ACTAUNIVERSITATIS

UPSALIENSISUPPSALA

2019

Digital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology 1819

Electrochemical Characterizationsof Conducting Redox Polymers

Electron Transport in PEDOT/Quinone Systems

MIA STERBY

ISSN 1651-6214ISBN 978-91-513-0674-2urn:nbn:se:uu:diva-383026

Dissertation presented at Uppsala University to be publicly examined in Häggsalen, 10132,Ångström, Lägerhyddsvägen 1, Uppsala, Friday, 30 August 2019 at 09:15 for the degreeof Doctor of Philosophy. The examination will be conducted in English. Faculty examiner:Professor Peter Pickup (Memorial University of Newfoundland · Department of ChemistryCanada).

AbstractSterby, M. 2019. Electrochemical Characterizations of Conducting Redox Polymers. ElectronTransport in PEDOT/Quinone Systems. Digital Comprehensive Summaries of UppsalaDissertations from the Faculty of Science and Technology 1819. 59 pp. Uppsala: ActaUniversitatis Upsaliensis. ISBN 978-91-513-0674-2.

Organic electrode materials for rechargeable batteries have caught increasing attention sincethey can be used in new innovative applications such as flexible electronics and smartfabrics. They can provide safer and more environmentally friendly devices than traditionalbatteries made from metals. Conducting polymers constitute an interesting class of organicelectrode materials that have been thoroughly studied for battery applications. They have highconductivity but are heavy relative to their energy storage ability and will hence form batterieswith low weight capacity. Quinones, on the other hand, are low weight molecules that participatein electron transport in both animals and plants. They could provide batteries with high capacitybut are easily dissolved in the electrolyte and have low conductivity. These two constituents canbe combined into a conducting redox polymer that has both high conductivity and high capacity.In the present work, the conducting polymer PEDOT and the simplest quinone, benzoquinone,are covalently attached and form the conducting redox polymer used for most studies in thisthesis. The charge transport mechanism is investigated by in situ conductivity measurementsand is found to mainly be governed by band transport. Other properties such as packing,kinetics, mass changes, and spectral changes are also studied. A polymerization technique isalso analyzed, that allows for polymerization from a deposited layer. Lastly, two different typesof batteries using conducting redox polymers are constructed. The thesis gives insight into thefundamental properties of conducting redox polymers and paves the way for the future of organicelectronics.

Keywords: Conducting Redox Polymer, PEDOT, Quinone, Charge transport, Conductivity,Organic Battery

Mia Sterby, Department of Engineering Sciences, Nanotechnology and Functional Materials,Box 534, Uppsala University, SE-75121 Uppsala, Sweden.

© Mia Sterby 2019

ISSN 1651-6214ISBN 978-91-513-0674-2urn:nbn:se:uu:diva-383026 (http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-383026)

Ju mer man läser och ju mer man förstår,

desto mer man begriper hur litet man vet,

ty livet är kort och snabbt det förgår

och på det mesta i livet så går vi dock bet

Edvin ”Pang” Malmsjö

List of Papers

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

I Sterby M., Emanuelsson R., Huang X., Gogoll A., Strømme M., and Sjödin M. (2017) Characterization of PEDOT-Quinone Conducting Redox Polymers for Water Based Secondary Bat-teries. Electrochimica Acta, 235:356-364

II Emanuelsson R., Sterby M., Strømme M., and Sjödin M. (2017) An All-Organic Proton Battery. Journal of the American Chem-ical Society, 139(13):4828–4834

III Sterby M., Emanuelsson R., Strømme M., and Sjödin M. (2019) Investigating electron transport in a PEDOT/Quinone conducting redox polymer with in situ methods. Electrochimica Acta, 308:277-284

IV Sterby M., Strietzel C., Emanuelsson R., Strømme M., and Sjödin M. (2019) Post-Deposition Polymerization – A Method for Circumventing Processing of Insoluble Conducting Poly-mers. Manuscript

Reprints were made with permission from the respective publishers.

The author´s contribution to the included papers

Paper I. I participated in the planning of the study and performed all the electrochemical experiments as well as associated data analysis. I wrote the initial manuscript and contributed to the continued writing process.

Paper II. I participated in the planning of the study and performed some data analysis. I contributed to the writing process of the manuscript.

Paper III. I participated in the planning of the study and performed all the electrochemical experiments as well as associated data analysis. I wrote the initial manuscript and contributed to the continued writing process.

Paper IV. I participated in the planning of the study and performed all the experiments, except synthesis, as well as associated data analysis. I wrote the initial manuscript and contributed to the continued writing process.

Also published

Sjödin M., Karlsson C., Huang H., Sterby M., Gogoll A., Emanuelsson R., and Strømme M. (2015) Quinone Based Conducting Redox Polymers for Electrical Energy Storage. 10th International Frumkin Symposium on Elec-trochemistry, Moscow (Russia) Sterby M., Emanuelsson R., Strømme M., Gogoll A., and Sjödin M. (2016) Characterization of a PEDOT/Quinone Conducting Redox Polymer. Gordon Research Conference and Seminar: Electronic Processes in Organic Mate-rials, Barga (Italy) Sterby M., Emanuelsson R., Strømme M., Gogoll A., and Sjödin M. (2016) Linker Effect in PEDOT/Quinone Organic Battery Materials. 67th Annual Meeting of the International Society of Electrochemistry, The Hague (The Netherlands) Sjödin M., Huang H., Emanuelsson R., Sterby M., Huang X., Yang L., Karlsson C., Gogoll A., and Strømme M. (2016) Conducting Redox Poly-mers for Secondary Batteries. 67th Annual Meeting of the International Soci-ety of Electrochemistry, The Hague (The Netherlands) Sterby M., Emanuelsson R., Strømme M., and Sjödin M. (2017) Electro-chemical Characterization of Conducting Redox Polymers – Redox Match-ing and Mass Transport. E-MRS Spring Meeting, Strasbourg (France) Sterby M., Emanuelsson R., Strømme M., and Sjödin M. (2017) Electro-chemical Properties of PEDOT-Based Conducting Redox Polymers. Organic Battery Days, Uppsala (Sweden) Sjödin M., Emanuelsson R., Sterby M., Strietzel C., Yang L., Huang H., Wang H., Huang X., Gogoll A., and Strømme M. (2017) Conducting Redox Polymer Based Batteries. Organic Battery Days, Uppsala (Sweden) Emanuelsson R., Sterby M., Strømme M., and Sjödin M. (2017) All-Organic Proton Batteries from Conducting Redox Polymers. Organic Battery Days, Uppsala (Sweden)

Sjödin M., Emanuelsson R., Sterby M., Åkerlund L., Huang H., Huang X., Gogoll A., and Strømme M. (2017) Organic Batteries Based on Quinone-Substituted Conducting Polymers. IUPAC 17th International Symposium on MacroMolecular Complexes (MMC-17), Tokyo (Japan)

Sterby M., Emanuelsson R., Strømme M., and Sjödin M. (2018) Electronic properties of a PEDOT/Quinone Conducting Redox Polymer. Gordon Re-search Conference and Seminar: Electronic Processes in Organic Materials, Barga (Italy)

Sterby M., Emanuelsson R., Strømme M., and Sjödin M. (2018) In-situ Methods for Understanding Charge Transport in a Conducting Redox Poly-mer. MRS Fall Meeting & Exhibit, Boston (USA)

Contents

1. Motivation ........................................................................................... 13

2. Aim of Thesis ...................................................................................... 14

3. Redox Active Organic Materials for Battery Applications .................. 15 3.1 Redox Reaction .................................................................................. 15 3.2 Conducting Polymers ......................................................................... 16

3.2.1 Polymerization ............................................................................ 17 3.2.2 Conductivity ............................................................................... 18 3.2.3 PEDOT ....................................................................................... 21

3.3 Conducting Redox Polymers .............................................................. 23 3.4 Quinones............................................................................................. 23

3.4.1 Proton Activity............................................................................ 25

4. Electrochemical Methods .................................................................... 27 4.1 Cyclic Voltammetry ........................................................................... 27 4.2 In Situ Methods .................................................................................. 29

4.2.1 In Situ EQCM ............................................................................. 29 4.2.2 In Situ IDA ................................................................................. 30 4.2.3 In Situ EPR ................................................................................. 31 4.2.4 In Situ UV-vis ............................................................................. 31

5. Quinone Based CRPs and Post-Deposition Polymerization ................ 32 5.1 Polymerization ................................................................................... 33

5.1.1 Electrochemical Polymerization ................................................. 33 5.1.2 Post-Deposition Polymerization ................................................. 33

5.2 Characterizations ................................................................................ 36 5.2.1 Substitution Effect ...................................................................... 37 5.2.2 pH-Dependence .......................................................................... 38 5.2.3 Mass Changes ............................................................................. 38 5.2.4 Conductance ............................................................................... 39 5.2.5 Temperature Dependence ........................................................... 40 5.2.6 In Situ Spectroscopy ................................................................... 41 5.2.7 Trimer Packing ........................................................................... 43

6. Batteries ............................................................................................... 45 6.1 Basic Batteries .................................................................................... 45

6.1.1 Testing a Battery ......................................................................... 46 6.2 pH Battery .......................................................................................... 46 6.3 All-Organic Proton Battery ................................................................ 48

7. Concluding Remarks ........................................................................... 50

8. Sammanfattning på svenska ................................................................ 52

9. Acknowledgement ............................................................................... 54

10. References ........................................................................................... 56

Abbreviations

Ac Acetylene AQ Anthraquinone BG Band gap BQ Benzoquinone CB Conduction band CE Counter electrode CP Conducting polymer CRP Conducting redox polymer CV Cyclic voltammogram EDOT 3,4-Ethylenedioxythiophene FS Fluorescence HOMO Highest occupied molecular orbital HQ Hydroquinone LUMO Lowest unoccupied molecular orbital MO Molecular orbital PAc Polyacetylene PEDOT Poly(3,4-ethylenedioxythiophene) RE Reference electrode SHE Standard hydrogen electrode UV-vis Ultraviolet-visible VB Valence band WE Working electrode

Symbols

Symbol Name Unit E Potential V E Energy eV E0 Standard potential V E0´ Formal potential V F Faraday constant C mol-1 G Conductance S I Current A m Mass kg n Amount of substance mol Q Charge C R Gas constant J mol-1 K-1 z Number of charges - λ Wavelength nm

13

1. Motivation

A more environmentally friendly society is necessary to ensure a tolerable future for generations to come. One major issue today is the use of fossil fuels for energy production. Combustion of fossil fuels emits large amounts of carbon dioxide into the atmosphere which contribute to global warming. We are using more and more energy on this planet, and to ensure a more sustainable society the energy needs to be produced from renewable energy sources instead, such as wind and solar power. The drawback of these meth-ods is that they do not produce energy on demand since the wind does not always wine and the sun does not always shine. Batteries can stabilize the energy output by working as a buffer between the energy production and energy users.

Most commercial batteries contain different kind of metals that are nonre-newable and leave a large carbon footprint with their mining and refining processes. To find more sustainable energy storage materials one can instead look toward biomass and will find a wealth of organic molecules used for charge storage in different systems.

