Influences of Growth Conditions and Porosity on Polypyrrole for Supercapacitor Electrode Performance by Joanna Wing Yu Lam B.A.Sc, University of British Columbia, 2007 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF APPLIED SCIENCE in THE FACULTY OF GRADUATE STUDIES (Electrical and Computer Engineering) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) April 2011 © Joanna Wing Yu Lam, 2011 Abstract Supercapacitors are electrical energy storage devices that offer high power density and high energy density. The current energy density of supercapacitors is, however, not sufficient to meet the requirements of many applications. By using polypyrrole (PPy) as an alternate electrode material, the energy density of supercapacitors can be increased. Approaches for simplifying and speeding the production of PPy electrodes are investigated, as are means of increasing power density. The tradeoffs in performance of PPy are investigated when electrochemical deposition conditions – current density and temperature -- are modified to reduce costs. Although the surface morphology changes according to deposition conditions, PPy’s performance in capacitance and charging time is not greatly affected. The best electrode performance is obtained using electrodeposition conditions in which a current density of 0.125 mA/cm2 is used and the temperature is held at -30°C. Higher temperatures and faster deposition rates can lead to more voluminous films which are lower in density, volumetric and specific capacitances. Further work is needed to investigate the impact of growth conditions on cycle and shelf life. To decrease the charging time of PPy the hypothesis is that additional porosity will help by creating channels of high ionic mobility. The porosity is achieved by polymerizing onto carbonized polyacrylonitrile nanofibres (NF). PPy-coated NF samples with a density of 1.2 g/cm3 exhibit similar volumetric (160 F/cm3) and specific capacitances (130 F/g) similar to that of pure PPy. The use of NF can increase the apparent ionic conductivities of PPy, allowing NF/PPy samples to charge just as quickly as pure polypyrrole electrodes that are four times less capacitive. However, based on the current model, the advantages of increasing porosity should be more dramatic, suggesting that other mechanisms such as uncompensated resistances and ion depletion may also influence charging time. As such, further work on NF/PPy is needed to determine and hopefully to mitigate the effects of such mechanisms. ii Table of Contents Abstract ......................................................................................................................... ii Table of Contents .......................................................................................................... iii List of Tables ..................................................................................................................v List of Figures ............................................................................................................... vi List of Abbreviations................................................................................................... viii Acknowledgements ....................................................................................................... ix 1 2 Introduction .............................................................................................................1 1.1 Supercapacitors .................................................................................................1 1.2 Operation of a Supercapacitor ...........................................................................2 1.3 Applications of Supercapacitors ........................................................................6 1.4 Motivation and Aim ..........................................................................................7 1.5 Thesis Organization ..........................................................................................9 Polypyrrole from Different Growth Conditions ...................................................... 10 2.1 Background..................................................................................................... 10 2.2 Modelling ....................................................................................................... 12 2.3 Experimental Setup ......................................................................................... 16 2.3.1 Deposition Setup ...................................................................................... 16 2.3.2 Sample Test Setup ................................................................................... 18 2.4 2.4.1 Data Analysis ........................................................................................... 19 2.4.2 Results and Discussion ............................................................................. 23 2.5 3 Results and Data Analysis ............................................................................... 19 Summary ........................................................................................................ 38 Polypyrrole on Nanofibres ..................................................................................... 39 3.1 Background..................................................................................................... 39 3.2 Experimental Setup ......................................................................................... 42 3.2.1 Deposition ............................................................................................... 42 3.2.2 Sample Test Setup ................................................................................... 44 3.3 Results and Data Analysis ............................................................................... 45 3.3.1 Data Adjustment ...................................................................................... 45 3.3.2 Results and Discussion ............................................................................. 46 iii 3.4 4 Summary ........................................................................................................ 60 Conclusion ............................................................................................................. 62 4.1 Research Contributions ................................................................................... 62 4.2 Future Work.................................................................................................... 63 Bibliography ................................................................................................................. 64 Appendix A Cyclic Voltammetry Measurements Without Filtering .............................. 67 Appendix B MATLAB Code........................................................................................ 68 iv List of Tables Table 1 Activated carbon characteristics [5]. ..................................................................3 Table 2 Deposition conditions for samples 1 – 6........................................................... 18 Table 3 Area, thickness and mass of PPy prepared. ...................................................... 23 Table 4 Capacitance, ionic conductivity, and time constant from CV at 1 mV/s. ........... 31 Table 5 Capacitance, ionic conductivity, and time constant from CV at 10 mV/s. ......... 31 Table 6 Approximation of uncompensated resistances in CV tests on Samples 1 – 6. ... 37 Table 7 Average thicknesses of NF/PPy Samples 4h – 16h. .......................................... 49 Table 8 Volume and mass of polypyrrole-coated nanofibre samples. ............................ 51 Table 9 Volumetric and specific capacitances of NF/PPy samples. ............................... 54 Table 10 Estimated ionic conductivity and time constant results of NF/PPy samples from CV at 1mV/s. .............................................................................................. 56 Table 11 Approximation of uncompensated resistances in CV tests on Samples 4h – 16h. ................................................................................................................... 59 v List of Figures Figure 1 Diagram of the cross-section of a supercapacitor showing the placement of current collectors, electrodes, electrolyte and separator. An asymmetrical device is when the two electrodes use different materials. .............................2 Figure 2 Electrochemical doping and dedoping of polypyrrole [17]. ............................. 11 Figure 3 Equivalent circuit of porous electrode............................................................. 12 Figure 4 Simplified equivalent circuit of porous electrode. ........................................... 13 Figure 5 Polypyrrole deposition setup using glassy carbon. Picture to the right is a top view of the setup. ........................................................................................ 17 Figure 6 Polypyrrole on glassy carbon. ......................................................................... 18 Figure 7 Three-electrode test setup on PPy. .................................................................. 19 Figure 8 Sample 1 CV at 1mV/s showing data before and after filtering. ...................... 20 Figure 9 CV at 1 mV/s from -0.8 V to +0.6 V vs. Ag/AgNO3 of Sample 1. ................... 21 Figure 10 CV at of 10 mV/s from -0.8 V to +0.6 V vs. Ag/AgNO3 of Sample 1. ........... 21 Figure 11 Relationship between density and thickness for PPy films from different growth conditions. By comparing Sample 1 to Sample 6, an increase in thickness results in a lower apparent density. .............................................. 24 Figure 12 SEM images of PPy grown at -30°C and 25°C ............................................. 26 Figure 13 CV at 1 mV/s from -0.8 V to +0.6 V vs. Ag/AgNO3. .................................... 28 Figure 14 CV at 10 mV/s from -0.8 V to +0.6 V vs. Ag/AgNO3. .................................. 29 Figure 15 Capacitances from 1 mV/s and 10 mV/s CV. ................................................. 32 Figure 16 Degradation of Sample 1 results in a decrease in charging/discharging capacitance, a longer rise time at reduced potentials, as well as a larger tail of current at -0.8V vs. Ag/AgNO3. .................................................................. 33 Figure 17 Estimated ionic conductivity of PPy estimated by fitting a model to cyclic voltammetry results. .................................................................................... 34 Figure 18 Time constant values of PPy from CV at 1 mV/s. ......................................... 36 Figure 19 Deposition setup using nanofibres. ............................................................... 42 Figure 20 Sample 4h, 12h and 16h dimensions from left to right (in cm). ..................... 43 Figure 21 Three-electrode test setup for NF/PPy. ......................................................... 44 Figure 22 NF/PPy after electrodeposition. PPy is clearly visible on NF. ...................... 46 Figure 23 SEM images of NF before deposition. .......................................................... 47 Figure 24 SEM images of Sample 12h after deposition. ................................................ 48 Figure 25 Thickness distribution of Sample 12h coated with PPy (in cm unless indicated). ................................................................................................... 50 Figure 26 Thickness distribution of Sample 16h coated with PPy (in cm unless indicated). ................................................................................................... 50 Figure 27 Density versus PPy mass on NF/PPy. ........................................................... 51 vi Figure 28 Figure 29 Figure 30 Figure 31 CV at 1 mV/s between -0.8 V to +0.4 V vs. Ag/AgNO3 on Sample 4h.......... 52 CV at 1 mV/s between -0.8 V to +0.6 V vs. Ag/AgNO3 on Sample 12h. ....... 53 CV at 1 mV/s between -0.8 V to +0.4 V vs. Ag/AgNO3 on Sample 16h. ....... 53 Change in the CV with time seen in Sample 4h. The sample becomes more resistive as it is cycled. Beginning represents the 10th cycle, and end represents the 30th cycle. ............................................................................. 55 Figure 32 Ionic conductivity of NF/PPy versus density from 1 mV/s CV...................... 57 Figure 33 Time constant versus density from 1 mV/s CV. ............................................ 58 vii List of Abbreviations AC Activated carbon ACN Acetonitrile CE Counter electrode CV Cyclic voltammetry EC Electrochemical capacitor F Farad GPS Global positioning system IL Ionic liquid LED Light emitting diode NF Carbonized nanofibre PAN Polyacrylonitrile PC Propylene carbonate PPy Polypyrrole RE Reference electrode SEM Scanning electron microscope SMT Surface mount technology TBAP Tetrabutylammonium hexafluorophosphate TEAB Tetraethylammonium hexafluoroborate TEAP Tetraethylammonium hexafluorophosphate WE Working electrode viii Acknowledgements First of all, I thank God for His many blessings. I thank my supervisor Dr. John Madden, for his guidance, consideration, and patience throughout my time in and out of the lab. I thank all of my past/present labmates, especially Eddie for his help, Nicole for supplying nanofibres, and Ali I. for introducing me to this research project. I reminisce about the lunch breaks at the Kaiser Atrium with Ali B., Andre and Diego, and I feel very fortunate to have made such good friends. I also thank Danny, for listening, for criticizing, but most importantly, for motivating. Finally, I thank my parents and my brother Paul for their constant reminder to complete my degree, as well as their unconditional love, respect and support. ix 1 Introduction 1.1 Supercapacitors Many electronic applications require a form of energy storage capability to properly function. Although the most cost-effective electrical energy storage technology is lead-acid batteries, applications such as portable electronics and vehicles require energy storage technologies that can also supply high power, an atypical characteristic of lead-acid batteries. Supercapacitors, also known as ultracapacitors or electrochemical capacitors (ECs), are a developing technology which combines a relatively high energy density and high power density. The supercapacitor’s energy density is significantly larger than traditional capacitors and bridges the gap between capacitors and batteries [1]. Traditional capacitors rely on charge-separation, where the movement of charges cause little damage to the capacitor, thus allowing capacitors to have high cycle life. Capacitors are able to discharge energy in a short period of time, but only a very small amount of energy can be stored. On the other hand, batteries store electrical energy through chemical reactions. The rate of reaction in batteries is much slower than the charge separation process in capacitors although batteries are able to store larger amounts of electrical energy. Also, batteries are more susceptible to effects of aging due to processes or mechanisms such as sulfation, loss of electrolyte, corrosion and so forth. To date, batteries offer more than one order of magnitude higher energy density than supercapacitors. Since the energy density offered by commercial supercapacitors cannot meet the energy storage requirement of many applications, it is unlikely that supercapacitors can entirely replace batteries. However, its ability to bridge batteries and capacitors has allowed supercapacitors to find niche applications. Applications of supercapacitors are discussed in the next section. Some manufacturers of the commercially-available supercapacitors are Panasonic, Maxwell and Cap-XX. Supercapacitors are available in different forms of packaging tailored to application requirements. For example, Cap-XX offers supercaps 1 in flexible packaging, whereas Panasonic’s “Gold Capacitors” are available in SMT, suited for PCB [1,2]. Unfortunately the high cost of supercapacitors is a barrier to wide adoption for this storage technology [3]. Typical cost parameters include $/Wh, $/kW and $/kg. Like traditional capacitors, supercapacitors consist of the following components: Electrodes, current collectors, electrolyte, and packaging[4]. 1.2 Operation of a Supercapacitor Similar to the operation of a traditional capacitor, applying a voltage to a supercapacitor causes anions in the electrolyte are attracted to the positive electrode whereas cations are attracted to the negative electrode. This interface between the charges in the electrodes and the respective ions from the electrolyte form two electrochemical double-layers, one at each of the electrodes. Current Collector Current Collector Active Material Separator Active Material Figure 1 Diagram of the cross-section of a supercapacitor showing the placement of current collectors, electrodes, electrolyte and separator. An asymmetrical device is when the two electrodes use different materials. If C1 and C2 are capacitors formed at the positive and negative electrodes, then the overall capacitance (CT) of the device is: 2 1 1 1 CT C1 C2 . (1) With CT, both the specific capacitance and volumetric capacitance can be found by normalizing to the mass or volume, respectively. The total amount of energy able to be stored by a device is then found by: 1 𝑈 = 2 𝐶𝑇 𝑉 2 . (2) Energy density in units of Wh/g or Wh/cm3 can be found by normalizing the total energy by either mass or volume. Another parameter that is of interest is the power density in W/g, or W/cm3. Commercial supercapacitors are carbon-based and symmetrical, meaning both the anode and cathode are based on carbon, thus C1 is essentially equal to C2. A typical carbon electrode is made with activated carbon powders – such as ones listed in Table 1 – mixed with a binder, and pressed onto an aluminum backing which acts as a current collector. Carbon electrodes are cost efficient but the capacitance of 100F/g, and only 60F/g (if including binder), is low compared to some other electrode materials. Supplier Capacitance Surface Area Maxsorb 110 F/g 2500 m2/g Opti 100 F/g 1600 m2/g Table 1 Activated carbon characteristics [5]. A type of electrode material which is higher in specific capacitance and electronic conductivity is based on conducting polymers like polypyrrole (PPy) and polythiophene. Largest relatively high capacitance value for conducting polymers is 220 F/g [6]. The charged state of a conducting polymer is when it is oxidized, and the discharged state is when it is reduced. Conducting polymers are however limited in charge/discharge speeds by their ionic conductivity – when the polymer charges and discharges, ions from the electrolyte migrate in and out of the polymer’s matrix. More on polypyrrole is discussed later. 3 All types of electrode materials are mounted onto current collectors like aluminum foil for consistent and reliable contact. Because supercapacitors are primarily power devices, all sources of resistance within the cell need to be minimized. If not properly designed, the contact resistance between the collector and the electrode material can limit the power of a device. There has been research in nanostructured current collectors which offer an increased contact area to minimize this resistance. Aside from electrode materials which can dictate capacitance of a supercapacitor, the electrolyte used can also influence the capacitance and operating voltage range of a supercapacitor. There are two main classes of electrolytes: aqueous and organic. Using aqueous electrolytes can provide high capacitance and power density, but at the expense of a lower voltage range of 0.6-0.8 V [7]. Aqueous electrolytes allow a lower cell voltage because the electrolyte voltage window is limited by water decomposition. Two common organic solvents are acetonitrile (ACN) and propylene carbonate (PC). Current activated carbon supercapacitors using PC containing tetraethylammonium tetrafluoroborate ions, offer a cell voltage of 2.5-3V. PC is higher in viscosity and has also a higher boiling point compared to ACN. The conductivity and density of 1 M TEAB/PC are 1.2 S/m and 1.191 g/cm3, while 1 M TEAB/ACN are 4.8 S/m and 0.861 g/cm3 [8]. Activated carbon in organic electrolyte is in the range of 50150 F/g, whereas in aqueous electrolyte the specific capacitance of 150-250 F/g is can be obtained [7]. Organic solvents however cannot withstand elevated temperatures since the stability decreases with increasing temperatures [8]. Another type of electrolyte employs ionic liquids (ILs), which are room temperature molten salts composed of only cations and anions. They can achieve even higher specific voltages than organic solvents, but the trade-off in using ionic liquids is their slow response time since the viscosity of ionic liquids is much higher than organic solvents [9]. Known for their wide electrochemical stability window, chemical stability at high temperature and reasonable if not outstanding conductivity at room temperature, ILs have garnered more interest for activated carbon supercapacitors and polymer-based supercapacitors [10]. Although supercapacitors using ILs can achieve a cell voltage of 3.5V and higher, ILs are most expensive compared to other aqueous and organic 4 electrolytes, which can drive up the cost of supercapacitors further. In addition, the cyclability of supercapacitors using ILs has yet to be confirmed [11]. A supercapacitor also requires a separator, a film that physically separates the anode and cathode to prevent them from shorting. An ideal separator is electrically nonconductive but is high in ionic conductivity; for example, Gore’s (Delaware, USA) EDG3, has a high ionic conductivity of ~3·10-4 S/cm in 1M TEAP ACN, is 12 μm thick and 50% porous [12]. If a separator that is low in ionic conductivity is used, the high ionic resistance lowers the power density of the device. Ideal separators provide ionic conductivities similar to the bulk solution conductivity. 5 1.3 Applications of Supercapacitors One of the uses of supercapacitors is in vehicles. In 2006, supercapacitor busses were piloted in Shanghai. Quick charge buses were designed and developed by a joint venture between Sinautec Automobile Technologies and Shanghai Aowei Technology Development Company, and manufactured by Sunwin Bus in partnership with Volvo of Sweden. Supercapacitors used in these buses are provided by Shanghai Aowei and are made of activated carbon with an energy density of 6 Wh/kg [13]. To improve the energy density going forward, Sinautec is working with the Massachusetts Institute of Technology to use vertically aligned carbon nanotube structures in the supercapacitor to increase surface area for holding charge [14]. To recharge the supercapacitor banks, stored along the base of the bus, overhead charging stations are strategically located at various bus stops. A collector rises from the top of the bus to make contact with the overhead charging line. The typical charging time is approximately 30 seconds. A fully charged bus can cover a distance of approximately 5 to 9 km depending on load. Other than vehicular applications, which typically exploit the high power density of supercapacitors, a number of applications now use supercapacitors because the high cycle life is ideal for low-maintenance systems. As a baseline comparison to the following performance data of supercapacitors, battery equipped systems usually need to replace batteries every 5 years. Ciralight’s SunTrackerTwo is a solar-powered sunlight tracker which uses supercapacitors as energy storage and a GPS system to control the motor and mirror to optimize the amount of daylight into a building [15]. In addition, Skyfuel from Ontario offers SolarCaps, which are solar paving LED lights powered by supercapacitors [16]. The ratings of the supercapacitor used in this system are 2.3 V and 120 F, with an operational time of 12-20 hours and lasting 100,000 full charge/discharge cycles; the overall expected lifetime of SolarCaps is ~10 years as it is limited by the LED lifetime of 100,000 hours [16]. 6 1.4 Motivation and Aim The goal of this project is to increase the energy density of supercapacitors as current supercapacitors using activated carbon are unable to meet the energy storage requirements of certain applications despite the high power and cycle life offered. In an asymmetric configuration, a different material is used for each of the electrodes in the supercapacitor. In using different materials, the energy stored and delivered can be increased because of two reasons: an increase in capacitance of the active material, and an increase in the operational voltage. Khomenko has demonstrated that symmetric electrochemical capacitors in aqueous medium obtained an energy density of 3 Wh/kg, whereas asymmetric electrochemical capacitors achieved up to 10 Wh/kg [7]. An asymmetrical supercapacitor using polypyrrole, a p-dopable polymer, as its cathode, along with the anode using activated carbon (AC), can achieve a total voltage window up to 3 V, whereas typical carbon-carbon supercapacitors operate up to 2.7 V. Also, because the specific capacitance of PPy, which ranges from 90-480 F/g depending on the electrolyte and synthesis method, is significantly higher than activated carbon, the energy density of PPy/AC supercapacitors can be further increased. According to calculations, the energy density attainable by PPy/AC supercapacitors can reach up to 6 Wh/kg, almost a 70% increase over commercial supercapacitors. Though this high energy density seems attractive, there are problems to be rectified in order for polypyrrole to become a full substitute of carbon. As such, this thesis investigates possible solutions to two underlying challenges encountered when polypyrrole is substituted for carbon. The first problem addresses to the high cost of supercapacitors. It is projected that for supercapacitors to be widely adopted, the cost of supercapacitors needs to be below 0.5 cents per Farad (F) [3]. While activated carbon fabrication techniques are mature and optimized, activated carbon currently costs roughly 2-3 cents/F [3]. With PPy, there is limited knowledge with cost-savings with regards to manufacturing, thus the cost of polypyrrole is suspected to be also far higher than the target of 0.5 cents/F. 7 In the Molecular Mechatronics Lab, polypyrrole is synthesized by applying a low current to an electrochemical cell at a low temperature [17]. This slow process – estimated to produce 1 μm of PPy per hour – can drive up costs of manufacturing polypyrrole. If a higher current is applied during synthesis, a larger rate of polypyrrole deposition is possible, and production can be less costly. Hence the aim of the first portion of this thesis is to investigate the tradeoffs in performance of polypyrrole if electrochemical deposition conditions are modified to minimize the time of preparation and synthesis. Specific capacitance, volumetric capacitance, density, and ionic conductivity are investigated to suggest whether or not changing synthesis techniques can improve or worsen the performance of PPy fabricated using the existing method. The second problem focuses on the ionic conductivity and charging time of polypyrrole. It is known that as the thickness of polypyrrole increases, the ionic resistance increases. The increase in ionic resistance combined with the increase in capacitance of thicker layers leads to longer charging times (in proportion to thickness squared), lowering the power output. Although PPy can increase energy density, it comes at the cost of lower power density. Since a key advantage of supercapacitors is supposed to be their high power density, this characteristic should ideally be maintained to justify its usefulness as an energy storage technology. To maintain high ionic conductivity and low charging times, our strategy is to increase the porosity of polypyrrole. The hypothesis is that an increase in porosity will allow fast transport of ions into the polypyrrole through the liquid electrolyte filled pores (the ions then travel less distance through solid polypyrrole). In this thesis, the porosity of polypyrrole is increased by synthesizing PPy onto carbonized nanofibres (NF), a high surface area conductive electrode which acts as a scaffold for PPy to mount on. By coating carbonized nanofibres with PPy rather than a solid piece of PPy, the thickness of PPy is significantly reduced as PPy is distributed onto each fibre, requiring less time is taken for ions to migrate into the matrix of polypyrrole. Consequently shorter charging time and higher ionic conductivity while maintaining the high energy density offered by PPy can be achieved. Samples of carbonized nanofibres coated with different amounts of PPy – referred to as NF/PPy – are investigated to determine the benefits and drawbacks of the addition of NF. 8 1.5 Thesis Organization Chapter 2 describes the work behind polypyrrole synthesized under different conditions. The different depositions take place at +25°C and -30°C, where constant currents of 0.125 mA/cm2, 0.250 mA/cm2 and 0.500 mA/cm2 are applied in each temperature, synthesizing a total of six samples. For all six samples, SEM images of the surfaces are taken. Also modelling is applied to characterize the capacitance and charging/discharging behaviour of the samples. Samples are compared in terms of ionic conductivity, specific capacitance, and volumetric capacitance to suggest whether deposition conditions affect the performance of PPy. Chapter 3 explains the work on increasing the porosity and ionic conductivity of polypyrrole. By coating carbonized nanofibres with PPy, the effective thickness of polypyrrole on each nanofibre is drastically reduced, thereby increasing porosity and ionic conductivity and most importantly charging/discharging speeds. Three samples are investigated, each containing a different amount of polypyrrole. SEM images of the three samples are taken. In addition, modelling is applied to obtain the ionic conductivity and specific capacitance of each sample. Finally these samples of NF/PPy are compared to the performance of pure PPy, as obtained in Chapter 2, and an analysis is made to see whether experimental data is consistent with expectations. Chapter 4 concludes the thesis with a summary, list of contributions, recommendations and possible future work. 9 2 Polypyrrole from Different Growth Conditions The aim of this section is to identify the differences in performance of polypyrrole (PPy) electrodeposited with varying current densities of 0.125 mA/cm2, 0.250 mA/cm2, and 0.500 mA/cm2 at two temperatures, 25°C and -30°C. The analysis of samples, which includes determining densities, capacitances, ionic conductivities and rates of charging are based on SEM images, cyclic voltammetry (CV), and modelling. It is found that while depositions in higher temperatures may be more convenient, the condition which produced the highest performing PPy is 0.125 mA/cm2 at -30°C. 2.1 Background Polypyrrole, a conducting polymer, is an attractive electrode material for supercapacitors due to its high charge density relative to carbon. Aside from supercapacitors, PPy is also investigated for a range of other applications, such as actuators and polymer light emitting diodes. Some of the other advantages in using polypyrrole include its low cost, ease of production and slow degradation. To charge and oxidize polypyrrole, a positive potential is applied to the polymer, which results in the removal of electrons from the polymer. At the same time, anions from the electrolyte move into the polymer to balance this charge. By inducing a positively charged backbone in the polymer, it becomes p-doped and electronically conductive. In order to discharge and reduce polypyrrole, a negative potential is applied, adding electrons back into the polymer. Ions originally in the matrix exit the polymer, leaving a neutral and uncharged backbone. This mechanism is shown in Figure 2. 10 Figure 2 Electrochemical doping and dedoping of polypyrrole [17]. There are two ways to synthesize PPy: chemically and electrochemically. In chemical deposition, pyrrole monomers are added to a solution containing an oxidizing agent such as FeCl3, and small bits of PPy form in the solution [18]. Usually the bits of PPy formed are filtered from solution and prepared as solid electrodes by pressing. With electrochemical deposition, a constant voltage, voltage sweep or constant current signal is applied to the system and a layer of polypyrrole film deposits onto the working electrode. Yamaura, Smela, Kaneto and others all have proposed “recipes” to synthesize polypyrrole electrochemically [19]. Each deposition method has its own advantages and disadvantages. For example, the electrochemical method is limited in terms of the mass production and it tends to be slow. Although chemical deposition can yield higher amounts of polypyrrole in a shorter time frame, electrochemical deposition methods can produce more capacitive and conductive PPy samples; PPy from electrochemical deposition achieved up to 187 F/g, whereas PPy from chemical deposition achieved up to 70 F/g [9]. The Yamaura method in particular is slow, but leads to the highest electronic conductivities [20]. 11 2.2 Modelling The response of polypyrrole during oxidation and reduction can be described by the use of a finite transmission line equivalent circuit. In this equivalent circuit, the porous electrode is separated into a liquid (ionic) phase and solid (electronic) phase with resistive and capacitive elements, as shown in Figure 4. This analysis was previously shown in detail by Fok [21]. A brief description of the electrical components in the finite transmission line follows. Figure 3 Equivalent circuit of porous electrode. The electronic phase potential Ф1(ɣ,t) and ionic phase potential Ф2(ɣ,t) are both functions of the normalized thickness 𝛾 and time 𝑡. The normalized thickness is: 𝑥 𝛾 = 𝑙. (3) The actual thickness of a PPy sample is represented by l, where x varies from 0 to l, thus 𝛾 varies between 0 and 1. The electronic and ionic resistances are represented by Re and Ri, respectively. In addition, Rc is contact resistance to the electrode while Rs is solution resistance, which can be dictated by the electrolyte’s concentration and the distance between polypyrrole to the reference electrode. Finally C is capacitance per surface area. It is seen that the total capacitance is influenced by thickness; as thickness increases, so does the capacitance. 12 Although efforts are made to minimize Rc and Rs so that charging and discharging time constants are dominated by either Re and Ri, the contact and solution resistances can be large enough to affect the accuracy of modelling. As such, it is necessary to consider all resistances, although the model presented in this section assumes that Rc and Rs are minor. Contact and solution resistances approximations will be presented later in Section 2.4 Results and Discussion. Another assumption made in this thesis is that Ri is much larger relative to Re, or that the electronic conductivity 𝜎 is much smaller than ionic conductivity κ. The range of electronic conductivity of polypyrrole is 300 S/cm in the fully oxidized state and 0.5 S/cm in the reduced state [22]. In comparison, κ is typically in the range of 10 -6 S/cm depending on the dopant and electrolyte used. It is then reasonable to assume that the electronic conductivity is negligible to the behaviour of polypyrrole, and that to charging speeds of PPy is limited by ionic conductivity. The basis of the differential equation used in this equation is shown in Figure 4. Figure 4 Simplified equivalent circuit of porous electrode. Other assumptions are made with this model include: Ionic conductivity κ is significantly lower than electronic conductivity so the electronic conductivity can be neglected; The effect of ion concentration gradients is small compared to that of field gradients of ion transport; No electronic voltage drop along the length of the sample; No Faradaic reactions; No variation in capacitance with oxidation state. 13 With the symbol 𝜑 represents the voltage between the solid and liquid phase (Ф1 – Ф2), the response of a polypyrrole sample can be shown with the following differential equation: 𝜏 𝑑𝜑 𝑑𝑡 = 𝑑2𝜑 𝑑𝛾2 . (4) where 𝜏 = 𝑅𝑇 𝐶𝑇 = 𝑙 𝜅𝐴 𝐶𝑣𝑜𝑙 𝐴 𝑙 = 𝐶𝑣𝑜𝑙 𝑙 2 𝜅 . (5) The time constant τ, a function of the total resistance RT and total capacitance CT, represents the charging or discharging time in seconds. 𝐶𝑣𝑜𝑙 is the volumetric capacitance in units of F/cm3, and ionic conductivity is represented by κ, in units of S/cm. Also, l represents the thickness while A is the apparent area of the sample. Once RT is known, approximations of Rc and Rs are compared to the total resistance. In the case that Rc and Rs are less than 10% of the total resistance, the uncompensated resistances are considered negligible, and the charging time constant will be dominated by the ionic resistance. However, in the case that Rc and Rs represent a large fraction of RT, ionic conductivities quoted will be upper limits. Although in an ideal capacitor the time constant to charge and discharge is independent of voltage and direction of charge, in PPy the ionic conductivity can depend on the state of oxidation and thus charge and discharge rates can differ. Capacitance also changes with oxidation state, as will be shown in Section 2.4. In order to account for these change with voltage to first order, ionic conductivity and Cvol will be separated into two cases: κ+ and Cvol+ for when the sample is close to the fully a reduced state; and κand Cvol- for when the sample is oxidized and being reduced. Cyclic voltammetry (CV) is a technique often used to characterize electrochemical systems. In a cyclic voltammetry (CV) experiment, the input signal is a voltage ramp specified within a voltage range. Once the defined voltage is reached, the ramp will reverse in polarity. The input signal is represented by kt, where k is the scan rate in V/s. Using Fourier Series, solutions are found for the positive and negative scans. 14 In the positive scan, the current density jM (t) in A/cm2 predicted by the model in Figure 4 is: k 4 ( 2 m1) 2 t 8 e j M (t ) 1 2 2 L m 0 ( 2m 1) . 2 m (6) In the negative scan, k is negative. The measured current density hence is: k 4 ( 2 m1) 2 t 8 e j M (t ) 1 2 2 L m 0 ( 2m 1) . 2 m (7) At long times, jM (t) approaches κτk/L and -κτk/L, equivalent to kCT and - kCT. This follows the equation, 𝑖= 𝐶𝑑𝑣 𝑑𝑡 . (8) . (9) where 𝑘= 𝑑𝑣 𝑑𝑡 Hence we are able to estimate the total capacitance of a sample using slow scans, as determined by simply dividing the current by the ramp rate. 15 The model is applied to simulate the CV response of each sample in this thesis using the following steps: 1. Calculate the area and thickness of each sample. 2. Specify the voltage range and scan rate of each scan. 3. Generate the current response at each step in time. 4. Apply the simulated current vs. voltage curves to the data. 5. Fit the current at long times by adjusting Cvol separately for positive and negative scans. 6. Fit the discharge/charge rate by adjusting conductivity κ. 2.3 Experimental Setup 2.3.1 Deposition Setup The setup shown in the Figure 5 is used to synthesize polypyrrole. A glassy carbon disc, diameter 2.54 cm, from Sigradur (HTW Hochtemperatur-Werkstoffe GmbH, Germany) is first mechanically polished with alumina polish to ensure PPy synthesis takes place on a smooth surface. Following polishing, the carbon disc which acts as a working electrode is secured by a metal clamp to the bottom of the glass cell, with a mylar o-ring sandwiched in between the glass cell and glassy carbon disc to prevent electrolyte from leaking. 16 CE WE Figure 5 Polypyrrole deposition setup using glassy carbon. Picture to the right is a top view of the setup. After the bottom of the cell is assembled, approximately 20 mL of solution containing 0.05 M tetrabutylammonium hexafluorophosphate (TBAP, Sigma-Aldrich), 0.06 M distilled pyrrole (Sigma-Aldrich) and 1 vol% of deionized water in propylene carbonate (PC, Sigma-Aldrich) is poured in. An activated carbon electrode from Gore (Newark, Delaware, USA) acts as the counter electrode, running through a septum thus creating a closed system for PPy deposition. Finally a syringe needle connected to a nitrogen gas line is inserted through the arm of the glass cell to bubble the solution with nitrogen gas. Constant current polymerization is carried out at both room (+25°C) and low temperature (-30°C) environments. At each temperature, films are deposited using one of three deposition current densities: 0.125 mA/cm2, 0.250 mA/cm2, or 0.500 mA/cm2. The total amount of time for each deposition is adjusted so each deposition has a total of 450 C per cm2 of electrode area, such that the same amount of polypyrrole is deposited in each deposition condition. After deposition each sample is washed in acetonitrile and then dried in air for 24 hours. Later sections in this thesis refer to sample by number, where samples 1 through 6 were prepared with different current densities and at different 17 temperatures, as shown in Table 2. Figure 6 is a picture of resulting PPy on glassy carbon. Sample 1 Sample 2 Sample 3 Sample 4 Sample 5 Sample 6 Temperature Current Density Total Deposition 2 (°C) (mA/cm ) Time (h) -30 0.125 4 -30 0.250 2 -30 0.500 1 25 0.125 4 25 0.250 2 25 0.500 1 Table 2 Deposition conditions for samples 1 – 6. PPy Glassy Carbon Figure 6 Polypyrrole on glassy carbon. 2.3.2 Sample Test Setup Samples are tested in 0.2 M TEAP in anhydrous acetonitrile using the setup in Figure 7. The solution is first bubbled in nitrogen gas to remove any oxygen which can degrade polypyrrole. Placement of an Ag/AgNO3 reference electrode is approximately 5 mm away from the surface of the working electrode, keeping the solution resistance to a minimum without interfering the migration of ions into the PPy film. Activated carbon electrodes (2cm by 2 cm) as counter electrode are clamped on by an alligator clip with a lead that goes through a septum. A sealed environment is crucial to avoid any changes in the concentration of the electrolyte because acetonitrile is volatile. After electrochemical testing, thicknesses and masses of PPy samples are measured by first removing PPy films 18 from glassy carbon discs with a razor blade. The standalone films are also forwarded to the UBC Bioimaging Lab for SEM imaging. CE RE WE Figure 7 Three-electrode test setup on PPy. 2.4 Results and Data Analysis 2.4.1 Data Analysis For CV tests at 1 mV/s, a noticeable amount of noise is seen because the signal is relatively small. One reason for ripples in the data shown in Figure 8 is suspected to be the AC grid, as the potentiostat is powered at 120 V and 60 Hz. Another reason could be the digital input of the potentiostat. To remove ripples, data is filtered in MATLAB using a moving average filter taking 100 samples, resulting in the second curve in Figure 8. 19 CV’s of other samples showing before and after filtering can be found in Appendix A. From this point on, only the filtered data curves are presented. Figure 8 Sample 1 CV at 1mV/s showing data before and after filtering. All samples tested are initially discharged and in a reduced state at -0.8V vs. Ag/AgNO3. When a voltage ramp is applied, current of the sample increases as PF6- ions move into polypyrrole, consequently charging it. At 0 V vs. Ag/AgNO3, the sample is fully oxidized and charged; as such, the current is flat, similar to a fully-charged capacitor. However, beyond 0.2V vs. Ag/AgNO3, current further increases until 0.6V, where the input voltage is reversed and discharges the sample. The tail observed in the 0.2 to 0.6 V vs. Ag/AgNO3 region suggests that there are parasitic reactions such as the breaking down of water, as moisture in the experimental setup can exist. Graphs in Figure 9 and Figure 10 below show the results of Sample 1 at scan rates of 1 mV/s and 10 mV/s. 20 Tail Oxidizing/Positive Scan Reducing/Negative Scan Figure 9 CV at 1 mV/s from -0.8 V to +0.6 V vs. Ag/AgNO3 of Sample 1. Tail Figure 10 CV at of 10 mV/s from -0.8 V to +0.6 V vs. Ag/AgNO3 of Sample 1. 21 At both scan rates, data in the positive scan diverges from the expected steady capacitive current in the 0.4 V to 0.6 V region, as indicated by the circles in Figure 9 and Figure 10. For the case of 1 mV/s, the average difference in current between the data and a purely capacitive model in this 0.4 V - 0.6 V region is 1.3·10-4 A, whereas at 10 mV/s, the difference in current is also 1.3·10-4 A. Because the average differences at the two scan rates are the same, this current is likely due to parasitic reactions as the current from such electron transfer kinetics limited reactions will be independent of scan rate and only dependent on the applied voltage (Butler-Volmer equation), whereas capacitive current is proportional to scan rate. Also the Butler-Volmer relationship predicts an exponential increase in current with voltage, which is consistent with the rising voltage seen in Figure 9 and Figure 10. Since the response involves some Faradaic as well as capacitive components, particularly in the 0.4-0.6V vs. Ag/AgNO3 region, this response is removed before fitting of the resistive-capacitive transmission line model of Figure 4 (the ideal model assumes no Faradaic reactions occur). To account for the existence of Faradaic reactions, the current difference found in the positive scan is added to the current in the 0.4 – 0.6V vs. Ag/AgNO3 region during the negative scan. This compensation is required to treat the sample as if there were no parasitic reactions; with compensation, the rise time to capacitive (constant current) response is much faster during the reducing scan, where PPy is oxidized and being discharged. Current of reduced PPy in Figure 9 and Figure 10 also diverges from the fit in the -0.6 V to -0.8 V window during the reducing scan. Using the same strategy employed to interpret the positive voltage scan, if the source of this increase in current is again due to kinetics limited parasitic reactions then the rising currents should be comparable in the test results between 1 mV/s and 10 mV/s. However, this is not the case; the average current in the -0.6 to -0.8 V in 1 mV/s scan, after subtracting estimated capacitive current, is -8.1·10-4 A, while the average current in the 10mV/s case is -7.3·10-3 A. As the average currents are not in the same range, the increase in current is not due parasitic reactions and cannot be compensated in the same fashion; in fact, because the difference in currents scale with scan rate, it suggests that this tail seen at 22 -0.6 to -0.8 V vs. Ag/AgNO3 is due to a voltage dependent capacitance, and that the capacitance of polypyrrole can be dependent on the oxidation state. 2.4.2 Results and Discussion Table 3 presents the areas, thicknesses, masses and finally densities for the 6 samples of PPy grown on glassy carbon. The samples’ dry masses result in the same range, consequently suggesting the amount of pyrrole polymerized is directly proportional to the number of coulombs supplied in a deposition. As current density increases, thicker films with higher volumes are produced, resulting in lower densities. Apparent densities and thicknesses are presented in Figure 11. The uncertainties in thickness and mass are estimated to be 10% and 20%, respectively. This is to account for the chances of samples being in different oxidation states, as PPy at different oxidation states can contain different amounts of salt. The average uncertainty in density is Current Density (mA/cm2) Total Deposition Time (h) Area (cm2) Thickness (µm) -30 0.125 4 3.14 7.0±0.7 2 -30 0.250 2 3.14 10±1 3 -30 0.500 1 3.14 9.8±1 4 25 0.125 4 3.14 9.0±0.9 5 25 0.250 2 3.14 9.5±1 6 25 0.500 1 3.14 18±2 0.0032 ±0.0006 0.0044 ±0.0009 0.0038 ±0.0008 0.0044 ±0.0009 0.0041 ±0.0008 0.0049 ±0.001 Density (g/cm3) Temperature (°C) 1 Mass (g) Sample No. determined by the root mean square of the uncertainties in volume and mass. 1.5±0.4 1.4±0.4 1.2±0.3 1.6±0.4 1.4±0.4 0.9±0.2 Table 3 Area, thickness and mass of PPy prepared. 23 Figure 11 Relationship between density and thickness for PPy films from different growth conditions. By comparing Sample 1 to Sample 6, an increase in thickness results in a lower apparent density. In literature, increased compactness of the films produced at lower current density is expected as more time is allowed for better alignment of pyrrole chains [24]. As for samples grown in room temperature, the same trend is observed with respect to a lower current density yielding denser films. Features seen at the surface of samples 4-6 are much larger than samples 1-3. Scanning electron microscope (SEM) images of each sample, taken by Derrick Horne at the UBC Bioimaging Facility, are shown in Figure 12. The images presented are representative of the entire surface of each sample, as several images at various areas of the samples were taken to check for consistency. Sample 1 is dense and uniform, with white residue, suspected to be left behind from acetonitrile. The surface of sample 2 is similar to that of sample 1. As for sample 3, it is noticeably rougher with larger clouds of polypyrrole. As for samples grown in room temperature, features of the film are much larger, more pronounced, and more closely spaced. On sample 6, globules of PPy appear 24 to be more voluminous, consistent with the increased thickness mentioned in Table 5. The SEM images of PPy obtained are consistent with those published in literature [23]. 25 25°C Sample 1 Sample 4 Sample 2 Sample 5 Sample 3 Sample 6 0.500 mA/cm2 0.250 mA/cm2 0.125 mA/cm2 -30°C Figure 12 SEM images of PPy grown at -30°C and 25°C 26 Cyclic voltammetry tests can characterize the charging and discharging behaviour of PPy. Figure 13 and Figure 14 show CV results at scanning rates of 1 mV/s and 10 mV/s between -0.8 V and + 0.6 V vs. Ag/AgNO3. While typical cyclic voltammetry curves have axes of I and V, y-axes in these figures have been modified to normalized current, with units of A/cm2, due to convenience of model-fitting. The model fit employs the ionic conductivity and volumetric capacitance in the transmission line of Figure 4. Polypyrrole is fully reduced at -0.8 V vs. Ag/AgNO3 and fully oxidized at +0.6V. Over this wide potential range, PPy transitions from a discharged to charged state like a capacitor, evident by the rise in current at negative potentials to a flat current response at positive potentials. This capacitance response is exhibited in all 6 samples, at both scanning rates of 1 mV/s and 10 mV/s. Despite the consistent capacitive performance, it appears a “hump” exists in particularly sample 4 and sample 5 when PPy transitions from a reduced to oxidized state. Warren explains the appearance of this “hump” in CV is likely a combined effect of the large changes in electronic and ionic conductivity when PPy transitions from insulating to conducting [22]. This “hump”, however, was observed in earlier cycles for all other samples. As the sample is cycled, the magnitude of the hump at ~ -0.4 V vs. Ag/AgNO3 decreases, eventually flattening out. Since this effect can only be observed in certain samples, it is suspected that each sample requires a different number of chargedischarge cycles to overcome this “first-cycle effect”. A sample of its sequential changes in CV is presented in the Appendix. Of all the CV’s, sample 6 results resemble more of a parallelogram rather than a rectangle. After disassembling the setup for sample 6, it was found that the PPy film was no longer fully contacting the glassy carbon backing. Therefore the shape of a parallelogram suggests that the sample was high in electronic resistance. 27 25°C 0.500 mA/cm2 0.250 mA/cm2 0.125 mA/cm2 -30°C Figure 13 CV at 1 mV/s from -0.8 V to +0.6 V vs. Ag/AgNO3. 28 25°C 0.500 mA/cm2 0.250 mA/cm2 0.125 mA/cm2 -30°C Figure 14 CV at 10 mV/s from -0.8 V to +0.6 V vs. Ag/AgNO3. 29 The model described in Section 2.2 is fit to CV’s to estimate capacitance, ionic conductivities and time constants of PPy. Values obtained from 1 mV/s and 10 mV/s CV’s are presented in Table 4 and Table 5. Each sample is estimated for charging capacitance C+, discharging capacitance C-, and Cavrg is the average of the charging and discharging capacitances. The volumetric capacitance, Cvol, and specific capacitance, Cspec, are based on Cavrg. Finally, ionic conductivities and time constants are also separated into charging and discharging: κ+, κ-, τ+, and τ-. With the ionic conductivity and Cvol values obtained from the fit, time constants are then found based on equation 3. The uncertainties in C+ and C- are found by varying the fit of current density in the simulation. Over a ±5% difference in C+ and C- causes the simulated current density to no longer align with CV measurements. The uncertainty in Cspec is a root mean square average of the uncertainties in mass and capacitance. Likewise, the uncertainty in Cvol is an average of the uncertainties in volume and capacitance. The uncertainty in κ is based on the sensitivity of changing ionic conductivity in the simulation (below a ±15% change was unnoticeable in simulation). Finally, the uncertainty in τ is a root mean square of the uncertainties of volumetric capacitance, thickness, and ionic conductivity. 30 Sample No. C+ (F) C(F) Cavrg (F) Cvol (F/cm3) Cspec (F/g) κ+ (S/cm) τ+ (s) κ(S/cm) τ(s) 1 2 3 4 5 6 0.44±0.02 0.39±0.02 0.41±0.02 0.50±0.02 0.55±0.03 0.54±0.03 0.31±0.02 0.30±0.02 0.31±0.02 0.35±0.02 0.43±0.02 0.23±0.01 0.38±0.02 0.35±0.02 0.36±0.02 0.42±0.02 0.49±0.02 0.38±0.02 1.7±0.3·102 1.1±0.2·102 1.2±0.2·102 1.5±0.3·102 1.6±0.3·102 6.8±1.0·101 1.2±0.2·102 7.8±1.6·101 9.4±2.0·101 9.6±2.0·101 1.2±0.2·102 7.8±1.4·101 2.8±0.4·10-7 4.9±0.7·10-7 4.5±0.7·10-7 3.5±0.5·10-7 3.0±0.5·10-7 2.2±0.3·10-7 3.0±0.8·102 2.3±0.6·102 2.5±0.7·102 3.5±0.9·102 5.0±1.3·102 1.0±0.3·103 8.5±0.1·10-7 1.1±0.2·10-6 1.1±0.2·10-6 1.0±0.2·10-6 9.5±0.1·10-7 1.0±0.2·10-6 9.8±0.3·101 1.0±0.3·102 1.0±0.3·102 1.2±0.3·102 1.6±0.4·102 2.2±0.6·102 κ(S/cm) τ(s) Table 4 Capacitance, ionic conductivity, and time constant from CV at 1 mV/s. Sample No. C+ (F) 1 2 3 4 5 6 0.33±0.02 0.36±0.02 0.38±0.02 0.43±0.02 0.48±0.02 0.37±0.02 C(F) Cavrg (F) Cvol (F/cm3) Cspec (F/g) κ+ (S/cm) τ+ (s) 0.33±0.02 0.33±0.02 1.5±0.3·102 1.0±0.2·102 1.1±0.2·10-7 6.7±1.7·102 2.3±0.4·10-7 0.31±0.02 0.34±0.02 1.0±0.2·102 7.6±1.6·101 1.5±0.2·10-7 7.3±1.9·102 2.5±0.4·10-7 0.36±0.02 0.37±0.02 1.2±0.2·102 9.7±2.0·101 2.2±0.3·10-7 5.3±1.4·102 4.4±0.7·10-7 0.41±0.02 0.42±0.02 1.5±0.3·102 9.5±2.0·101 1.5±0.2·10-7 8.0±2.1·102 3.0±0.5·10-7 0.44±0.02 0.46±0.02 1.5±0.3·102 1.1±2.3·102 1.5±0.2·10-7 9.3±2.4·102 3.0±0.5·10-7 0.30±0.01 0.33±0.02 5.9±1.0·101 6.8±1.4·101 1.5±0.2·10-7 1.3±0.3·103 3.0±0.5·10-7 Table 5 Capacitance, ionic conductivity, and time constant from CV at 10 mV/s. 3.2±0.8·102 4.4±1.1·102 2.6±0.7·102 4.0±1.0·102 4.6±1.2·102 6.4±1.6·102 31 Results in Table 4 and Table 5 are presented in plots for easier comparison between 1 mV/s and 10 mV/s. In Figure 15, capacitances during positive scan (charging) and negative scan (discharging) at scanning rates of 1 mV/s and 10 mV/s are shown. Figure 15 Capacitances from 1 mV/s and 10 mV/s CV. Capacitance of PPy does not depend on scan rates at long times; however, it appears charging capacitances at 1 mV/s are consistently higher while discharge capacitances are lowest compared to capacitances achieved at 10 mV/s. Charge efficiency is the amount of charge available for discharge divided by the amount of charge input during charging. One way to determine charge efficiencies is based on comparing discharging capacitances to charging capacitances, as the voltage for charging and discharging are the same. A more accurate method for calculating efficiency is to sum all positive currents to determine the total amount of charge input, and compare this with the sum of all negative currents. With this method, the efficiency for sample 1, stays consistent at ~82% between earlier and later cycles, whereas the efficiencies based on capacitances are 70% for earlier cycles, dropping down to 60% for 32 later cycles. Part of the inefficiency is due to a gradual degradation of the films with cycling, and to parasitic reactions in general. The large potential range the films are cycled in is associated with some degradation, causing polypyrrole to reduce in capacitance and to charge more slowly. Samples were first tested at 10 mV/s, then at 1 mV/s; as such, results at 1 mV/s were obtained from later cycles. Figure 16 is a CV on the second scan and 10th scan on Sample 1. The capacitive current for both charging and discharging decreases, resulting in a lower capacitance. As well, the rise time at -0.8 V vs. Ag/AgNO3 becomes longer, suggesting the sample to become more resistive as it is cycled. A larger tail of current between -0.6 V to -0.8 V vs. Ag/AgNO3 shows that more charge is consumed by parasitic reactions at later cycles. Figure 16 Degradation of Sample 1 results in a decrease in charging/discharging capacitance, a longer rise time at reduced potentials, as well as a larger tail of current at -0.8V vs. Ag/AgNO3. 33 For simplicity, the average of the charging and discharging capacitances are used for analysis. In selecting the appropriate polypyrrole growth conditions, it is noted that the deposition current density and temperature do not appear to affect the capacitance of PPy, as the capacitance of each sample is dependent on the total number of coulombs supplied for deposition. Nevertheless, as previously mentioned, the volume is affected by the deposition current density. Hence in order to attain high volumetric capacitance with polypyrrole, low deposition current densities are required. Aside from capacitance, another important parameter to characterize the performance of PPy is ionic conductivity. A high ionic conductivity within the electrode suggests that ions can easily transport through the film for charging/discharging. Figure 17 is a plot of reduced and oxidized PPy ionic conductivities found from CV’s at 1 mV/s and 10 mV/s. Figure 17 Estimated ionic conductivity of PPy estimated by fitting a model to cyclic voltammetry results. 34 At both scan rates, the estimated ionic conductivities of reduced polypyrrole are lower than those observed in the oxidized state. The oxidized state (positive voltages) ionic conductivities found are consistent with estimates from literature, which range from 8.5·10-7 S/cm to 1.1·10-6 S/cm [24,25]. It is known that the oxidation degree controls the ionic conductivity of the polymer -- when polypyrrole is oxidized and swollen, the ionic conductivity is higher than when PPy is reduced and compact [26]. This suggests that a sample which is less dense is expected to have higher ionic conductivities. However, given the amount of uncertainty in approximating ionic conductivities, this trend was not clearly observed for the six samples for which densities vary from 0.9 g/cm3 to 1.6 g/cm3. It has been found that polymers grown at low temperatures are higher in electronic conductivity [23]. Synthesis at low temperatures produce PPy with fewer structural defects, as more unwanted side reactions occur at higher temperatures [27]. Although the samples grown at room temperature have lower electronic conductivities, it is their ionic conductivities which continue to dominate the rate of charging/discharging as these samples are adhered onto a conductive backing. The effects of deposition temperature and current density on the ionic conductivity of PPy are not significant. Finally, the last parameter being considered is the time constant, which is directly related to the volumetric capacitance, thickness, and ionic conductivity, as in equation (3). Figure 18 is a plot of time constant (τ) versus the sample number. 35 Figure 18 Time constant values of PPy from CV at 1 mV/s. The resulting time constants lead to a number of observations similar as before. Not only does the scan rate affect ionic conductivity, time constants are also affected; samples with lower ionic conductivities result in larger time constants. In addition, unlike the other samples, sample 6 did not adhere well onto the glassy carbon disc; as such, the time constant is the largest, requiring the most time to charge and to discharge, at both scan rates of 1 mV/s and 10 mV/s. Moreover, Sample 6 took the longest time to charge due to its increased thickness. The thickness of PPy can dictate the charging time times required, since PF6- ions in the electrolyte need to travel a longer distance to charge a thicker film entirely. Hence in order to reduce charging times, the PPy layer needs to be kept as thin as possible while being able to offer the capacitance required. To consider the effect of uncompensated resistances mentioned in Section 2.3, approximations are presented here. First, solution resistance Rs follows the equation below, where the solution conductivity σs is assumed to be on the order of 10-2 S/cm [28], and the distance between the reference and working electrode, l, is ~ 0.5cm. Area of each sample, A, is 3.14 cm2. 36 𝑅𝑠 = 𝑙 𝜎𝑠 𝐴 = 16 Ω (10) Second, the contact resistance Rc, measured with a 2-point probe, is approximated to be 10 Ω. Thus the total uncompensated resistance, Ru, is 26 Ω. According to Table 6, Ru represents a small fraction of RT+ and RT-, which confirms that the modelling is correct in assuming uncompensated resistances are negligible. Sample 1 2 3 RU (Ω) RT+ (Ω) RT(Ω) 26 ± 7 796±199 669±167 698±174 819±205 1008±252 2606±651 262±66 298±75 285±71 287±72 318±80 573±143 4 5 6 Table 6 Approximation of uncompensated resistances in CV tests on Samples 1 – 6. While similar gravimetric capacitance can be obtained from PPy grown at room temperature and at higher rates, based on the results in this chapter, electrosynthesis of polypyrrole electrodes should continue to be performed at low current density and at low temperatures if performance is the primary concern. PPy grown in this condition generally exhibits higher volumetric/specific capacitance, density, and higher electronic conductivity[29]. These are all characteristics which contribute to a high quality electrode material for supercapacitors. Where time (and hence cost) are important, room temperature deposition can be an alternative, though producing films with slightly poorer performance. The key questions to be resolved in deciding on temperature and rate will involve weighing cost, performance and perhaps most importantly, the impact on lifetime, cycle life and leakage of different deposition methods – and this will require further experimentation. Further investigation into such issues using specifically PF6 - as the dopant ion will be required. 37 2.5 Summary To summarize, the performance of polypyrrole synthesized with different current densities and at different temperatures are analyzed. Here are the observations: Low deposition current density of 0.125 mA/cm2 at -30°C is capable of producing films of volumetric and specific capacitances up to (1.7±0.3)·102 F/cm3 and (1.2±0.2)·102 F/g; Low deposition current density of 0.125 mA/cm2 at -30°C created a more ordered film with a density of 1.5±0.4 g/cm3, whereas higher deposition current of 0.500 mA/cm2 at 25°C produced a more voluminous PPy sample with density of 0.9±0.2 g/cm3; Ionic conductivities found from 1 mV/s CV measurements are not clearly affected by deposition conditions: reduced PPy have ionic conductivities ranging from (2.2 to 4.9)·10-7 S/cm, while oxidized PPy have ionic conductivities varying from (8.5 to 11)·10-7 S/cm; PPy’s performance in specific and volumetric capacitance, ionic conductivity and rate of charging is not hugely affected by deposition current density and temperature; however, a comprehensive analysis also requires effects on the rate of degradation, cycle life, and cost to be analyzed, for which further work is required. 38 3 Polypyrrole on Nanofibres The hypothesis explored in this chapter is that porous electrodes can charge faster since ions are able to travel faster through liquid filled pores than through polymer. This chapter is divided into 4 sections. Section 3.1 explains the use of nanofibres (NF) to increase the porosity of polypyrrole (PPy). Section 3.2 explains the test equipment and implementation, and Section 3.3 analyses experimental results obtained from cyclic voltammetry (CV) of NF samples coated with PPy. It is found that PPy on NF can charge and discharge faster as a result of higher ionic conductivities. However, the improvement in response time is less than expected, partly due to uncompensated resistances in the experiment which cause mass transport within the film to no longer dominate charging time. 3.1 Background To improve on energy and power density of supercapacitors, the ratio of active electrode material versus total device mass can be possibly improved. One way to increase energy density is to increase active electrode material thickness to minimize the relative amount of collector and separator materials also used in the cell. However as thickness increases in polypyrrole (PPy), ions from the electrolyte travel a longer path to charge and discharge the sample, leading to slower charging and reducing power density. To use PPy in large amounts (as required for larger supercapacitors), one option may be to increase the porosity of polypyrrole so that thick samples of PPy can nevertheless charge quickly, likely at the expense of some energy density due to the reduction of active material per volume. There has been some previous research to increase the porosity of PPy, mainly through incorporating a very porous structure for PPy to “mount” on. Tan and Izadi electrochemically deposited different amounts of PPy onto carbon fibres and analyzed the subsequent benefits of using a scaffold [30]. They found that beyond a certain loading of polypyrrole, the rate of charging slows, likely due to clogging of pores. Jurewicz found that if carbon nanotubes are modified by polypyrrole, the specific 39 capacitance was enhanced up to 163 F/g (per gram of active material of electrode) in 1M H2SO4 can be achieved[31]. Carbon nanofibres (NF) remain more cost effective to manufacture than carbon nanotubes. Thus a number of groups are exploring the use of nanofibres coated with conducting polymers to reduce charging time constants. For example, Kim and Sharma performed an in situ chemical polymerization on vapour grown NF[32]; they demonstrated that in 6M KOH, a specific capacitance of ~545 F/g was obtained with a fast scan rate 200 mV/s for PPy on NF, whereas the theoretical specific capacitance of pure PPy is 620 F/g [33]. Typically at high scan rates the capacitance attained drops because ions cannot charge the entire thickness of the film in time. In addition, Kim and Zhang investigated the performance of polypyrrole on NFs, where composite films of NF and PPy were found to have electronic conductivities of 40 – 63 S/cm [34]. Finally Jang et al. has also deposited polyaniline (PANI) onto NF[35]. It was found that both the electrical conductivity and specific capacitance were functions of the thickness of PANI – as the thickness of PANI increased, the electrical conductivity reached a maximum of 30 S/cm and a specific capacitance of 264 F/g in 1M H2SO4. In this chapter we investigate how the incorporation of carbon nanofibres affects the performance of PPy. Carbon nanofibres prepared by other research groups, as shown in [36,37],measure 10-40 μm in length, and their electronic conductivity is typically several S/cm. Nanofibres samples used in this study are supplied by Nicole Lee, a member of Dr. Frank Ko’s group in Materials Engineering at UBC. These nanofibre samples are fabricated by electro-spinning polyacrylonitrile (PAN) into very fine fibres, and are then carbonized to become electronically conductive (15-18 S/cm). Typical densities of the carbonized NF samples are 0.150 to 0.200 g/cm3. Using the NF samples as electrodes, polypyrrole is electrochemically deposited onto them. The final product with PPy coated NF is referred to as NF/PPy in this thesis. The surface area of carbonized PAN samples is on the order 10 m2/g. Since the surface is large, the same amount of polypyrrole deposited onto a higher surface area template will lead to an overall thinner layer of polypyrrole per fibre. Providing that some empty space is left between fibres after deposition, the sample is porous and permeable, increasing bulk ionic conductivity. Electrical conductivity may also be improved by the 40 addition of the conductive nanofibres, at least in the reduced state, thereby potentially increasing rate and improving charge transfer. By using NF with PPy, the assumption is that ionic conductivity of PPy can be increased to approach solution conductivity. The following analysis can further explain the potential benefit of using PPy on NF. The analysis assumes that all pores are in parallel and can contact solution in the same fashion, as well as migration dominating the movement of ions. The effective ionic conductivity of NFs is a relationship between the solution conductivity κs, ionic conductivity κi, and porosity ɣ: 𝜅𝑒𝑓𝑓 = 𝜅𝑠 𝛾 + 𝜅𝑖 1 − 𝛾 . (11) The overall ionic conductivity of NF/PPy samples κeff is limited to the solution conductivity κs when NF is entirely porous, and is limited to the actual ionic conductivity of PPy through the thickness, κi, when NF is not porous ( ɣ=0 ). Because κs is estimated to be 10-2 S/cm [29] and κi of PPy is in the range of 10-7 S/cm [25], the following simplification is made: 𝜅𝑒𝑓𝑓 = 𝜅𝑠 𝛾. (12) Pure polypyrrole’s volumetric capacitance Cvo and thickness lo are also modified to consider porosity: 𝐶𝑣 = 𝐶𝑣𝑜 1 − 𝛾 . 𝑙𝑜 𝑙= . 1−𝛾 (13) (14) In this case the thickness, l, is chosen such that total capacitance remains constant for a given electrode area. Recall equation (3) gives a relationship between volumetric capacitance, thickness and ionic conductivity. Hence, the improvement in time response between NF/PPy and pure PPy is expected to be: 𝜏𝑁𝐹/𝑃𝑃𝑦 𝜏𝑃𝑃𝑦 𝐶𝑣 𝑙 2 𝜅𝑒𝑓𝑓 𝜅𝑖 = 𝑜 2 = . 𝐶𝑣 𝑙𝑜 𝜅𝑠 𝛾 1 − 𝛾 𝜅𝑖 (15) 41 In the case where a PPy-coated NF sample has the same amount of PPy as another pure PPy sample, NF can potentially improve the charging time of PPy by 2500 times, assuming the porosity factor measures 0.6. This is true if it the model, which assumes the charging time through the pores dominates the time response of PPy, is correct. However, if the benefits of using NF with PPy shown are not as dramatic, perhaps other mechanisms may be limiting the charging time of PPy despite increasing the porosity. Examples of such mechanisms can include uncompensated resistances, and a depletion of ions in the solution which can cause concentration gradients. 3.2 Experimental Setup 3.2.1 Deposition Rather than depositing PPy onto glassy carbon, PPy is deposited onto the carbon nanofibre samples. Bare nanofibre samples prior to deposition measure approximately 50-55 μm in thickness. For even deposition of PPy onto nanofibres, a counter electrode is placed on each side of the sample, as illustrated in Figure 19. CE WE Figure 19 Deposition setup using nanofibres. 42 Contact leads to respective electrodes run from outside to the inside of a glass cell through a septum at the top. One end of the lead is a stainless steel clamp which is used to secure the position of a NF sample; the clamp also minimizes contact resistance, which can otherwise be significant and determine charging rate. Once electrodes are secured into place, the deposition solution containing pyrrole monomers is poured into the cell. To avoid the deposition solution from being contaminated by the lead and clamp, depth of the contact lead is adjusted so only nanofibres are submerged in solution. Once solution inside the cell has been bubbled with nitrogen gas with a needle through the septum, the glass cell is placed in an ethylene glycol bath, already cooled to -30°C. After the setup has been in the bath for approximately 3 hours, PPy can be deposited on the NF. A constant current of 0.125 mA/cm2 is applied for electrodeposition of PPy. Three samples of NF/PPy are produced, labelled as sample 4h, 12h, and 16h, with deposition times of 4 h, 12 h and 16 h, respectively. The different deposition times produce NF/PPy samples with different polypyrrole loadings. Areas used to determine the applied deposition currents are the external areas of nanofibres exposed to solution, as defined in the Figure 20. Note that the effective current per electrode area is much lower than that used in the previous chapter since the internal surface area of the NF samples is very high. After deposition, samples are washed with acetonitrile and air-dried overnight before final masses and thicknesses are determined. Figure 20 Sample 4h, 12h and 16h dimensions from left to right (in cm). 43 3.2.2 Sample Test Setup As in Chapter 2, an electrolyte of 0.2 M TEAP in ACN is used. After bubbling the electrolyte with nitrogen gas, a reference electrode is placed in the cell, as shown in Figure 21. A porous carbon electrode (Gore EXCELLERATOR, Delaware, USA) is used as the counter electrode. Cyclic voltammetry is performed on the different samples, with a scan rate at 1mV/s between -0.8 to +0.4V or +0.6V vs. Ag/AgNO3. The maximum voltage was decreased from 0.6V to 0.4V vs. Ag/AgNO3 to reduce parasitic reactions. WE CE RE Figure 21 Three-electrode test setup for NF/PPy. 44 3.3 Results and Data Analysis 3.3.1 Data Adjustment To enable comparison with results from Chapter 2, time constants need to be normalized. In Chapter 2, only one face of the PPy film is exposed to solution while the other side is adhered to glassy carbon. When only one face of a sample is exposed to solution, ions travel the thickness, 𝑙, for charging/discharging. However, NF/PPy samples are free-standing and both faces are exposed to electrolyte. As such, for the case with two faces exposed, the ion penetration thickness is 𝑙/2, which affects the time constants calculated. In this chapter, results are normalized to having just one side accessed by electrolyte. Recall that the time constant equation is: 𝐶𝑣𝑜𝑙 𝑙 2 𝜏= . 𝜅 (16) If the effective thickness of NF/PPy is reduced by half, time constants used for comparison should then be calculated by: 𝜏= 𝐶𝑣𝑜𝑙 𝑙 2 . 4𝜅 (17) In addition, CV data on NF/PPy are filtered and compensated for any parasitic reactions in oxidized states, as presented in Section 2.5.1. 45 3.3.2 Results and Discussion Because PPy is deposited onto a portion of the NF sample, there is an obvious visual difference between bare nanofibres and nanofibres coated with PPy, making it easy to determine the depth to which to submerge samples. The shade variation along the length of the film in Figure 22 is due to a varying amount of polypyrrole deposited onto the NF sample; this variance leads to a thickness difference along the length of the sample. NF/PPy Bare NF Figure 22 NF/PPy after electrodeposition. PPy is clearly visible on NF. SEM images below, taken by Nicole Lee, are on Sample 12h (12 h deposition) before and after deposition. Figure 23 is an SEM image taken on an NF sample before deposition. According to this image, bare nanofibres have an approximate diameter of 400 nm. 46 Figure 23 SEM images of NF before deposition. After deposition, the diameter of fibres in the center of Sample 12h have increased to 500 nm, thus the thickness of polypyrrole on fibres is on average 50 nm. As shown at the top of Figure 24, each individual fibre in the center can still be easily distinguished even after deposition, suggesting that the polymer thickness is very thin. Salt crystals seen adhered onto the sample are a result of cycling as SEM images were taken after electrochemical tests. 47 Centre Edge Figure 24 SEM images of Sample 12h after deposition. The bottom of Figure 24 is an SEM image of the edge of Sample 12h. Notice the difference between the centre and edge of the sample; individual fibres can no longer be easily distinguished as the edges become like a bulk piece of PPy. Since the separation distance of the fibres along the same layer as well as between layers is small, this makes it easy for PPy to grow across the same layer as well as through the depth. Due to this build up of PPy at the edges, film thicknesses at the edges for Sample 12h and 16h are consequently 30-40 µm thicker than the center of the samples. More PPy collected at the edges of samples is likely due to the stronger electric fields at edges. Ions are transported to the edge of the sample at a higher rate and more polypyrrole is formed at that particular area. Similarly, migration of ions is expected to 48 be faster at the edges than in the centre. Because the bare nanofibres are very low in density, a sample not significantly loaded with PPy – like Sample 4h -- does not change in thickness after deposition. To fit our CV data to the model, a constant thickness is required to define the length of the transmission line model for each sample. Therefore an average NF/PPy layer thickness is used in the modelling. To determine the average thickness, areas on samples 12h and 16h are first probed and defined to particular thicknesses, as shown in Figure 25 and Figure 26. The total volumes are then found by summing all the volumes from each region together. Finally average thicknesses are found by dividing the total volume by the apparent area of each sample, and these averages are displayed in the last Deposition Time (hr) Thickness Before Deposition (µm) Thickness After Deposition (µm) Average Thickness After Deposition (µm) Sample column of Table 7. 4h 4 55 ± 0.6 55 55 ± 0.6 12h 12 52 ± 0.5 52 to 90 (edges) 67 ± 0.7 16h 16 52 ± 0.5 52 to 85 (edges) 75 ± 0.8 Table 7 Average thicknesses of NF/PPy Samples 4h – 16h. 49 Figure 25 Thickness distribution of Sample 12h coated with PPy (in cm unless indicated). Figure 26 Thickness distribution of Sample 16h coated with PPy (in cm unless indicated). Table 8 summarizes the physical dimensions and masses for samples 4h-16h. Each sample holds a different amount of polypyrrole, with sample 16h loaded with the most PPy since the deposition time is longest. 50 Volume (cm3) Mass Before Deposition (g) Mass After Deposition (g) Mass of PPy (g) Density (g/cm3) Area (cm2) Sample 1.1 ±0.1 2.2 12h ±0.3 2.2 16h ±0.3 4h (5.8±0.9)·10-3 (2.4±0.2)·10-3 (5.0±1.0)·10-3 (2.6±0.3)·10-3 (1.5±0.3)·10-2 (3.4±0.3)·10-3 (1.4±0.3)·10-2 (1.1±0.2)·10-2 (1.6±0.3)·10-2 (4.9±0.5)·10-3 (2.0±0.4)·10-2 (1.5±0.3)·10-2 0.87 ±0.2 0.96 ±0.3 1.2 ±0.3 Table 8 Volume and mass of polypyrrole-coated nanofibre samples. Figure 27 shows the change in density of the samples with the amount of PPy deposited. The first point on the graph at 0 g of PPy represents the density of NF alone, at 0.17 g/cm3. Also noticed is that the density of the sample increases as more polypyrrole is loaded on it. Sample 16h reaches an overall density of 1.2 ± 0.3 g/cm3, approaching the bulk PPy density of 1.5 ± 0.3 g/cm3. Figure 27 Density versus PPy mass on NF/PPy. 51 Sample 12h is first tested at 1 mV/s between -0.8 V to 0.6 V vs. Ag/AgNO3, but for samples 4h and 16h tested, this range is modified to -0.8 V to 0.4 V to minimize parasitic reactions. The model described in Section 2.2 is applied to the CV data for Samples 4h-16h, as shown in Figure 28 to Figure 30. Figure 28 CV at 1 mV/s between -0.8 V to +0.4 V vs. Ag/AgNO3 on Sample 4h. 52 Figure 29 CV at 1 mV/s between -0.8 V to +0.6 V vs. Ag/AgNO3 on Sample 12h. Figure 30 CV at 1 mV/s between -0.8 V to +0.4 V vs. Ag/AgNO 3 on Sample 16h. 53 In Table 9, specific and volumetric capacitances of PPy and NF/PPy are shown. Sample 1 is pure PPy grown with 0.125 mA/cm2 at -30°C for 4 hours. Each sample is estimated for charging capacitance C+, discharging capacitance C-, and Cavrg is the average of the charging and discharging capacitances. The average capacitance Cavrg is PPy mass, yielding Cspec PPy. C+ (F) PPy 0.44±0.02 4h 0.3±0.02 0.38±0.02 (1.7±0.3)·102 Cspec PPy (F/g) Cvol (F/cm3) total mass of NF/PPy, yielding Cspec NF/PPy ; Cavrg (F) C(F) volume, yielding Cvol; Sample Cspec NF/PPy (F/g) normalized using three parameters: -- (1.2±0.1)·102 0.37±0.02 0.22±0.01 0.29±0.01 (5.1±0.9)·101 (5.9±1.2)·101 (1.1±0.2)·102 12h 1.7±0.1 1.6±0.08 1.6±0.1 (1.1±0.2)·102 (1.1±0.2)·102 (1.5±0.3)·102 16h 2.8±0.1 2.3±0.2 2.6±0.1 (1.6±0.3)·102 (1.3±0.3)·102 (1.7±0.3)·102 Table 9 Volumetric and specific capacitances of NF/PPy samples. Bare NF samples were tested with CV and it was found that the specific capacitance of NF is negligible relative to the specific capacitance of pure PPy. Therefore the capacitances exhibited by NF/PPy samples are due to the presence of PPy. As NF samples are loaded with more PPy, Cavrg, Cvol, Cspec PPy, and Cspec NF/PPy all increase. Sample 4h is loaded with the least PPy and Sample 16h with the most PPy, therefore Sample 4h results the lowest Cavrg while Sample 16h results the highest. Cvol for sample 4h is the lowest of the three NF/PPy samples because much of the sample is void. Sample 12h results in a higher Cvol than Sample 4h, and Sample 16h approaches (1.7 ± 0.3)·102 F/cm3, the Cvol of pure PPy; Sample 16h becomes much like a thick piece of PPy. Charge transfer efficiencies, as in Chapter 2, can be calculated by comparing either the discharging to charging capacitances, or the average of the summation of 54 discharge to charging currents. For Sample 4h, the coulombic efficiency based on capacitance is 60%, or more precisely 77% based on total input/ouput charges. The low efficiency attained by Sample 4h may in part be due to changes in resistance of the cell, as evident in Figure 31 from the more sloped CV at longer times. As for Sample 12h, efficiency based on capacitance is 94% and 100% based on charges. Finally for Sample 16h, the efficiency obtained is 82% with capacitance, and higher at 94% based on charge. Figure 31 Change in the CV with time seen in Sample 4h. The sample becomes more resistive as it is cycled. Beginning represents the 10th cycle, and end represents the 30th cycle. Hughes et al. demonstrated that the specific capacitance of PPy is increased with increased porosity and suggested that this is because ions more easily access the bulk of the film [38]. The specific capacitance normalized to the total weight of each sample results in Cspec NF/PPy as listed in Table 9. In Sample 4h, the ratio of NF mass to PPy mass is high; whereas in Sample 16h, the ratio of NF mass to PPy is lower, leading to a specific capacitance of 130 F/g, similar to pure PPy. Given the amount of uncertainty, it is not evident that Cspec PPy is changed by the NF. However, when considering the mass 55 of PPy alone, there does appear to be an increase in specific capacitance in the porous samples, though not for the 4h hour sample. The effect of NF on ionic conductivity and time constant are also investigated. In Table 10, κ+ represents charging ionic conductivity, κ- represents discharging ionic conductivity, and τ+ and τ- are calculated based on equation 3. Properties of PPy are of Sample 1 from Chapter 2. Sample Cvol (F/cm3) κ+ (S/cm) τ+ (s) κ(S/cm) τ(s) PPy (1.7±0.3)·102 (2.8±0.4)·10-7 (3.0±0.8)·102 (8.5±1.3)·10-7 (9.8±2.5)·101 4h (5.1±0.9)·101 (5.3±0.8)·10-6 (7.2±1.9)·101 (1.2±0.2)·10-5 (3.2±0.8)·101 12h (1.1±0.2)·102 (6.8±1.0)·10-6 (1.8±0.5)·102 (1.3±0.2)·10-5 (9.5±0.2)·101 16h (1.6±0.3)·102 (9.0±1.4)·10-6 (2.5±0.6)·102 (1.5±0.2)·10-5 (1.5±0.4)·102 Table 10 Estimated ionic conductivity and time constant results of NF/PPy samples from CV at 1mV/s. The charging ionic conductivity, κ+, is found during the positive scan of a CV, and the discharging ionic conductivity, κ-, is found from the negative scan. Ionic conductivities are plotted versus apparent density in Figure 32. 56 Figure 32 Ionic conductivity of NF/PPy versus density from 1 mV/s CV. From the CV results at scan rates of 1 mV/s and 10 mV/s in Chapter 2, recall that the charging rate of PPy is limited by the mass transport of ions through the thickness of the film. With the use of NF, the higher internal surface area reduces the distance PF6ions to travel to charge the film. As such, PPy/NF samples all have higher charging and discharging ionic conductivities than pure PPy. Particularly with pure PPy and Sample 4h, although the two have similar amounts of PPy, the ionic conductivities of NF/PPy increase by an order of magnitude. Even with large amounts of PPy on Sample 16h, the ionic conductivities for both charging and discharging are still at least one order higher than bulk PPy. Faster accessibility of ions to the bulk of the electrode therefore increases ionic conductivity[11]. Charging and discharging time constants of NF/PPy samples are presented in Figure 33. Comparing the two samples with ~10 mg of PPy, both the charging and discharging time constants are reduced considerably in the porous film when compared to the pure PPy, despite the much greater thickness of the porous film. Thus for larger amounts of PPy, the advantages of NF are even more obvious. Although Sample 12h holds almost four times as much PPy than the pure PPy sample, the time constants are 57 not significantly different. Furthermore, Sample 16h is five times more capacitive than Sample 1 due to the increased PPy loading, yet the charging time is only somewhat increased compared to Sample 12h. The advantage of using a porous electrode is thus demonstrated by showing a similar charging time for a much larger capacitor. Figure 33 Time constant versus density from 1 mV/s CV. Despite the improvements seen with NF in κ and τ, the results are not as large as expected. In fact, the benefits of using NF with PPy might be expected to be more drastic given that electrolyte ionic conductivities are 1000 times higher than that in bulk PPy. One possible reason is that the assumption the migration of ions through the pores is the rate limiting factor may not be entirely correct. Although results from bulk PPy suggest that it is indeed the ionic conductivity though the thickness of the film which limits charging, it seems to be more complex with using porous electrodes. 58 Another factor which can reduce the effectiveness of NF is uncompensated resistances. Although effort is made to minimize such resistances, they can, however, add up so that the mass transport of ions no longer dominates the response times of NF/PPy samples. The different types of uncompensated resistances not accounted for in modelling are: Rnf, which is the electronic resistance along the length of the NF sample; Rs, which is the solution resistance in the electrolyte; Rc, which is the contact resistance between the instrument and the sample. This analysis is based on the dimensions of Sample 7. Since the electronic conductivity of NF obtained from measurements is ~15 S/cm, the resulting RNF is: 𝑅𝑁𝐹 = 𝜎 𝑙 𝑁𝐹 𝐴 = 13 Ω. (18) For solution conductivity σs, it is assumed to be in the order of 10-2 S/cm [28], and the distance between the reference and working electrode, l, is ~ 0.5cm. The area, A, is two times the apparent area as solution can access both sides of a sample. As such, the solution resistance is for Sample 4h is 23 Ω. Finally, contact resistance, Rc, which was measured with a 2-point probe from the stainless steel clamp to the NF sample is approximated to be 15 Ω. Thus the total uncompensated resistance, Ru, is 51 Ω. Resistances for Sample 8 and Sample 9 are also found using the approach above and are reported in Table 11. Sample RNF (Ω) RS (Ω) RC (Ω) RU (Ω) RT+ (Ω) RT(Ω) 4h 13 ± 3 24 ± 6 15 ± 4 52 ± 19 248 ± 62 110 ± 28 12h 16 ± 4 11 ± 3 15 ± 4 42 ± 14 112 ± 28 60 ± 15 16h 21 ± 5 11 ± 3 15 ± 4 47 ± 15 96 ± 24 57 ± 14 Table 11 Approximation of uncompensated resistances in CV tests on Samples 4h – 16h. 59 Recall that in equation (3) the time constant is a product of the sample’s effective capacitance and resistance. On Sample 4h, Cavrg is 0.29 F while the charging time constant is 72 s, and the discharging time constant is 32 s; the resulting resistances are 248 Ω for R+ (charging) and 110 Ω for R- (discharging). Charging and discharging resistances of NF/PPy samples are also shown in Table 11. If the approximation of Ru is accurate, it is, in fact, large enough to cause a more sluggish response in charging and discharging in CV’s, therefore affecting the fit of ionic conductivities and time constants. Especially for Sample 12h and Sample 16h, the fact that Ru is in the same range as R- indicates the response time for oxidized NF/PPy is limited by the uncompensated resistance, and not the mass transport of PF6 - ions. As mentioned in Section 2.2, because Ru represents a significant fraction of RT, ionic conductivities reported in this chapter therefore provide upper-bound values. The actual ionic conductivities are likely underestimated in the experimental results; to increase accuracy, improvements particularly for the three-electrode test setup include: Reducing the distance between WE and RE, l, to reduce Rs; Increasing solution conductivity by increasing concentration to reduce solution R s; Increasing contact area and force of clamp to reduce Rc. 3.4 Summary The use of nanofibres (NF) can improve the porosity of polypyrrole (PPy), thereby improving the ionic conductivity and time response of PPy. Different amounts of PPy are polymerized onto carbonized polyacrylonitrile nanofibres supplied by Nicole Lee from Dr. Frank Ko’s lab. The samples are analyzed with SEM imaging and threeelectrode electrochemical testing using 0.2 M TEAP ACN, and here are the observations of PPy-coated NF (NF/PPy): Density of the samples increased to that of pure PPy (1.3 g/cm3) with increased deposition of polypyrrole: the density of bare NF is 0.15 g/cm3 and PPy/NF with a 16-hour deposition has a density of 1.2±0.3 g/cm3; 60 Specific capacitance normalized to the entire sample’s mass, NF and PPy, reached a maximum of (1.7±0.3)·102 F/g, similar to that of pure PPy; the specific capacitance of the polypyrrole itself is generally higher in the porous electrodes than in pure polypyrrole. Due to relatively large uncompensated resistances (contact resistance, solution resistance and electronic resistance) in the experimental setup, upper-bound ionic conductivities and time constants are reported: o The estimated ionic conductivity of reduced PPy is more than 10 times larger than that of pure PPy, with pure PPy having a conductivity of (2.8±0.4)·10-7 S/cm and NF/PPy having a much higher ionic conductivity of (5.3±0.8)·10-6 S/cm. This increase in speed is likely because the thickness of PPy on NF is significantly lower than in the pure polypyrrole film and thus the mass transport rate is improved; o The estimated ionic conductivity of oxidized PPy is also increased by more than 10 times, with pure PPy having a conductivity in the range of 10-7 S/cm, whereas PPy/NF has an ionic conductivity of 10 -5 S/cm, again assisted by the thinner layers of PPy; o A NF/PPy containing over four times more polypyrrole than the pure PPy film still retained high ionic conductivities of (6.8±1.0)·10-6 S/cm (reduced) and (1.3±0.2)·10-5 S/cm (oxidized), leading to a similar charging time despite the larger capacitance. For more accurate results in estimating ionic conductivities and time constants, factors which can hinder charging time of NF-coated PPy need to be further identified, because the current model suggests a more drastic improvement that what was achieved. 61 4 Conclusion 4.1 Research Contributions This thesis has two main objectives. The first objective was to determine the tradeoffs in varying electrodeposition current density and temperature of polypyrrole to reduce the cost of production. Through cyclic voltammetry measurements and mathematical modelling, PPy synthesized with the current density of 0.125 mA/cm2 at 30°C produced the densest films (1.5 g/cm3), with high volumetric (170 F/cm3) and specific capacitances (120 F/g). As such, this condition produced the best performing films, with ionic conductivities of 2.8·10-7 S/cm when reduced and 8.5·10-7 S/cm when oxidized. The worst performing film, which offered only 68 F/cm3 and 78 F/g, was produced with the current density of 0.500 mA/cm2 at 25°C. Although deposition conditions did not substantially affect ionic conductivities of PPy, the recommended method for deposition of PPy is at a current density of 0.125 mA/cm2 at -30°C, if time is not a critical factor. The second objective was to increase the porosity of polypyrrole by adhering polypyrrole onto carbonized polyacrylonitrile nanofibres that form a conductive yet highly porous substrate. Varying amounts of PPy were deposited on nanofibres, which were tested with cyclic voltammetry in 0.2M TEAP in acetonitrile. It was demonstrated that with the use of nanofibres, polypyrrole with a normalized thickness of up to 30 µm was able to reach ionic conductivities of 9.0·10-6 S/cm when reduced, and 1.5·10-5 S/cm when oxidized, ten times higher than that achieved by thinner samples of pure polypyrrole. In addition, results suggested that the specific capacitance of polypyrrole could be enhanced to 170 F/g. A NF/PPy sample with a comparable density to bulk PPy, achieved higher ionic conductivities of 9.0·10-6 S/cm and 1.5·10-5 S/cm, for reduced and oxidized states. NF/PPy samples four times in capacitance were able to charge at the same rate as bulk PPy films with significantly lower capacitances. Therefore, by using a porous electrode, the charging time required by PPy is reduced, thus increasing power density. Despite the improvements in ionic conductivities, the reduction in charging time 62 was not as large as expected. This was a result, at least in part, of uncompensated resistances which represented a significant fraction of the total resistance determining charging time. Other suspected mechanisms such as ion depletion may have led to lower than expected results. 4.2 Future Work Possible improvements and future work include: Reducing the amount of error in mass of PPy by washing PPy samples in a volatile solvent such as acetonitrile and allowing samples to dry sufficiently before weighing. Also consider reducing samples first so that the oxidation state is consistent across samples, as dopants in the film can lead to changes in mass. Reducing parasitic reactions in experiments by using distilled pyrrole, purified salts and anhydrous solvents. All experiment setups, whether during deposition of electrochemical testing, should be enclosed as moisture can affect performance of polypyrrole. Reducing solution resistance by increasing the concentration of electrolyte, which effectively increases solution conductivity. Another is to decrease the separation distance between the reference and working electrodes. Investigating the rate of degradation and effects of ambient temperature to fully characterize the performances of PPy synthesized with different conditions, and PPy/NF. Reducing electronic resistance of NF by making NF samples more flexible, as carbonized PAN samples were extremely brittle and often led to breakage. Determine other factors which also can affect charging time of NF/PPy. 63 Bibliography [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] “CAP-XX.” [Online]. Available: http://www.cap-xx.com/products/products.htm. [Accessed: 04-Mar-2011]. “Gold Cap Electric Double Layer Capacitors - Panasonic.” [Online]. Available: http://www.panasonic.com/industrial/electronic-components/capacitiveproducts/gold-cap-electric.aspx. [Accessed: 04-Mar-2011]. A. Burke, “R&D considerations for the performance and application of electrochemical capacitors,” Electrochimica Acta, vol. 53, no. 3, pp. 1083-1091, Dec. 2007. P. Sharma and T. S. Bhatti, “A review on electrochemical double-layer capacitors,” Energy Conversion and Management, vol. 51, no. 12, pp. 2901-2912, Dec. 2010. P. Simon and Y. Gogotsi, “Materials for electrochemical capacitors,” Nat Mater, vol. 7, no. 11, pp. 845-854, Nov. 2008. M. Mastragostino, C. Arbizzani, and F. Soavi, “Conducting polymers as electrode materials in supercapacitors,” Solid State Ionics, vol. 148, no. 3-4, pp. 493-498, Jun. 2002. V. Khomenko, E. Raymundo-Pinero, and F. Beguin, “A new type of high energy asymmetric capacitor with nanoporous carbon electrodes in aqueous electrolyte,” Journal of Power Sources, vol. 195, no. 13, pp. 4234-4241, Jul. 2010. Y. Zhang et al., “Progress of electrochemical capacitor electrode materials: A review,” International Journal of Hydrogen Energy, vol. 34, no. 11, pp. 48894899, Jun. 2009. A. Izadi-Najafabadi, “Nanotube and Polypyrrole Supercapacitors,” Master of Applied Science, University of British Columbia, 2006. A. Balducci et al., “High temperature carbon-carbon supercapacitor using ionic liquid as electrolyte,” Journal of Power Sources, vol. 165, no. 2, pp. 922-927, Mar. 2007. E. Frackowiak, “Carbon materials for supercapacitor application,” Phys. Chem. Chem. Phys., vol. 9, no. 15, pp. 1774-1785, 2007. “Separators.” [Online]. Available: http://www.gore.com/en_xx/products/electronic/battery/separators.html. [Accessed: 04-Mar-2011]. “AOWEI TECHNOLOGY.” [Online]. Available: http://www.aowei.com/english/case_news01.htm?type_id=182&id=82. [Accessed: 04-Mar-2011]. T. Hamilton, “Next Stop: Ultracapacitor Buses - Technology Review.” [Online]. Available: http://www.technologyreview.com/energy/23754/. [Accessed: 04-Mar2011]. “Ciralight.” [Online]. Available: http://www.ciralightglobal.com/. [Accessed: 04Mar-2011]. “Skyfuel.” [Online]. Available: http://www.skyfuel.ca/Product3.html. [Accessed: 04-Mar-2011]. J. D. Madden, “Conducting Polymer Actuators,” PhD, Massachusetts Institute of 64 [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] Technology, 2000. V. Khomenko, E. Frackowiak, and F. Beguin, “Determination of the specific capacitance of conducting polymer/nanotubes composite electrodes using different cell configurations,” Electrochimica Acta, vol. 50, no. 12, pp. 24992506, Apr. 2005. M. Cole, “Feasibility of Minature Polypyrrole Actuated Valves,” Master of Applied Science, University of British Columbia, 2006. Kohlman, R. S. and Epstein, A. J., “Insulator-metal transistion and inhomogeneous metallic state in conducting polymers,” in Handbook of Conducting Polymers, New York: Marcel Dekker, 1998, pp. 85-122. C. W. E. Fok, “Framework for the Characterization and Performance Projection of Electrochemical Capacitor Electrodes,” Master of Applied Science, University of British Columbia, 2005. M. Warren, “Electronic and Structural Effects on the Electrochemistry of Polypyrrole,” Master of Science, University of British Columbia, 2005. D. S. Maddison and J. Unsworth, “Optimization of synthesis conditions of polypyrrole from aqueous solutions,” Synthetic Metals, vol. 30, no. 1, pp. 47-55, Apr. 1989. C.-H. Tso, J. D. Madden, and C. A. Michal, “An NMR study of PF6- ions in polypyrrole,” Synthetic Metals, vol. 157, no. 10-12, pp. 460-466, Jun. 2007. J. D. W. Madden, P. D. W. Madden, and I. W. Hunter, “Characterization of polypyrrole actuators: modeling and performance,” Proceedings of SPIE 8th Annual Symposium on Smart Structures and Materials: Electroactive Polymer Actuators and Devices, pp. 72-83, Nov. 2001. M. J. Ariza and T. F. Otero, “Ionic diffusion across oxidized polypyrrole membranes and during oxidation of the free-standing film,” Colloids and Surfaces A: Physicochemical and Engineering Aspects, vol. 270-271, pp. 226231, