 89. Platonov, A. Y.; Evdokimov, A. N.; Kurzin, A. V.; Maiyorova, H. D. J. Chem.

Eng. Data. 2002, 47, 1175-1176. 90. Kolb, H. C.; Finn, M. G.; Sharpless, K. B. Angew. Chem. Int. Ed. 2001, 40,

2004-2021. 91. Zhang, L.; Chen, X. G.; Xue, P.; Sun, H. H. Y.; Williams, I. D.; Sharpless, K.

B.; Fokin, V. V.; Jia, G. C. J. Am. Chem. Soc. 2005, 127, 15998-15999. 92. Worrell, B. T.; Malik, J. A.; Fokin, V. V. Science 2013, 340, 457-460. 93. Miyaura, N.; Suzuki, A. Chem. Rev. 1995, 95, 2457-2483.

81

94. Stille, J. K. Angew. Chem. Int. Edit. 1986, 25, 508-523. 95. King, A. O.; Okukado, N.; Negishi, E. I. J. Chem. Soc. Chem. Comm. 1977,

683-684. 96. Tamao, K.; Sumitani, K.; Kumada, M. J. Am. Chem. Soc. 1972, 94, 4374-4376. 97. Sonogashira, K. J. Organomet. Chem. 2002, 653, 46-49. 98. Heck, R. F.; Nolley, J. P. J. Org. Chem. 1972, 37, 2320-2322. 99. Yamamoto, T. Macromol Rapid Comm. 2002, 23, 583-606. 100. Yang, L.; Mihali, V. A.; Brandell, D.; Stromme, M.; Sjodin, M. J. Phys. Chem.

C 2014, 118, 25956-25963. 101. Schottland P., S. O., Le Gall P.-Y., Chevrot C. Journal de Chimie Physique et

de Physico-Chimie Biologique 1998, 95, 1258-1261. 102. Hu, D. F.; Lu, B. Y.; Duan, X. M.; Xu, J. K.; Zhang, L.; Zhang, K. X.; Zhang, S.

M.; Zhen, S. J. RSC Adv. 2014, 4, 35597-35608. 103. Balog, M.; Rayah, H.; Le Derf, F.; Salle, M. New J. Chem. 2008, 32, 1183-1188. 104. Segura, J. L.; Gomez, R.; Blanco, R.; Reinold, E.; Bäuerle, P. Chem. Mater.

2006, 18, 2834-2847. 105. Conte, S.; Rodriguez-Calero, G. G.; Burkhardt, S. E.; Lowe, M. A.; Abruna, H.

D. RSC Adv. 2013, 3, 1957-1964. 106. Sassi, M.; Mascheroni, L.; Ruffo, R.; Salamone, M. M.; Pagani, G. A.; Mari, C.

M.; D'Orazio, G.; La Feria, B.; Beverina, L. Org. Lett. 2013, 15, 3502-3505. 107. Patai, S., John Wiley & Sons Ltd.: Bristol, 1978; Vol. Patai Series: The

Carbon-Carbon Triple Bond Volume 1 & 2. 108. Stang, P. J. D., F., Modern Acetylene Chemistry. VCH Verlagsgesellschaft mbH:

Weinheim, 2007. 109. Chinchilla, R.; Najera, C. Chem. Rev. 2014, 114, 1783-1826. 110. Boyarskiy, V. P.; Ryabukhin, D. S.; Bokach, N. A.; Vasilyev, A. V. Chem. Rev.

2016, 116, 5894-5986. 111. Atkins, P. W.; Friedman, R., Molecular quantum mechanics. 4th ed.; Oxford

University Press: New York, 2005; p xiv, 573 p. 112. Seminario, J. M.; Politzer, P., Modern density functional theory : a tool for

chemistry. Elsevier: Amsterdam ; New York, 1995; p xi, 405 p. 113. Namazian, M.; Lin, C. Y.; Coote, M. L. J. Chem. Theory Comput. 2010, 6,

2721-2725. 114. Marenich, A. V.; Cramer, C. J.; Truhlar, D. G. J. Phys. Chem. B 2009, 113,

6378-6396. 115. Ben-Na`îm, A., Statistical thermodynamics for chemists and biochemists.

Springer: New York, 2013. 116. Truhlar, D. G.; Garrett, B. C.; Klippenstein, S. J. J. Phys. Chem. 1996, 100,

12771-12800. 117. Ribeiro, R. F.; Marenich, A. V.; Cramer, C. J.; Truhlar, D. G. J. Phys. Chem. B

2011, 115, 14556-14562. 118. Zhu, T. H.; Li, J. B.; Hawkins, G. D.; Cramer, C. J.; Truhlar, D. G. J. Chem.

Phys. 1998, 109, 9117-9133. 119. Hodgson, J. L.; Namazian, M.; Bottle, S. E.; Coote, M. L. J. Phys. Chem. A

2007, 111, 13595-13605. 120. M. J. Frisch, G. W. T., H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R.

Cheeseman, G. Scalmani, V. Barone, G. A. Petersson, H. Nakatsuji, X. Li, M. Caricato, A. Marenich, J. Bloino, B. G. Janesko, R. Gomperts, B. Mennucci, H. P. Hratchian, J. V. Ortiz, A. F. Izmaylov, J. L. Sonnenberg, D. Williams-Young, F. Ding, F. Lipparini, F. Egidi, J. Goings, B. Peng, A. Petrone, T. Henderson, D. Ranasinghe, V. G. Zakrzewski, J. Gao, N. Rega, G. Zheng, W. Liang, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y.

82

Honda, O. Kitao, H. Nakai, T. Vreven, K. Throssell, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, T. Keith, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, J. M. Millam, M. Klene, C. Adamo, R. Cammi, J. W. Ochterski, R. L. Martin, K. Morokuma, O. Farkas, J. B. Foresman, and D. J. Fox Gaussian 09, Revision A.02. 2016.

121. Zhao, Y.; Truhlar, D. G. Theor. Chem. Acc. 2008, 120, 215-241. 122. Lu, T.; Chen, F. W. J. Comput. Chem. 2012, 33, 580-592. 123. Lu, T.; Chen, F. W. Acta Chim. Sinica 2011, 69, 2393-2406. 124. Leriche, P.; Blanchard, P.; Frere, P.; Levillain, E.; Mahon, G.; Roncali, J. Chem.

Commun. 2006, 275-277. 125. Semenikhin, O. A.; Ovsyannikova, E. V.; Ehrenburg, M. R.; Alpatova, N. M.;

Kazarinov, V. E. J. Electroanal. Chem. 2000, 494, 1-11. 126. Laviron, E. J. Electroanal. Chem. 1979, 101, 19-28. 127. Ofer, D.; Crooks, R. M.; Wrighton, M. S. J. Am. Chem. Soc. 1990, 112, 7869-

7879. 128. Zotti, G. Synthetic Met. 1998, 97, 267-272. 129. Park, L. Y.; Ofer, D.; Gardner, T. J.; Schrock, R. R.; Wrighton, M. S. Chem.

Mater. 1992, 4, 1388-1395. 130. Kaiser, A. B. Adv. Mater. 2001, 13, 927-941. 131. Levi, M. D.; Gofer, Y.; Aurbach, D.; Lapkowski, M.; Vieil, E.; Serose, J. J.

Electrochem. Soc. 2000, 147, 1096-1104. 132. Shaktawat, V.; Jain, N.; Saxena, R.; Saxena, N. S.; Sharma, T. P. J.

Optoelectron. Adv. Mater. 2007, 9, 2130-2132. 133. Masubuchi, S.; Kazama, S. Synthetic Met. 1995, 74, 151-158. 134. Lee, K.; Cho, S.; Park, S. H.; Heeger, A. J.; Lee, C. W.; Lee, S. H. Nature 2006,

441, 65-68. 135. Bubnova, O.; Khan, Z. U.; Wang, H.; Braun, S.; Evans, D. R.; Fabretto, M.;

Hojati-Talemi, P.; Dagnelund, D.; Arlin, J. B.; Geerts, Y. H.; Desbief, S.; Breiby, D. W.; Andreasen, J. W.; Lazzaroni, R.; Chen, W. M. M.; Zozoulenko, I.; Fahlman, M.; Murphy, P. J.; Berggren, M.; Crispin, X. Nat. Mater. 2014, 13, 190-194.

136. Coropceanu, V.; Cornil, J.; da Silva, D. A.; Olivier, Y.; Silbey, R.; Bredas, J. L. Chem. Rev. 2007, 107, 926-952.

137. Hayes, W.; Pratt, F. L.; Wong, K. S.; Kaneto, K.; Yoshino, K. J Phys C Solid State 1985, 18, L555-L558.

138. Kvarnstrom, C.; Neugebauer, H.; Ivaska, A.; Sariciftci, N. S. J. Mol. Struct. 2000, 521, 271-277.

139. Keiichi, K.; Shogo, U.; Katsumi, Y.; Yoshio, I. Jpn. J. Appl. Phys. 1984, 23, L189.

140. Jiang, X. Q.; Patil, R.; Harima, Y.; Ohshita, J.; Kunai, A. J. Phys. Chem. B 2005, 109, 221-229.

141. Kenji, M.; Makoto, H.; Shin-ichi, M.; Shigeo, K.; Kiyoshi, K. Jpn. J. Appl. Phys. 1994, 33, L1239.

Acta Universitatis UpsaliensisDigital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology 1604

Editor: The Dean of the Faculty of Science and Technology

A doctoral dissertation from the Faculty of Science andTechnology, Uppsala University, is usually a summary of anumber of papers. A few copies of the complete dissertationare kept at major Swedish research libraries, while thesummary alone is distributed internationally throughthe series Digital Comprehensive Summaries of UppsalaDissertations from the Faculty of Science and Technology.(Prior to January, 2005, the series was published under thetitle “Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology”.)

Distribution: publications.uu.seurn:nbn:se:uu:diva-334562

ACTAUNIVERSITATIS

UPSALIENSISUPPSALA

2017