Organic materials for energy storage can have several advantages over in-organic materials:1-2 the basic building blocks (H, C, O, N, and S) are abun-dant and accessible in nature from biomass, they can easily be recycled, they can be more light-weight than metals, and their properties can be fine-tuned by substitution to fit various applications. The electrolyte used for batteries based on organic materials can be water based, something that does not work for lithium batteries due to lithium’s instability in water. Using water as elec-trolyte instead provides improved safety of the batteries since they are non-flammable. Organic batteries can also be used for applications that lithium-based batteries are not suitable for, e.g. in flexible electronics, for applica-tions in smart fabric that can be washed without a safety risk, and in applica-tions were the battery cannot easily be regained but instead can be combust-ed with its device.

Batteries continue to be the limiting factor when developing portable elec-tronics and new systems that have the potential to provide increased stabil-ity, higher specific capacity, and increased safety etc. need to be found. Or-ganic batteries have the possibility for new ground breaking discoveries that could move the field of batteries forward faster and make a big leap towards a more environmentally friendly society.

14

2. Aim of Thesis

The overall aim of this thesis has been to investigate the fundamental proper-ties of conducting redox polymers. Focus has been on the redox reactions of the polymer backbone and the pendant group as well as possible interaction between them. This is of importance for conducting redox polymers since separate properties of the backbone and the pendant group are desired, and they should not have a negative impact on each other.

Polypyrrole has previously been studied as backbone for conducting re-dox polymers but was found to twist and loose conductivity during the redox reaction of the pendant group.3 This work has instead focused on the con-ducting polymer PEDOT that is often used for its stability and rigidity. The work has been focused towards materials for battery applications and two battery systems have also been tested. The specific aims of this work are as follows:

• To study the redox process of the quinone pendant group with re-gards to kinetics, pH dependence, and mass changes during the re-dox reaction (Paper I and III).

• To investigate the charge transport mechanism through the polymer

backbone (Paper I and III).

• To investigate the relationship between the pendant group and the backbone with regards to redox matching and the pendant groups in-fluence on conductivity (Paper I, II and III).

• To understand the mechanisms occurring during post-deposition

polymerization of trimers (Paper IV).

• To analyze the effect of packing in trimer solids (Paper IV).

• To use conducting redox polymers as active material in both the pos-itive and the negative electrode in all-organic batteries (Paper II).

15

3. Redox Active Organic Materials for Battery Applications

There is much research on organic materials for lithium ion batteries (LIBs) to replace the heavy metals currently used in the positive electrode.1, 4-6 There are also some attempts at constructing all-organic batteries, which do not include any metals.7-9 There are several different categories of organic materials for battery applications,10 both for LIBs and all-organic batteries, including radicals,11-12 conducting polymers (CPs),13-15carbonyl compounds,6,

16-17 and organosulfur compounds.18-20 The energy storage mechanism differ between the different categories and they have different strengths and weak-nesses. Radicals usually have a high redox potential, which is beneficial for the positive side of a battery and show a rapid charge transport through the material, but can be chemically unstable.10, 21-22 One of the most studied radi-cals for battery applications is 2,2,6,6-Tetramethylpiperidine 1-oxyl (TEM-PO). TEMPO has a robust radical which can be both oxidized and reduced, allowing it to be used both as p- and n-type materials.23 CPs have high con-ductivity but low capacity and will be discussed in detail in Section 3.2. Car-bonyl compounds have high charge storage capacity, by the redox reaction of carbonyls, and they can be tuned to different potentials and can therefore be used as both negative and positive electrode.10, 24-25 A commonly studied class of carbonyl compounds is quinones, which will be discussed further in Section 3.4. The last category of organic materials for battery applications is organosulfur compounds. They utilizes the breaking and reforming of disul-fide bonds which is a slow reaction, but organosulfur compounds can have a high energy density.10

3.1 Redox Reaction The main requirement for materials applied in rechargeable batteries is to be reversibly redox active, i.e. to be oxidized and reduced repeatedly, without major losses in capacity. The redox reaction occurring in a battery material is: + ⇌ (1)

16

where Ox is the oxidized species, Red the reduced species, and z the number of electrons involved in the reaction. The potential of a half-cell reaction is defined as the reduction potential and the potential of a full cell reaction is calculated as the reduction potential of the species being reduced minus the reduction potential of the species being oxidized. The potential of a redox reaction is always measured against a reference electrode (RE) and is de-fined against the standard hydrogen electrode (SHE), which is the reduction of hydrogen on a platinum electrode. A Ag/AgCl electrode, with a known potential vs. SHE, is normally used in water electrolytes and the potential is then recalculated to give a value vs. SHE. In organic electrolytes the poten-tial is instead reported vs. the ferrocene redox potential (Fc0/+).

The Gibbs free energy (ΔG) of a reaction is related to the potential (E) by: = − (2)

where F is the Faraday constant. A higher reduction potential means a higher driving force for the reducing reaction since ΔG gets more negative. The standard potential (E0) of a reversible redox reaction is defined at standard conditions as the potential where the activity of all reactants is 1. E0 is com-pared to the SHE at the same temperature. When standard conditions are not met, the potential can be calculated by Nernst equation:26

= + × (3)

where R is the gas constant, T the temperature, F the Faraday constant, and Cox and Cred the concentration of oxidized and reduced forms respectively.

3.2 Conducting Polymers Most organic materials are insulators and require conducting additives for electrons to be able to move through the whole material.27 This is of im-portance for battery applications in order to utilize the entire capacity of the material. No conducting additives are required when using CPs, since they can be made conducting on their own. CPs are used in a wide range of appli-cations including organic light emitting diodes,28 sensors,29-30 and solar cells.31

Polymers consist of many small units, called monomers, which are cova-lently bonded together forming a repetitive chain. Most CPs have a conju-gated backbone, i.e. a continued π-system. This can be achieved by double bonds between carbon atoms and with heteroatoms such as N, O, and S. Non-conjugated polymers can have charge transport only through hopping, a

17

mechanism discussed in Section 3.2.2. The fact that polymers can be con-ducting was first discovered in 197732 and was awarded a Nobel Prize in Chemistry 2000.33

3.2.1 Polymerization The simplest CP is polyacetylene (PAc) which is built from the monomer acetylene (Ac), H-C≡C-H, and consists of a long chain of sp2 hybridized carbon atoms with one hydrogen bonded to each carbon. Molecular orbitals (MOs) of PAc can be constructed by doing a linear combination of atomic orbitals.34 The orbitals of interest for PAc are the π-orbitals and the π*-orbitals since they are the ones forming the highest occupied molecular or-bital (HOMO) and the lowest unoccupied molecular orbital (LUMO), and only these orbitals will be considered further on (the σ-orbitals are lower in energy and the σ*-orbitals are higher in energy). Two MOs are formed when combining two atomic orbitals, one filled having lower energy and one emp-ty having higher energy, Figure 1a. The sp-hybridized carbons in Ac con-tribute with two p-orbitals each, but the py- orbitals will be hybridized in the dimer, forming σ-bonds between Ac-units. This leaves only the pz-orbitals which will contribute to the π-orbitals in the polymer, and thus only the pz-orbitals are depicted in Figure 1. When Ac is dimerized, the two MOs of each Ac unit combine into four new MOs with two filled π-orbitals, Figure 1b. The dimers then combine to form tetramers, Figure 1c, and so on, with a narrowing of the HOMO-LUMO gap when longer and longer chains are formed. Theoretically, HOMO and LUMO would therefore have no energy difference in an infinite long polymer, forming a continuous band of energy levels that is half-filled. The polymer would then have metallic conductivity since no energy is required for electrons to move to higher energy levels.35 For this to be true, the electrons would have to be delocalized over the whole polymer. However, the π-electrons are actually localized over the double bond, creating alternating single and double bonds in the polymer. This is due to the Peierls distortion in which a localization of π-electrons stabilizes the π-orbitals.34-35 Even though the energy of the filled σ-orbitals increase, the total energy of the system decreases. The MOs have now generated two bands, one filled band that is called the valence band (VB) and one empty band called the conduction band (CB), Figure 1d. These bands have a band gap (BG) between them and the polymer is an insulator, or in some cases a semiconductor.35 The method used for making the polymers conducting is discussed in Section 3.2.2.

One common way of polymerizing CPs is by oxidation.36-39 There are two major techniques for oxidative polymerization from a monomer solution.

18

Figure 1. The polymerization steps of PAc with π-orbitals, showing the formation of two bands with a BG narrower than the HOMO-LUMO gap in Ac a) The pz-orbitals from the carbons in Ac combine to form two MOs. b) Dimerization of Ac forming four new π-orbitals. c) Two dimers combine to a tetramer with four filled and four empty π-orbitals. d) Two bands are formed for the polymer, one filled and one empty.

One is to add a chemical oxidant to the monomers and allowing the reaction to continue until the radical becomes stable or the polymer becomes large enough to precipitate. The other technique is to remove the electrons by ap-plying an oxidative potential through an electrode to the monomer solution and stop it when the desired degree of polymerization is reached. This will form a polymer on the electrode surface.

3.2.2 Conductivity The conductivity of polymers is defined as the ease of which electrons (or holes) can move through it. Polymers are insulators in their natural state and charges need to be introduced in a polymer in order to achieve conductivity.34 This can be accomplished by electrochemical methods or by adding a chemical oxidant or reductant. Charges can be formed by either adding or removing electrons, creating a negatively doped (n-doped) or posi-tively doped (p-doped) polymer respectively and conductivity is achieved after a threshold doping level is reached. In this thesis only p-doping is dis-cussed.

When an electron is removed from the polymer, a polaron is formed, i.e. a positive charge together with a radical that is delocalized over several mon-omer units, as depicted in Scheme 1b. Polaron formation results in sp3 hy-bridized carbons on the chain and the bonds in the polymer can change char-acter to minimize the energy of the system. This distortion causes the intro-duction of new energy levels in the BG that are taken from the VB and CB, Figure 3b, of which the lower one is half-filled.

19

Scheme 1. P-doping of PEDOT. a) Un-doped PEDOT. b) Polarons are formed by oxidation of PEDOT. The structure of the backbone changes to a quinoide structure between the positive charge and the radical. c) Upon further doping, two polarons can combine into a bipolaron.

Additional polarons are formed by removing more electrons from the poly-mer and two polarons can combine to form a bipolaron, a species with two positive charges and no unpaired electrons, as seen in Scheme 1c. The charge repulsion between the two positive charges in a bipolaron will favor the polaron state. But the distortion will be almost the same for two polarons and one bipolaron which will favor the formation of bipolarons. The new energy levels in the BG are empty for bipolarons, Figure 3c.

The conductivity is dependent on the number of charge carriers and their mobility. The number of charge carriers can be controlled by the extent of oxidation but has a maximum value due to charge repulsion. The mobility of positive charge carriers always move in the direction of the applied electrical field and can be affected by different variables depending on which charge transport mechanism occur. There are two major conductivity mechanisms for CPs, band transport and hopping.34 The band transport only transpires in conjugated polymers and consists of charge carriers moving along the poly-mer chain over the delocalized orbitals, Figure 2a. Lattice vibrations, also known as phonons, disrupt the charge transport and causes a lower mobility.

20

Figure 2. Charge transport mechanisms with the electric field direction marked. Ex-amples of a) band transport, b) interchain hopping, and c) intrachain hopping.

The number of phonons increases with temperature and the mobility is thus proportional to the inverse temperature. When the polymer chain ends, is not parallel to the applied electrical field, or has a kink so that the charge gets trapped, the charge carriers have to jump to another polymer strand. This jumping is a process called interchain hopping, Figure 2b. Hopping, i.e. thermally activated tunneling, requires an activation energy to move from one orbital to the next. The activation energy is easier to overcome at higher temperatures and charge transport through hopping displays a positive tem-perature dependence. Interchain hopping occurs between neighboring close-ly packed polymer chains which makes the polymer morphology important. Hopping can also occur intrachain when the polymer has kinks or other de-fects, Figure 2c, or for non-conjugated polymers with aromatic pendant groups (PGs). The more localized charge carriers in a polymer, the more hopping will occur, resulting in lower mobility.

Charge neutrality needs to be restored when positive charges, both polar-ons and bipolarons, are introduced in the polymer. This can be accomplished by incorporation of anions from the electrolyte and/or expulsion of cations. The polymer can be forced to cycle cations if the anion is covalently at-tached to the polymer, or is too bulky and thus immobile. Solvent molecules can follow the ions which will lead to a swelling or shrinking of the polymer film.

3.2.2.1 Spectroscopic Transitions A CP in the un-doped state has one major spectral transition, i.e. the BG transition, Figure 3a. The BG transition, called transition 1 in Figure 3, is typically between 1.5 eV and 4 eV for conjugated polymers40 which allows it to be investigated by UV-visible (UV-vis) spectroscopy. The new energy levels that are created when polarons are formed, Figure 3b, are taken from the orbitals closest to the BG in the VB and CB. This results in a slightly larger BG and a blueshift in the spectrum. The polaron energy levels also

21

Figure 3. Possible spectral transitions for CPs for a) un-doped polymer, b) polaron states, and c) bipolaron states.

allows for two new transitions, one from the VB to the lower polaron state, transition 2 in Figure 3, and one between the two in-gap states, transition 3 in Figure 3. These transitions will both lie at lower energies than the BG in a spectrum, so for polymers with a low BG energy these transitions might be outside the UV-vis spectral window. Transition 3 disappears for bipolarons, since the lower in-gap state is now empty, and only transition 1 and 2 are possible.

3.2.3 PEDOT Poly(3,4-ethylenedioxythiophene) (PEDOT) is a CP that was first reported in 1991 by Jonas et al.41 It is widely studied due to its stability, rigidity, and high electrical conductivity.34, 42-44 PEDOT consists of repeating units of EDOT, Scheme 2, that are connected at the α-carbons and forming a π-conjugated system containing C-C double bonds and sulfur atoms.

22

Scheme 2. Polymerization of EDOT through radical formation. Two radicals combine with the expulsion of two electrons and two protons. After multiple steps, this process eventually leads to the formation of a PEDOT polymer.

Polymerization of PEDOT starts with the oxidation of shorter segments, usually monomers, to form radicals, Scheme 2. Two radical monomer seg-ments then combine to form a dimer. This process continues, with larger and larger segments (dimers combine to tetramers, tetramers combine to octam-ers, etc.) with the expulsion of two protons in each step, until the polymeri-zation process stops, as described in Section 3.2.1.

The polymerization process of PEDOT is in principal without defects (side reaction that does not form the desired polymer) due to carbon 3 and 4 being substituted, in contrast to thiophene.43

The amount of charges that are introduced in a polymer is called doping level. The maximum number of charges introduced in PEDOT has been re-ported to one charge per three monomer units, i.e. a doping level of 0.33, but this can be affected by the environment around the polymer.44-46

PEDOT is insoluble in aqueous solutions but can be mixed with polysty-rene sulfonate (PSS) to improve processability. PSS consists of immobilized anions that can counteract the charges formed on PEDOT during p-doping. PEDOT:PSS is the most successful CP in practical applications due to a high electrical conductivity and stability42, e.g. in solar cells,47-48 capacitors,49-50 and thermoelectric materials.51-52

3.3 Conducting Redox Polymers For some applications a specific property is wanted, e.g. light absorption, catalysis, or charge storage, and a specific molecular structure is required to gain that property. The desired structure can be incorporated in the CP back-

23

bone, but this could alter the conjugation of the backbone and reduce the conductivity.53

Instead of changing the conducting backbone of the polymer to fit the re-quirements, a PG with the desired property can be covalently attached to it. If the PG is redox active the combination can be called a conducting redox polymer (CRP). The PG can be added to give different properties to the CRP, such as electrochromism, electrical energy storage, or photosensitizing effects. To preserve the individual properties of the backbone and CP, the PG should not be in conjugation with the backbone as it could disturb the conductivity of the backbone or change the properties of the PG.3 This is controlled by the choice of linker between PG and backbone. A longer and more flexible linker have been shown to best make use of the separate prop-erties of the PG and backbone since a stiff linker can cause a twist in the backbone upon oxidation which will result in a loss of conductivity due to a loss in conjugation. The linker can also affect the redox potential of the PG since a substitution on a redox system can alter its energy levels.

An important concept for CRPs is redox matching, i.e. that at the redox potential of the PG, the polymer backbone has to be conducting. This needs to be considered when choosing what PG and backbone to use.

Polypyrrole with a quinone PG has previously been investigated as CRP for energy storage3 but the redox chemistry of the PG was found to negative-ly influence the conductivity of the polymer. A PEDOT backbone will be used in this thesis instead, since it has high stability.

CRPs can be used for e.g. different types of sensors with an organometal-lic PG54-55, batteries with a terephthalate56 or quinone57-59 PG, electrochromic devices with a naphthalene diimide PG60, and photochromic devices with a diarylethene PG.61

This thesis will focus on CRPs for energy storage. In CRPs for battery applications the CP is responsible for the conductivity while the small PG gives the high charge storage capacity and the constant redox potential need-ed for a stable battery voltage.

3.4 Quinones Quinones belong to the carbonyl compounds category of organic battery materials and have been used as redox active species for the CRPs in this thesis. A wide range of quinones have previously been studied for different battery applications.62-64 Quinones are naturally occurring and can be found in both the photosynthesis of plants and bacteria65-67 and the respiratory sys-tem in animals68-69 as charge transport units.

The simplest quinone, a benzoquinone (BQ), can be seen in Scheme 3. The BQ can, in protic electrolytes, be reduced to a hydroquinone (HQ) by the addition of two electrons and two protons.

24

Scheme 3. BQ to HQ redox conversion, a 2e2H reaction.

The reduction of BQ to HQ can take several routes depending on experi-mental conditions. The square scheme of the proton coupled BQ/HQ redox reaction, Figure 4, shows the different routes, with proton transfer in hori-zontal direction and electron transfer in vertical direction. The semiquinone (SQ) is unstable and will disproportionate into BQ, named Q in Figure 4, and HQ, named QH2 in Figure 4, since two SQs have a higher total energy than one Q and one QH2.

Quinones can store two charges, or in some cases more, which give them a high charge storage capacity, ranging from 100 mAh/g to 600 mAh/g.62 Other quinones can be constructed by adding substituents to the benzene ring, which will change the energy levels of the quinone and thus giving it a different redox potential. The redox potential of the quinone is also depend-ent on the cation coordinating to the negatively charged oxygens in the

Figure 4. Square scheme of the quinone redox reaction from Q to QH2. H+ denotes proton reactions and e- denotes redox reactions. (For the sake of simplicity, BQ is denoted Q and HQ is denoted QH2 when talking about the square scheme).

25

Table 1. E0 of selected quinones in 1 M p-TsOH70 and the electronic properties of the substituents.71

Structure E0 (V vs. SHE) Electronic properties of substituents

0.64 -

0.60 EDG

0.57 EDG

0.09 EDG

0.71 EWG

0.70 EWG

reduced state. The redox potentials of quinones cycling lithium and sodium can, as an example, range from 2.0 to 3.3 V vs. Li+/Li in organic solvents.62 The substituents can be electron withdrawing groups (EWGs) or electron donating groups (EDGs) to the aromatic system of the quinone, compared to a proton. A more EWG will make the oxygen electron poor and thus making the oxy group more prone to accepting electrons, i.e. giving the quinone a higher redox potential. An EDG group will have the opposite effect and low-er the redox potential of the quinone. A study of quinones in 1 M p-TsOH70 shows the impact of substituents on the redox potential. Selected quinones are listed in Table 1.

3.4.1 Proton Activity The potential at which the BQ to HQ process occurs is dependent on the proton activity in the electrolyte. A higher proton activity, i.e. a lower pH in a water electrolyte, will make the reduction reaction from BQ to HQ, for-ward reaction in Scheme 3, easier and thus giving it a higher potential, i.e. a lower energy. The pH dependence of the 2e2H quinone reaction can be ex-plained by the Nernst equation, Equation (3), with the insertion of the reac-tants and product in Scheme 3:

26

E0´=E0+ RT2F ln =E0+ RT2F 2 ln( H+ ) =E0-59mV∙pH (4)

where E0´ is the formal potential, i.e. the apparent standard potential that applies when pH differ from 0. The quotient of BQ and HQ are 1 at E0´ and with the temperature set to room temperature (25 °C) the pH dependence of the potential is -59 mV per pH unit.

The pH dependence applies in aqueous solutions but protons can also be provided by other molecules. Many organic electrolytes are non-protic and a protic salt that can provide both a proton donor and a proton acceptor is re-quired if a battery cycling protons are sought. Quinones can also cycle other cations such as lithium and sodium, but in this thesis only proton cycling is considered, since it opens for a metal free battery.

27

4. Electrochemical Methods

A three electrode setup is used in all the electrochemical experiments in this thesis, except for the battery tests. The three electrodes are 1) a working electrode (WE) on which the redox reactions of interest occur after applica-tion of a controlled potential and whose current response is monitored; 2) a counter electrode (CE) that counteracts the charges formed on the WE; and 3) a RE, which is a half cell with known redox potential, used in order to ensure that the correct potential is applied to the WE. The potential of the WE is given as a potential vs. the RE but can after the measurement be con-verted to a different reference system. The potentials in this thesis will be reported vs. SHE for all characterization experiments which are conducted in water electrolytes and vs. ferrocene (Fc0/+) for polymerizations done in ace-tonitrile (MeCN) for easy comparison to other research.

4.1 Cyclic Voltammetry One of the most frequently used characterization methods in electrochemis-try is cyclic voltammetry. In cyclic voltammetry the potential is scanned between a lower potential and a higher potential at a fixed scan rate while the current is monitored. Oxidation occur at the WE when the potential is swept towards higher potentials, and when the potential is reversed (going back to lower potentials), reduction takes place. Cyclic voltammetry allows for information about the redox processes occurring at different potentials for both monomers and polymers. Cyclic voltammetry is also a common method of polymerization.

The current response differs between a diffusion limited reaction and a surface bound entity.26 A narrow and reversible redox peak is seen in the cyclic voltammogram (CV) for a surface bound material. The peak current (ip) for this type of reaction is directly proportional to the scan rate (ν), and a plot of ln(ip) vs. ln(ν) will give a slope of 1.

The oxidation and reduction peaks move apart when the scan rate is in-creased since there is a potential gradient in the material and the outer layers feel the applied potential later than the inner layers.

The back reaction can be ignored when the peaks have moved 200 mV apart and the reaction can then be viewed as a surface bound irreversible reaction and the peak potential (Ep) is given by:

28

=E0´ + RTαF ln ∙ (5)

where α is the transfer coefficient and k0 the rate constant. The apparent rate constant can be calculated by measuring Ep at different scan rates.

A fully reversible surface bound reaction where the whole material can feel the same potential is called a Nernstian system. The potential is then defined by the Nernst equation, Equation (3), and the peak split is zero, Fig-ure 5a. A surface bound reaction with a peak split up to 100 mV is called a semi-reversible reaction. For species in solution, the redox reaction is limited by diffusion of reactants to the electrode. For this system ip is proportional to ν½, and a plot of ln(ip) vs. ln(ν) will give a slope of 0.5. The peak in the CV, Figure 5b, is usually broad with a diffusion tail and with a large peak split. The peak split and the diffusion tail appears due to that the reactants close to the electrode surface are used up and it takes time for new molecules to dif-fuse to the electrode.

The activation energy (Ea) for the rate determining step of the redox reac-tion can be calculated by determining the rate constant at different tempera-tures with the Arrhenius equation: =A /( ) (5) where A is a constant.

CRPs for battery applications show two different current responses in dif-ferent areas of a CV, Figure 5c. The constant potential before and after the large peak comes from the capacitive behavior of the conducting backbone, even though the reaction is faradaic in nature. The capacitive behavior arises when each new charge that is being introduced in the material can feel the other charges by columbic interactions and a gradually increasing potential

Figure 5. Simulated CVs of a) a surface bound material without diffusion limitation (a Nernstian system) and b) a dissolved redox molecule with diffusion limitation. c) CV of a CRP, showing current responses from both the backbone and PG.

29

is needed to overcome the columbic repulsion. This applies for all CPs. The peak in Figure 5c arises from a faradaic process, a semi-reversible surface bound reaction, occurring at the PG, where the pendant molecules have a specific redox potential and the different redox sites do not influence each other.

All the following methods, except in situ EPR, utilize cyclic voltammetry for characterization.

4.2 In Situ Methods The phrase “in situ” is Latin for “on site” and in electrochemistry it means that a change in potential or current is applied while at the same time another parameter is measured such as mass changes, conductivity changes, changes in spin characteristics, or spectral changes. This allows for direct and accu-rate measurements without the risk of a change in potential that could other-wise occur if there is a time delay between the change in potential and the measurement.

4.2.1 In Situ EQCM The mass changes occurring in a material can be monitored with an electro-chemical quartz crystal microbalance (EQCM). The EQCM works by the piezoelectric effect (more precisely, the converse piezoelectric effect) which creates a motion (vibration) in the quartz crystal when an electric current is applied to it.72 The vibrational frequency is dependent on the mass on the crystal. A higher mass loading will make the crystal move with a lower fre-quency. This relationship was described by Sauerbrey in 1959 and gave rise to the Sauerbrey equation:73 ∆ = −∆ × /2( ) (6)

where Δm is the mass change, Δf the frequency change, Fq the reference frequency, A the area, ρq the quartz crystal density, and μq the AT-cut quartz constant. The Sauerbrey equation is only valid for a thin and rigid film that do not deform when a force is applied to it.72 The best EQCM devices can determine amounts down to ng/cm2.

Information about which species are moving in and out of the film during polymerization and characterization can be gained from EQCM experiments. These measurements will more specifically provide information about e.g. how much solvent is gained or expelled during polymerization and during characterization and which ions are cycled to counteract the charges formed when applying a potential to the material.

30

4.2.2 In Situ IDA An interdigitated array (IDA) electrode can be used to measure the conduc-tivity of a material and consists of two gold electrodes with bands in comb-like structures that are interdigitated. The many bands in the interdigitated structure form a large surface area which allows for more current to run through the material than in an electrode with only two bands. The sensitivi-ty to conductivity changes is thus increased. Figure 6 illustrates the conduct-ance measurement setup with the IDA electrode, RE, and CE connected to a bipotentiostat. Included is also a simplified circuit diagram for the currents involved. The polymer covers both WEs and a voltage bias (Ebias) is applied between the two sides of the IDA electrode. A current will flow between the two electrodes when the polymer starts to conduct, i.e. the resistance is low-ered. The bias current (Ibias) is determined by the resistance through the ma-terial (Rpol) and the applied Ebias according to ohms law:

= (7)

The sum of all currents in a point is zero according to Kirchhoff´s law.74 This law for point A and B in Figure 6 is used together with Equation (7) to calculate the conductance: Point A: + = (8) Point B: = + (9) where IC1 and IC2 are the convection currents from the faradaic reaction. Equation (8) and (9) can be combined to: − ( + ) = + − (10) and with ΔI = I2 – I1 and ΔIC = IC2 – IC1 the expression can be written as: = 2 + ΔI (11) The bias current can thus be calculated by the following equation (with the assumption that IC2 and IC1 are identical or negligible compared to the bias current): I = 2 (12)

The conductance (G) of the material is the inverse of the resistance and can be calculated by combining Equations (7) and (12) to:

31

Figure 6. Schematic image of an IDA electrode in a three electrode setup. A circuit diagram is also included. G = 1 = I = 2 ∙ (13)

The current for the CV (Itot) is calculated by adding I1 and I2 together.

4.2.3 In Situ EPR Electron paramagnetic resonance (EPR) is a technique that can observe spins (un-paired electrons) in a material. This is of interest for CRPs, and CPs in general, since the polarons formed upon doping has a spin whereas the dou-bly charged bipolarons formed upon further doping does not. For EPR measurements, a platinum wire was coated with polymer and subsequently immersed in a thin EPR cell filled with electrolyte. Potential steps were ap-plied to the polymer during an EPR measurement, instead of using cyclic voltammetry, due to the long sampling time of the EPR.

4.2.4 In Situ UV-vis The spectral changes occurring in a CP when going from a non-doped state to a polaron state and further into a bipolaron state are seen in Figure 3 and the possible spectral transitions have been discussed in Section 3.2.2.1. The spectral changes occurring during polymerization and characterization can be monitored by in situ UV-vis spectroscopy when the polymer is applied to a conducting and UV-vis-transparent indium tin oxide coated quartz (ITO) electrode. The electrochemical experiments were conducted in a quartz cu-vette.

32

5. Quinone Based CRPs and Post-Deposition Polymerization

In this thesis, PEDOT, Figure 7a, and two PEDOT-HQ CRPs, Figure 7b and c, is characterized in aqueous solutions. The different characterization tech-niques results in a better understanding of how charge transport occur in CRP-systems and compare the CRPs to pristine PEDOT. Three EDOT-units is combined into a trimer which allows for polymerization from a deposited layer. This polymer is compared to PEDOT formed by EDOT monomers and the packing of the trimer is also investigated. PEDOT≡HQ, Figure 7b, is also used for a pH battery where the electrolytes on the positive side and the negative side have different pH and thus the redox potential of PEDOT≡HQ is different for the two electrodes. An all-organic battery is also studied. This battery utilizes PEDOT-O-C=O-HQ, Figure 7c, as the positive electrode and an anthraquinone (AQ) as PG connected to PEDOT as the negative elec-trode. This battery utilizes an organic protic electrolyte since PEDOT-AQ cannot be cycled in water electrolytes.

Figure 7. Structure of a) PEDOT, b) PEDOT≡HQ, c) PEDOT-O-C=O-HQ, and d) EEE.

33

5.1 Polymerization

5.1.1 Electrochemical Polymerization Electrochemical polymerization from a monomer solution is the polymeriza-tion method used for most experiments in this thesis. In this method, an elec-trode is immersed in a monomer solution and the potential is changed by cyclic voltammetry, Figure 8. Radicals can start to form when the potential reaches a certain value, around 0.8 V vs Fc0/+ in Figure 8, and polymeriza-tion is initialized. The polymerization continues until the potential is swept below the radical formation potential. The current response increases for every consecutive sweep, indicated by arrows in Figure 8, which is due to the increasing amount of polymer on the electrode and associated increase in doping currents and accessible quinones.

Figure 8. Polymerization of EDOT≡HQ in TBAPF6/MeCN at a scan rate of 100 mV/s.

5.1.2 Post-Deposition Polymerization One problem with many CPs is the processability of the formed polymer, which is important when increasing the scale of production. PEDOT:PSS, see Section 3.2.3, is one example of increased processability by allowing PEDOT to be used in aqueous suspensions. Another approached is to in-crease the solubility of the polymer by adding functional groups, either polar groups for use in water solvents or non-polar groups for organic solvents. These approaches are targeted towards specific solvents and a more general approach would benefit the CP community.

Post-deposition polymerization is a technique employed in our group that enables polymerization from a deposited layer, Paper IV. The layer can be

34

deposited and polymerized on the substrate directly, avoiding the problem with polymer processing altogether.

The principles behind the method is to dissolve the building blocks in one solvent, make a layer on a substrate and let it dry, then finally polymerize the layer in a solvent that does not dissolve the building blocks. The monomers that are normally used for polymerization would not work with this method since they dissolves in most solvents and have no conductivity. Trimers were instead chosen as building blocks in Paper IV, since they are insoluble in some solvents but soluble in others, and could have inherent conductivity. DMSO was used for dissolving the trimers and the polymerization was con-ducted by cyclic voltammetry in sulfuric acid (0.5 M).

Since PEDOT displays good electrochemical properties, see Section 3.2.3, a trimer with three EDOT units, named EEE, was made, Figure 7d. During post-deposition polymerization of EEE, a large increase in conduc-tivity can be seen at 0.7 V, Figure 9a, which coincide with the oxidation peak in the CV and the beginning of the spectral changes in the UV-vis spec-tra, Figure 9b.

In order to hinder polymerization, methyl groups were added at the α-position of EEE, forming MeEEEMe. This was done to investigate if the trimer was conducting on its own and not only after polymerization. The trimer shows no indication of polymerization from the UV-vis spectra, Fig-ure 9d, where a probable reorganization occur instead during the first scan.

Figure 9. a) Conductance and CV and b) Uv-vis spectra during post-deposition polymerization of EEE in H2SO4 (0.5M). c) Conductance and CV and d) Uv-vis spectra during cycling of MeEEEMe in H2SO4 (0.5M). Both trimers were dissolved in DMSO prior to deposition on the IDA and ITO electrodes.

35

A small conductance can be seen above 0.7 V, Figure 9c, showing that the trimer is slightly conducting when doped which could be an explanation why post-deposition polymerization works and is completed after only one cyclic voltammetry cycle.

The mass change during post-deposition was also studied for EEE, Figure 10a and c. The trimer mass loading on the electrode can be measured pre-cisely since the trimer is dissolved in DMSO at a certain concentration with-out any additives and then dried under vacuum. This is harder to do with polymerization from a monomer solution, since the material on the electrode being weighed is already a polymer with solvent and ions incorporated. The mass loading was confirmed with the EQCM where the electrode was weighed before and after addition of the trimer. The mass loading of EEE was 22 µg which is 102 % of the theoretically calculated value. The number of water molecules per trimer that was incorporated in the polymer film dur-ing post-deposition polymerization was calculated from the mass changes at 0 V after the first cycle, Figure 10c, and charges extracted from Figure 10a,

Figure 10. a) CV and c) mass change during post-deposition polymerization of EEE. b) CV and d) mass change during characterization of pEEE.

36

and was found to be 14 molecules per trimer unit, which indicated a signifi-cant swelling of the polymer film (the number of water molecules per trimer is an overestimate since the polymer film cannot be completely de-doped in the electrolyte and some counter ions are still in the film at this potential). The doping level was calculated from the total trimer mass and from charac-terization of the 10th scan following polymerization, Figure 10b and d. The doping was 1.9 trimers per polaron, i.e. a doping level of 0.2 polarons per EDOT unit, which is close to the value of 0.3 polarons per EDOT discussed in Section 3.2.3.

5.2 Characterizations CRPs have two different current responses, see section 4.1. One comes from the capacitive behavior of the conducting backbone, dark colors in Figure 11a, and the other comes from the faradaic response of the PG redox reac-tion, bright colors in Figure 11a. The capacity originating from the redox

Figure 11. a) CV of PEDOT-O-C=O-HQ in nitrate buffer at 30 mV/s with a peak split of 17 mV, showing the different current responses. Dark blue corresponds to doping of the PEDOT backbone, bright blue corresponds to oxidation of the quinone PG, dark green corresponds to the de-doping of the PEDOT backbone, and bright green corre-sponds to the reduction of the quinone PG. b) CVs at scan rates between 10 mV/s (black line) and 30 V/s (brightest gray line). c) Peak potential vs. scan rate. d) Loga-rithmic graph of peak current vs. scan rate with a slope of 1.

37

reaction of the quinone is 83 % for PEDOT-O-C=O-HQ at 30mV/s and the remaining capacity of 17 % originates from the PEDOT backbone. Figure 11b shows CVs at scan rates between 10 mV/s and 30 V/s. The peak split is increasing with increasing scan rate, Figure 11c, and above 3 V/s the peak split is over 200 mV and Equation (5) is valid. The apparent rate constant can then be calculated and is 20 s-1 for the oxidation of PEDOT-O-C=O-HQ and 27 s-1 for the reduction. Figure 11d displays the linear relationship be-tween ln(Ip) and ln(ν) and the slope is 1 which says that the quinone redox reaction is a surface reaction without diffusion contributions. There is no diffusion tail even at high scan rates in the CVs in Figure 11b which strengthen the statement that the reaction is not diffusion limited.

5.2.1 Substitution Effect The linker connecting the PG to the backbone can affect the redox potential of the PG, Paper I. There is a 60 mV shift in redox potential when compar-ing PEDOT≡HQ, containing an alkyne linker, and PEDOT-O-C=O-HQ, containing an ester linker with aliphatic groups closest to the quinone, Figure 12. The alkyne is directly connected to the quinone and forces the quinone redox reaction to a higher potential due to its electron withdrawing ability.

Figure 12. Substitution effect. CVs in nitrate buffer (pH 0), showing the difference of an alkyne linker (black line) and an ester linker with alkane groups closest to the quinone (gray line) in PEDOT-HQ CRPs.

5.2.2 pH-Dependence The pH of the electrolyte can be changed to determine the pH dependence of the quinone, Paper I. The slope in Figure 13b shows a -61 mV behavior for PEDOT≡HQ which is close to the -59 mV per pH calculated for a quinone reaction, Equation (4). The formal potential was calculated to 0.70 V, which

38

Figure 13. pH dependence of PEDOT≡HQ in nitrate buffer (pH 0). The potential show a linear dependence with respect to pH and the slope was calculated to -61 mV per pH and the standard potential to 0.70 V. (The CV at pH 3 was measured after the polymer had been immersed in pH 10.)

is slightly higher than the 0.67 V for an unsubstituted BQ.75 The quinone displays one reversible redox reaction in the whole pH-interval measured.

5.2.3 Mass Changes The mass change, Figure 14, displays three distinct regions during oxidation of PEDOT≡HQ, Paper I. The first and third regions, dark grey in Figure 14, display a mass increase during polymer doping while the middle region, light grey in Figure 14, displays a mass decrease during oxidation of HQ. The average molecular weight of the species causing the mass changes is calculated from linear regions in a plot of mass change vs. charge. The aver-age molecular weight for the polymer doping regions is 22 g per mol charge and corresponds to a combination of influx and outflux of ions and solvent

Figure 14. Mass changes (solid line) and current response (dashed line) during cyclic voltammetry for PEDOT≡HQ in nitrate buffer (pH 0). Dark grey regions correspond to backbone doping and light grey region corresponds to quinone redox reaction.

39

molecules since 22 g/mol does not correspond to any of the involved cations or anions. The quinone PG region instead has an average mass decrease of 3 g/mol, which corresponds to the proton expulsion originating from the pro-ton release during oxidation of HQ to BQ. Only a small amount of solvent molecules move with the proton since 3 g/mol is close to the proton molar mass of 1 g/mol. The protons thus move through the polymer without the company of water molecules, which is possible due to the Grotthuss mecha-nism76 where a proton binds to a water molecule and forces another proton in that water molecule to leave. This proton binds to the next water and so on.

5.2.4 Conductance The current responses of PEDOT≡HQ, Paper III, from the two WEs in the IDA electrode are shown in Figure 15a. The currents move apart after -0.1 V which is a result of the polymer starting to get conducting, see Section 4.2.2. The quinone peak is visible on both electrodes and is equal in size, making Equation (12) valid and thus the conductance and total current, Figure 15b, can be calculated in accordance with Section 4.2.2. The conductivity rises quickly until a stable plateau is reached at around 0.1 V. The conductance stays at the same value before, during, and after the quinone redox reaction, which shows that the quinone redox reaction does not alter the conductivity of the backbone, as it did for the polypyrrole/quinone combination.3

The importance of redox matching is apparent in Paper II where non-redox matched systems results in loss of faradaic current from the quinone PG. Redox matching in PEDOT≡HQ is evident in Figure 15b where the redox reaction of the PG is well above the conductivity onset of the polymer.

PEDOT have a similar conductance value to PEDOT≡HQ, Paper III, but displays constant current responses during the whole potential region tested.

Figure 15. a) Current response of WE1 and WE2 in an IDA electrode in H2SO4 (0.5 M) and b) calculated conductance (solid line) and CV (dashed line) of PEDOT≡HQ. A 10 mV bias was applied between WE1 and WE2.

40

This shows that the PG in PEDOT≡HQ has not affected the mobility or number of charge carriers in the PEDOT backbone but the conductivity on-set is however affected, which could be due to morphology changes induced by the PG.

5.2.5 Temperature Dependence The conductance can be measured at different temperatures which will give information on the charge transport mechanism through the material, Paper III. A negative temperature dependence is seen for both PEDOT≡HQ, Fig-ure 16, and PEDOT. This indicates that band transport is a major contributor to the charge transport at these temperatures since the increased number of phonons at higher temperatures will cause increased scattering of the charge carriers and thus a lowered conductivity, see Section 3.2.2. A positive tem-perature dependence would be seen if hopping was the dominant part, unless the activation energy required for hopping was significantly smaller than the thermal energy around room temperature (26 meV), in which case the con-ductance would be temperature independent. Interchain hopping is usually involved when band transport occurs and the activation energy for this hop-ping is probably small for PEDOT and PEDOT≡HQ since it does not coun-teract the negative temperature dependence of the band transport.

The rate constant for the quinone redox reaction was also determined at different temperatures, Figure 17, and the activation energy, Equation (5), was calculated to 0.3 eV. The rate determining step for the quinone redox reaction is thus an activated process while the charge transport is a non-activated process. The charge transport through the backbone is thus not the rate determining step in the quinone redox conversion.

Figure 16. Temperature dependent conductance for PEDOT≡HQ. Showing a nega-tive temperature dependence.

41

Figure 17. Temperature dependence of the backbone conductance and the redox reac-tion of the PG for PEDOT≡HQ in H2SO4 (0.5M).

5.2.6 In Situ Spectroscopy The EPR spectra of PEDOT≡HQ, Figure 18a, displays one peak with a Lo-rentzian line shape, Paper III. Only one peak indicates that only one radical species is present in the material. The possible radicals in a PEDOT/quinone CRP are polarons and the semiquinone. The same EPR signal as for PE-DOT≡HQ was seen for pristine PEDOT, Paper III, which rules out the sem-iquinone. This is expected since the semiquinone rapidly disproportionate to BQ and HQ. The EPR signal then comes from polarons. The spin concentra-tion, Figure 18b, is proportional to the area under the integrated peak and is a measurement of the number of polarons in the polymer. It shows a slight decline at higher doping potentials which is in agreement with the combina-

Figure 18. a) EPR spectra and b) spin concentration (black squares) and signal width (gray circles) at different potentials for PEDOT≡HQ in H2SO4 (0.5M). Recorded in situ during potential steps.

42

tion of polarons into spinless bipolarons. The width of the EPR signal, Fig-ure 18b, increases with potential and has been connected to a decrease in mobility of the radical.77 The conductance, Figure 15b, stays at a plateau value at high doping levels while the number of polarons decline. Since the polarons at the same time become less mobile this is interpreted as bipolar-ons also being charge carriers.

The UV-vis absorption spectra of PEDOT≡HQ, Figure 19a, that were recorded during the oxidative sweep of cyclic voltammetry show a peak at 2.0 eV that are blueshifted and loses intensity upon doping. This peak corre-sponds to the BG transition of the polymer backbone, transition 1 in Figure 3. The peak at lower energy, 1.3 eV, also decreases in intensity upon doping and corresponds to polarons, transition 3 in Figure 3b, since it shows a simi-lar trend as the EPR results, Figure 18b. These peaks are also present in PE-DOT, Paper III, and all peaks go back to its original intensity after a reduc-tion to 0 V which show that all visible reactions are reversible. Transition 2 for polarons and bipolarons, Figure 3b and c, are at lower energy and outside the monitored spectral window. The spectra were subtracted with the spec-trum at 1 V for easier analysis of the peaks at higher energy, at 3.3 eV and

Figure 19. a) Absorption spectra and b) relative absorption (spectra are subtracted with the 1 V spectrum) of PEDOT≡HQ in H2SO4 (0.5M), recorded in situ during cy-clic voltammetry.

43

3.8 eV in Figure 19b. These two peaks show changes in absorption intensity after 0.6 V, the peak at 3.3 eV increases in intensity with increasing potential and thus corresponds to the BQ absorption while the peak at 3.8 eV decreas-es in intensity with increasing potential and corresponds to the HQ absorp-tion.

5.2.7 Trimer Packing The EEE trimer that was used for post-deposition polymerization was char-acterized by UV-vis and fluorescence (FS) spectroscopy, Paper IV. The trimer was studied both in solution and in solid form. The results are dis-played in Figure 20 and show an extra peak at 2.2 eV in the UV-vis spec-trum for the solid trimer, Figure 20a, compared to the trimer in solution, Figure 20b. This peak comes from a more packed region that only exists for the solid trimer. This region has increased orbital overlap, which decreases the energy difference between the HOMO and LUMO, thus forming a peak at lower energy in the UV-vis absorption spectrum.

In Figure 21, an energy level diagram displays the different transitions that give rise to the spectra for solid EEE. The absorption from the lowest vibrational state of the ground state to the lowest vibrational state in the first excited state is called the 0-0 transition. This is the absorption with lowest energy and reversely, the emission with the highest energy. The energy of this transition can be calculated by taking the interception between the ab-sorption spectra and the FS emission spectra. The energy of the 0-0 transi-tion is 3.0 eV for the trimer in solution and 2.8 eV for the large peak in the solid trimer spectrum. A small redshift, i.e. a lowered energy, is seen for the solid trimer in both the UV-vis and FS spectrum compared to the trimer in

Figure 20. a) UV-vis absorption spectrum (red line) and FS emission spectra (with excitation at 3.3 eV (yellow line), and with excitation at 2.2 eV (brown line)) for solid EEE. b) UV-vis absorption spectrum (blue line) and FS emission spectrum (with excitation at 3.1 eV (teal line)) for EEE in DMSO.

44

Figure 21. Energy level diagram of transitions in amorphous and more packed re-gions of EEE. The colors are corresponding to the absorptions and emissions dis-played in Figure 20. The black wavy lines show the vibrational relaxation to the lowest vibrational level.

solution. This shows that also in the more amorphous regions, the solid tri-mer has a slightly decreased energy gap between the ground state and the first excited state. The 0-0 transition for the packed region is 2.1 eV.

The trimer thus contains amorphous regions and more crystalline regions when in solid form. This is beneficial for charge transport since a more over-lapped orbital system can transport charges faster.

45

6. Batteries

The CRPs studied in this thesis have been targeted toward battery applica-tions. In this respect, two different battery types have been tested with CRPs at both electrodes. One in aqueous electrolyte, using PEDOT≡HQ as both positive and negative electrode, achieving a cell potential due to the shift in the quinone redox potential arising from a difference in pH of the electro-lytes. The other utilizes two different CRPs that have different redox poten-tials due to the substitution effect of quinones. This battery is using an or-ganic electrolyte slurry with both a proton acceptor and a proton donor.

6.1 Battery Basics A battery consists of two electrodes separated by a membrane and an elec-trolyte that can transport ions between the electrodes, Figure 22. An external circuit can be connected to the electrodes and thus allowing electrons to move between the electrodes. When the battery is charged, Figure 22a, the positive electrode material is oxidized and electrons move through the circuit to the negative electrode where the negative electrode material is reduced. Counterions simultaneously move through the membrane to counteract the charges formed by the electron movement. The battery works as an electro-lytic cell during charging and the process requires energy to push the elec-trons to the negative electrode and storing the electrical energy as chemical energy. This energy is then regained when the polarity is switched and elec-trons spontaneously move back to the positive side, Figure 22b. The battery now operates as a galvanic cell where the chemical energy stored at the elec-trodes is released and converted into electrical energy.

The redox potentials at which the redox reactions occur for the two elec-trode materials differ and the difference in potential gives the cell voltage.

46

Figure 22. Schematic picture of a battery with quinones as charge storage units. a) charge and b) discharge.

6.1.1 Testing a Battery A common way to test secondary batteries is by galvanostatic charging and discharging. In this method a constant current is applied that alternates in sign and the voltage of the cell is monitored. The cell voltage is plotted vs. time as in Figure 24c and d. The injection of current in a material will result in a rapid increase of the potential if no faradaic processes occur. This will transpire during the doping of the CP backbone since it displays a capacitive response in a CV, see Section 4.1. The potential change will decrease when the current have polarized the electrode to a potential where a faradaic pro-cess occurs. The potential will now be given by the Nernst equation instead. A plateau will be seen in the voltage vs. time graph when both materials in the battery have reached the potential where a faradaic response occurs. A long plateau represents a high charge storage capacity at a constant voltage output.

6.2 pH Battery Since the redox potential of quinones is changing with pH, a battery utilizing the same quinone on both the positive and negative electrode can be made with a concentration cell, i.e. different electrolytes in the two sides of the battery with a semipermeable membrane in between, Figure 23. The mem-brane used when the two electrolytes have different pH is an anion exchange membrane that only lets anions through and does not permit protons.

47

Figure 23. Schematic picture of the pH battery setup.

Figure 24. a) Glassy carbon electrodes with PEDOT≡HQ in pH 5 (black line) and pH 0 (grey line) before cycling, b) Capacity vs. cycle number, and c) Cell potential of the pH battery for the first three cycles. A plateau is seen at 0.3 V, corresponding to the potential difference of the two electrodes. d) Cell potential of the pH battery for the last three cycles. A plateau is seen around 0.23 V.

48

PEDOT≡HQ was electrochemically polymerized from a monomer solution onto two glassy carbon electrodes with five scans using the same conditions as in Figure 8.

The open cell battery in Figure 23 was tested for 168 charge and dis-charge cycles. The electrolyte for both sides of the battery was a sulfate buffer. The pH was set to 5 for the negative electrode side and to 0 for the positive electrode side, resulting in a theoretical voltage of 0.3 V for the battery, Figure 24a. This is consistent with Equation (4), where a 59 mV-difference is expected per pH unit. A small plateau is visible around 0.3 V in the cell voltage vs. time graph for the first three cycles, Figure 24c. This is consistent with the theoretical voltage value, and proves that the battery set-up works. The capacity is calculated from the time of one cycle and first decreases rapidly but then stabilizes after 30 cycles, Figure 24b. The last three cycles, Figure 24d, display a shorter plateau at a lower voltage com-pared to the first three cycles. The lower voltage is partially due to a de-crease in pH at the negative electrode (pH was about 3.5 after 168 cycles) as a result of leakage of protons through the membrane.

6.3 All-Organic Proton Battery This far, only aqueous electrolytes have been discussed. An AQ as PG was wanted in order to construct a battery with a higher voltage output, since the EWGs on the AQ is causing a dramatic decrease in its redox potential, Table 1. The redox potential for AQ is however close to the reduction potential of protons. This was solved by using an organic salt slurry, a concept built on the “water-in-salt” electrolyte78 where hydrogen reduction is suppressed by the high salt concentration. The quinone reaction requires a protic electrolyte to shuttle protons in and out of the material. Most organic solvents are non-protic so a protic salt is required. An all-organic battery based on PE-DOT/quinone CRPs has been made, Paper II, utilizing an electrolyte slurry containing a mixture of a protonated pyridine and a pyridine, which contrib-ute with both a proton donor and a proton acceptor. This allows for a re-versible proton coupled redox conversion of the quinone also in non-aqueous electrolytes.79 Pyridines can have different pKa values and the formal poten-tial of a quinone is shifted depending on the proton donor/acceptor strength in the electrolyte. The formal potential shows a pKa dependence similar to the pH dependence in aqueous electrolytes, although the pKa dependence has a different origin.

The two quinones used as PGs in the CRPs are an AQ on the negative side and a BQ on the positive side, Figure 25a. The AQ/BQ combination gives a theoretical battery voltage of 0.5 V, based on Equation (4). The ex-perimental battery voltage is seen in Figure 25c and is indeed 0.5 V.

49

Figure 25. a) Schematic of the all-organic battery. b) CVs of PEDOT-AQ and PEDOT-O-C=O-HQ. c) A charge and discharge curve from the battery test and d) a differential capacity plot made from the data in c). Reprinted with permission from Emanuelsson, R. et. al. J. Am. Chem. Soc. 2017, 139 (13), 4828-4834. Copyright (2017) American Chemical Society.

50

7. Concluding Remarks

The overall aim of this work was to understand the fundamental properties of CRPs with organic battery applications as a target. Organic batteries can provide safer batteries compared to batteries based on metals and can be used for new applications, such as in flexible electronics or in fabrics. The CRP approach provides a material with both high conductivity and high specific capacity. The rigidity of the backbone is of great importance since a twist in the conjugation could lower the conductivity. It was found that PE-DOT was not affected by the redox reaction of the PG as was previously seen for polypyrrole CRPs, which is an indication of a more stable back-bone. Other properties of the CRP system were also investigated by various in situ methods and compared to the properties of PEDOT:

Firstly, the redox reaction of quinone PG was studied with regards to sub-stitution effect, pH dependence, and kinetics. The choice of linker affected the redox potential of the quinone due to the different electronic properties of the part of the linker that was closest to the quinone. The potential of the quinone shifted with close to the characteristic -59 mV per pH and the qui-none showed a fast and reversible redox reaction in the whole pH range measured. The apparent rate constant was approximately 25 s-1 and the peak current at different scan rates provided evidence of a surface reaction with-out diffusion limitations. This means that the quinones are connected to the electrode by the conducting backbone and that the CRP concept works.

The mass changes during redox cycling displayed different responses dur-ing the doping of the backbone and during the redox reaction of the PG. An average mass change of 22 g per mol charge occurred during the backbone doping and an average mass change of 3 g per mol charge was seen during the redox reaction of the PG. The small mass change during the PG redox reaction is seen as proton cycling, without the accompanying of many water molecules.

The similar conductance value between pristine PEDOT and PEDOT≡HQ indicates that the number of charges and their mobility are equivalent. The conductivity onset on the other hand is at lower potentials for PEDOT, which could be caused by morphology differences between PEDOT and PEDOT≡HQ. The good conductivity for all polymers eliminates the need for conducting additives otherwise common in organic electrode materials. The in situ conductance measurements also provided evidence of redox matching between the chosen quinone and the PEDOT backbone.

51

The charge transport mechanism through the CRP material was measured with in situ conductance at different temperatures and was found to be most-ly consisting of a band transport, since it displayed a non-activated process at the measured temperatures. Some contribution of hopping with low activa-tion energy is also likely to occur. The apparent rate constant of the quinone redox reaction was also calculated at different temperatures and the rate lim-iting step was found to be an activated process with an activation energy of 0.3 eV. The rate limiting step can thus not be the charge transport through the polymer backbone since it had a non-activated process.

The spectral changes that occurred during redox cycling indicated that both polarons and bipolarons act as charge carriers in PEDOT and PE-DOT≡HQ. This was concluded since the number of polarons and their mo-bility decreased, while at the same time the conductivity stayed at a constant value. Overall the PG does not seem to affect the properties of the backbone in any major way.

The trimer solid was found to consist of two different regions where one had an increased packing and thus spectral transitions with lower energy than the more amorphous region.

The post-deposition polymerization method provides a way of polymeri-zation that increases the processability of CPs. The method is successful due to the use of trimers as polymer building blocks. The inherent conductivity of the trimers facilitates charge transport throughout the layer and allows the polymerization to reach the whole material. The exact mass loading can be calculated by knowing the concentration and amount of the deposited trimer layer, which allows for calculations of doping levels. The doping level of the polymer formed from EEE was calculated to 0.2 polarons per EDOT unit.

Lastly, two different all-organic battery types were demonstrated utilizing CRPs at both electrodes. The difference in redox potential for the two elec-trodes gave the battery voltage and was consistent with the theoretical calcu-lations for both battery types. The protic organic salt slurry used in one of the batteries was successful in hindering the proton reduction that would have been present in an electrolyte with a lower salt concentration.

52

8. Sammanfattning på svenska

Ett mer miljövänligt samhälle behövs för att säkerställa en acceptabel fram-tid för kommande generationer. Ett stort problem idag är användningen av fossila bränslen för energiproduktion som avger stora mängder koldioxid till atmosfären. För att säkerställa ett mer hållbart samhälle måste energi produ-ceras från förnybara energikällor, som vind- och solkraft. Nackdelen med dessa metoder är att de inte producerar energi kontinuerlig, eftersom vinden inte alltid viner och solen inte alltid skiner. Batterier kan agera som en buf-fert mellan energiproduktion och energianvändning genom att förvara ener-gin som produceras fram tills den behövs.

Kommersiella batterier innehåller metaller som inte kan återanvändas och kräver mycket resurser för att framställas. Mer hållbara energilagrings-material kan hittas med inspiration från naturen och kallas organiska materi-al vilket betyder att de består till mesta del av grundämnena väte, kol, syre, kväve och svavel, vilka alla finns tillgängligt på ytan av jorden. Batterier som baseras på dessa organiska material skulle även kunna användas till andra applikationer än vad vanliga litiumjonbatterier används till, såsom i böjbar elektronik och inbyggda i kläder.

De organiska materialen som passar till batteriapplikationer kan också fungera i andra system, som i solceller, katalysatorer och som färgföränd-rande material. Materialen som undersökts i denna avhandling är polymerer, d.v.s. en lång rad av likadana smådelar, monomerer, som sitter ihop kemiskt. Polymerer leder vanligtvis inte ström, plaster är t.ex. uppbyggda av polyme-rer, men kan göras ledande genom att man för in laddningar som kan röra sig fram och tillbaka genom polymeren. Ledningsförmågan är viktig för batteri-applikationer, eftersom elektroner måste kunna nå ut till hela materialet för att så mycket energi som möjligt ska sparas i batteriet. De ledande polyme-rerna tar inte upp särskilt många elektroner jämfört med sin vikt, d.v.s. de har dålig kapacitet. Det finns dock andra organiska material som är små och kan ta upp och avge flera elektroner. Dessa processer kallas reducering och oxidering och materialen kallas redoxmaterial. Dessa små grupper kan bin-das in med en länk till de ledande polymererna och på så sätt bilda ett material som både har hög ledningsförmåga och hög kapacitet, så kallade ledande redoxpolymerer, se Figur A. De redoxaktiva grupperna som har använts i denna avhandling kallas kinoner och de finns i olika former över-allt i naturen, i fotosyntesen hos växter och i andningssystemet hos djur. Kinonerna lagrar energi genom att reduceras och oxideras, samtidigt som de

53

Figur A. En schematisk bild av en ledande redoxpolymer som har undersökts i denna avhandling

skickar protoner in och ut ur materialet. Protoner är positivt laddade partiklar och motverkar de laddningar som de negativa elektronerna för med sig.

Huvudsyftet med avhandlingen var att ta reda på hur elektroner rör sig i materialen och hur redoxgrupperna och den ledande polymerbasen interage-rar med varandra. Det finns olika mekanismer för hur elektronerna rör sig genom ledande polymerer. För att ta reda på vilken mekanism som domine-rar i de undersökta materialen testades ledningsförmågan vid olika tempera-turer. Resultatet blev att ledningsförmågan blir lägre vid högre temperaturer; detta fenomen syns också i metaller där elektronerna rör sig nästintill obe-hindrat genom materialet. De ledande polymererna kan sägas ha metall-liknande ledningsförmåga. Orsaken till att ledningsförmågan blir sämre vid högre temperaturer är att atomerna i materialet då kan röra sig mer och stör elektronernas rörelse.

För ledande redoxpolymerer som tidigare undersökts har reaktioner som skett på redoxgruppen stört ledningsförmågan hos polymeren, men i denna avhandling har en stabilare polymerbas valts, och ledningsförmågan höll sig på en konstant nivå oavsett vad som skedde på redoxgruppen. Det visades också att redoxgrupperna, kinonerna, kan reagera snabbt med att ta upp och ge ifrån sig elektroner, vilket är en fördel för batterier och andra applikation-er. Protonerna, som motverkar de negativa laddningarna från elektronerna, rör sig in i materialet utan att ta med sig stora mängder vatten, något som annars är vanligt när joner (laddade partiklar) rör sig.

En metod för att tillverka polymerer har också undersökts. I denna metod används en bas av tre monomerer i rad, vilka läggs på önskat material och polymeriseras sedan direkt på materialet. I vanliga fall polymeriseras mo-nomerer från en lösning, men nackdelen med denna metod är att det är svårt att få den att fungera i större skala i en fabrik. Den nya metoden gör det lät-tare att tillverka många batterier på kort tid.

Till sist testades också två olika sorters batterier. Till den första användes samma ledande redoxpolymer till både plus- och minussidan i batteriet. Ki-nonernas redoxreaktion sker vid olika potentialer beroende på vilket pH som är i lösningen runtomkring. Den spänningsskillnad som behövs i ett batteri

54

uppstår genom att den ena sidan har lägre pH än den andra, d.v.s. är surare. Detta batteri visar att kinonernas kemi fungerar bra i de ledande redoxpoly-mererna. Hittills har alla experiment gjorts i vattenlösningar, men den andra sortens batteri använder organiska lösningsmedel (alkoholer och oljor är exempel på organiska lösningsmedel) med salter i som gör att kinonerna kan reagera ungefär som när de är i vatten. Fördelen med organiska lösningsme-del är att man kan ha högre spänning än i vattenbaserade batterier. Spänning i det organiska batteriet uppstår då två olika kinoner används, och dessa har olika potential där deras redoxreaktion sker.

Avhandlingen ökar kunskapen om hur ledande redoxpolymerer fungerar och banar väg för framtiden inom organisk elektronik.

55

9. Acknowledgement

First of all, I would like to deeply thank my main supervisor Prof. Martin Sjödin. I have really enjoyed working together with you and discovering the fundamentals of conducting polymers. I also want to thank you for your patience with all my questions during these years and that you always have your door open.

I would also like to thank my supervisor Dr. Rikard Emanuelsson for your guidance in the world of organic chemistry, even all the way back in my undergraduate course. And also for the times spent wondering about the crazy physics of it all.

I am very grateful to my supervisor Prof. Maria Strømme for your energy and positivity. I also want to thank my last supervisor Prof. Adolf Gogoll for your humor and assistance during my short time at BMC.

For all the people, past and present, at the Nano group: A big thank you for making work a fun place to be and or all the fun and interesting discus-sions during lunches and fika. Maria, Simon, and Christian, thank you for 10 good years together at Ångan with many discussion of science and much partying. Petter, thank you for being an awesome first roommate and helping me with all things a new PhD has troubles with, and also for taking part in various pranks that can light up a grey day. Thank you Rui for being a good present roommate who always have a gum ready when I need it. Teresa, thank you for entertaining discussions during lunches and fika (and some-times in between as well), and for joining me for workouts at Campus. Alex, thank you for inviting us for some awesome AWs at your place. Thank you to the seniors Natalia, Ocean, Ken, Viviana, Chao, Albert, Jonas, and Daniel for all things lab and office.

A special thank you to the people in the battery group: Huan, for discus-sions about interesting experiments and your jokes in the lab and during conferences. Hao, I really enjoyed traveling with you to GRC in Italy with much laughter and very excellent chocolate. I also appreciate your effort together with Maria of trying to find me a wine I like. A very big thank you to Lisa who has been with me almost from the start in both ups and downs. I am very happy that you decided to move to Uppsala and I had a lot of fun during all our conferences together and at your place. A big thanks to the former members of the battery group, Li, Xiao, and Christoffer for great discussion of literature, and to all the welcoming people at BMC: Magnus, Sandra, Michael, and Anna. I also want to thank all the undergraduate and

56

graduate students that have been involved in the battery project: Robin, Edvard, Andrew, Taku, Philipp, Oka, and Felicia, for forcing me to under-stand things on a higher level.

Kristina, I don´t know how I would have managed the first years at Ångan without you. Thank you for being my number one lab partner and for nice walks in the sun during our PhD days.

Tack alla vänner och släktingar för ert stöd och intresse under mina år som doktorand. Ett speciellt tack till Sofie, Malin och Ac för att ni har varit med mig genom allt ända sen lågstadiet och tack till alla barn i mitt liv som gör världen lite ljusare. Ellen, tack för all den trygghet du gett mig genom åren, tex när man följer staket i Tyskland, och för att du helt enkelt är den bästa storasyster man kan få. Tack till mina underbara systerdöttrar Tilda och Freja för att ni har varit mina glädjeämnen under era snart 4 år på denna jord. Ni har funnits i mina tankar ända sen jag började som doktorand, då ni låg i er mammas mage.

Det största tacket går till mamma Carina och pappa Bengt för ert stora stöd i stort och smått under alla mina 30 år. Tack för att ni har uppmuntrat mig att satsa på det jag tycker om och förståelse för alla mina egenheter.

Till sist, tack till min gammelmorfar Edvin “Pang” Malmsjö som gav mig ett fint citat att ha i början på avhandlingen.

57

10. References

1. Chen, H.; Armand, M.; Demailly, G.; Dolhem, F.; Poizot, P.; Tarascon, J.-M., ChemSusChem 2008, 1 (4), 348-355.

2. Larcher, D.; Tarascon, J. M., Nat. Chem. 2014, 7, 19. 3. Huang, H.; Karlsson, C.; Strømme, M.; Gogoll, A.; Sjödin, M., PCCP 2017, 19

(16), 10427-10435. 4. Xie, J.; Zhang, Q., J. Mater. Chem. A 2016, 4 (19), 7091-7106. 5. Han, X.; Chang, C.; Yuan, L.; Sun, T.; Sun, J., Adv. Mater. 2007, 19 (12). 6. Reddy, A. L. M.; Nagarajan, S.; Chumyim, P.; Gowda, S. R.; Pradhan, P.;

Jadhav, S. R.; Dubey, M.; John, G.; Ajayan, P. M., Sci. Rep. 2012, 2, 960. 7. Yang, B.; Hoober-Burkhardt, L.; Wang, F.; Surya Prakash, G. K.; Narayanan, S.

R., J. Electrochem. Soc. 2014, 161 (9), A1371-A1380. 8. Poizot, P.; Dolhem, F.; Gaubicher, J., Curr. Opin. Electrochem. 2018, 9, 70-80. 9. Emanuelsson, R.; Sterby, M.; Strømme, M.; Sjödin, M., J. Am. Chem. Soc.

2017, 139 (13), 4828-4834. 10. Liang, Y.; Tao, Z.; Chen, J., Adv. Energy Mater. 2012, 2 (7), 742-769. 11. Nakahara, K.; Iwasa, S.; Satoh, M.; Morioka, Y.; Iriyama, J.; Suguro, M.;

Hasegawa, E., Chem. Phys. Lett. 2002, 359 (Issues 5–6), 351–354. 12. Koshika, K.; Sano, N.; Oyaizu, K.; Nishide, H., Chem. Commun. 2009, (7),

836-838. 13. Nishide, H.; Oyaizu, K., Science 2008, 319 (5864), 737-738. 14. Song, H. K.; Palmore, G. T. R., Adv. Mater. 2006, 18 (13), 1764-1768. 15. Xuan, Y.; Sandberg, M.; Berggren, M.; Crispin, X., Org. Electron. 2012, 13 (4),

632–637. 16. Zhang, K.; Guo, C.; Zhao, Q.; Niu, Z.; Chen, J., Adv. Sci. 2015, 2 (5). 17. Geng, J.; Bonnet, J.-P.; Renault, S.; Dolhem, F.; Poizot, P., Energy Environ.

Sci. 2010, 3, 1929-1933. 18. Abruña, H. D.; Kiya, Y.; Henderson, J. C., Phys. Today 2008, 61 (12), 43-47. 19. Visco, S. J.; DeJonghe, L. C., J. Electrochem. Soc. 1988, 135 (12), 2905-2909. 20. Sotomura, T.; Uemachi, H.; Takeyama, K.; Naoi, K.; Oyama, N., Electrochim.

Acta 1992, 37 (10), 1851-1854. 21. Oyaizu, K.; Nishide, H., Adv. Mater. 2009, 21 (22), 2339-2344. 22. Wei, X.; Pan, W.; Duan, W.; Hollas, A.; Yang, Z.; Li, B.; Nie, Z.; Liu, J.; Reed,

D.; Wang, W.; Sprenkle, V., ACS Energy Letters 2017, 2 (9), 2187-2204. 23. Aydın, M.; Esat, B.; Kılıç, Ç.; Köse, M. E.; Ata, A.; Yılmaz, F., Eur. Polym. J.

2011, 47 (12), 2283-2294. 24. Armand, M.; Grugeon, S.; Vezin, H.; Laruelle, S.; Ribière, P.; Poizot, P.;

Tarascon, J. M., Nat. Mater. 2009, 8, 120. 25. Häupler, B.; Wild, A.; Schubert, U. S., Adv. Energy Mater. 2015, 5 (11),

1402034. 26. Bard, A. J.; Faulkner, L. R., Electrochemical Methods: Fundamentals and

Applications. 2nd ed.; John Wiley & Sons: New York, NY, 2001; p 864. 27. Song, Z.; Zhou, H., Energy Environ. Sci. 2013, 6, 2280-2301.

58

28. Dyer, A. L.; Thompson, E. J.; Reynolds, J. R., ACS Appl. Mater. Interfaces 2011, 3 (6), 1787-1795.

29. Xia, L.; Wei, Z.; Wan, M., J. Colloid Interface Sci. 2010, 341 (1), 1-11. 30. Sadik, O. A., Electroanal. 1999, 11 (12), 839-844. 31. Sista, P.; Wilson, M.; Holmes, N.; Kularatne, R. S.; Rainbolt, E. A.; Biewer, M.

C.; Dastoor, P. C.; Belcher, W.; Stefan, M. C., Sci. Adv. Mater. 2013, 5 (5), 512-518.

32. Shirakawa, H.; Louis, E. J.; MacDiarmid, A. G.; Chiang, C. K.; Heeger, A. J., J. Chem. Soc., Chem. Commun. 1977, (16), 578-580.

33. Heeger, A. J., Angew. Chem. Int. Ed. 2001, 40 (14), 2591-2611. 34. Geoghegan, M.; Hadziioannou, G., Polymer Electronics. Oxford University

Press: Oxford, 2013; Vol. 22. 35. Tolbert, L. M., Acc. Chem. Res. 1992, 25 (12), 561-568. 36. Gospodinova, N.; Terlemezyan, L., Prog. Polym. Sci. 1998, 23 (8), 1443-1484. 37. Bubnova, O.; Khan, Z. U.; Malti, A.; Braun, S.; Fahlman, M.; Berggren, M.;

Crispin, X., Nat. Mater. 2011, 10, 429. 38. De Jesus, M. C.; Fu, Y.; Weiss, R. A., Polym. Eng. Sci. 1997, 37 (12), 1936-

1943. 39. Groenendaal, L.; Zotti, G.; Aubert, P.-H.; Waybright, S. M.; Reynolds, J. R.,

Adv. Mater. 2003, 15 (11), 855-879. 40. Zade, S. S.; Bendikov, M., Org. Lett. 2006, 8 (23), 5243-5246. 41. Jonas, F.; Schrader, L., Synth. Met. 1991, 41 (3), 831-836. 42. Sun, K.; Zhang, S.; Li, P.; Xia, Y.; Zhang, X.; Du, D.; Isikgor, F. H.; Ouyang,

J., J. Mater. Sci. - Mater. Electron. 2015, 26 (7), 4438-4462. 43. Martin, D. C.; Wu, J.; Shaw, C. M.; King, Z.; Spanninga, S. A.; Richardson-

Burns, S.; Hendricks, J.; Yang, J., Polym. Rev. 2010, 50 (3), 340-384. 44. Groenendaal, L.; Jonas, F.; Freitag, D.; Pielartzik, H.; Reynolds, J. R., Adv.

Mater. 2000, 12 (7), 481-494. 45. Zotti, G.; Zecchin, S.; Schiavon, G.; Louwet, F.; Groenendaal, L.; Crispin, X.;

Osikowicz, W.; Salaneck, W.; Fahlman, M., Macromolecules 2003, 36 (9), 3337-3344.

46. Crispin, X.; Marciniak, S.; Osikowicz, W.; Zotti, G.; van der Gon, A. W. D.; Louwet, F.; Fahlman, M.; Groenendaal, L.; De Schryver, F.; Salaneck, W. R., J. Polym. Sci., Part B: Polym. Phys. 2003, 41 (21), 2561-2583.

47. Hong, W.; Xu, Y.; Lu, G.; Li, C.; Shi, G., Electrochem. Commun. 2008, 10 (10), 1555-1558.

48. Kurias, K. M.; Jasna, M.; Menon, M. R. R.; Antony, A.; Jayaraj, M. K., AIP Conf. Proc. 2019, 2082 (1), 050008.

49. Cuentas Gallegos, A. K.; Rincón, M. E., J. Power Sources 2006, 162 (1), 743-747.

50. Lei, C.; Wilson, P.; Lekakou, C., J. Power Sources 2011, 196 (18), 7823-7827. 51. Luo, J.; Billep, D.; Waechtler, T.; Otto, T.; Toader, M.; Gordan, O.; Sheremet,

E.; Martin, J.; Hietschold, M.; Zahn, D. R. T.; Gessner, T., J. Mater. Chem. A 2013, 1 (26), 7576-7583.

52. Kim, G. H.; Shao, L.; Zhang, K.; Pipe, K. P., Nat. Mater. 2013, 12, 719. 53. Baughman, R. H.; Bredas, J. L.; Chance, R. R.; Elsenbaumer, R. L.; Shacklette,

L. W., Chem. Rev. 1982, 82 (2), 209-222. 54. Higgins, S. J., Chem. Soc. Rev. 1997, 26 (4), 247-257. 55. Pietschnig, R., Chem. Soc. Rev. 2016, 45 (19), 5216-5231. 56. Yang, L.; Huang, X.; Mamedov, F.; Zhang, P.; Gogoll, A.; Strømme, M.;

Sjödin, M., PCCP 2017, 19 (36), 25052-25058.

59

57. Åkerlund, L.; Emanuelsson, R.; Renault, S.; Huang, H.; Brandell, D.; Strømme, M.; Sjödin, M., Adv. Energy Mater. 2017, 7 (20), 1700259.

58. Karlsson, C.; Huang, H.; Strømme, M.; Gogoll, A.; Sjödin, M., Electrochim. Acta 2015, 179, 336–342.

59. Oyaizu, K.; Choi, W.; Nishide, H., Polym. Adv. Technol. 2011, 22 (8), 1242-1247.

60. Salamone, M. M.; Sassi, M.; Beverina, L.; Mari, C. M.; Ruffo, R., Electrochim. Acta 2014, 140, 152-159.

61. Peters, G. M.; Tovar, J. D., J. Am. Chem. Soc. 2019, 141 (7), 3146-3152. 62. Wu, Y.; Zeng, R.; Nan, J.; Shu, D.; Qiu, Y.; Chou, S.-L., Adv. Energy Mater.

2017, 7 (24), 1700278. 63. Pan, F.; Wang, Q., Molecules 2015, 20 (11), 20499-20517. 64. Miroshnikov, M.; Divya, K. P.; Babu, G.; Meiyazhagan, A.; Reddy Arava, L.

M.; Ajayan, P. M.; John, G., J. Mater. Chem. A 2016, 4 (32), 12370-12386. 65. Feher, G.; Allen, J. P.; Okamura, M. Y.; Rees, D. C., Nature 1989, 339 (6220),

111-116. 66. Fischer, W. W.; Hemp, J.; Johnson, J. E., Annu. Rev. Earth Planet. Sci. 2016, 44

(1), 647-683. 67. Martin, W. F.; Bryant, D. A.; Beatty, J. T., FEMS Microbiol. Rev. 2017, 42 (2),

205-231. 68. Søballe, B.; Poole, R. K., Microbiology 1999, 145 (8), 1817-1830. 69. B Chance; W D Bonner, J., and; Storey, B. T., Annu. Rev. Plant Physiol. 1968,

19 (1), 295-320. 70. Huynh, M. T.; Anson, C. W.; Cavell, A. C.; Stahl, S. S.; Hammes-Schiffer, S.,

J. Am. Chem. Soc. 2016, 138 (49), 15903-15910. 71. Hansch, C.; Leo, A.; Taft, R. W., Chem. Rev. 1991, 91 (2), 165-195. 72. Ispas, A.; Bund, A., Electrochemical Quartz Crystal Microbalance. In

Encyclopedia of Applied Electrochemistry, Kreysa, G.; Ota, K.-i.; Savinell, R. F., Eds. Springer New York: New York, NY, 2014; pp 554-568.

73. Sauerbrey, G., Z. Phys. 1959, 155 (2), 206-222. 74. Cutnell, J. D.; Johnson, K. W.; Young, D.; Stadler, S., Physics. 2015. 75. Wedege, K.; Dražević, E.; Konya, D.; Bentien, A., Sci. Rep. 2016, 6, 39101. 76. Cukierman, S., Biochim. Biophys. Acta 2006, 1757 (8), 876-885. 77. Schärli, M.; Kiess, H.; Harbeke, G.; Berlinger, W.; Blazey, K. W.; Müller, K.

A., Synth. Met. 1988, 22 (4), 317-336. 78. Suo, L.; Borodin, O.; Gao, T.; Olguin, M.; Ho, J.; Fan, X.; Luo, C.; Wang, C.;

Xu, K., Science 2015, 350 (6263), 938-43. 79. Emanuelsson, R.; Huang, H.; Gogoll, A.; Strømme, M.; Sjödin, M., J. Phys.

Chem. C 2016, 120 (38), 21178-21183.

Acta Universitatis UpsaliensisDigital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology 1819

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-383026

ACTAUNIVERSITATIS

UPSALIENSISUPPSALA

2019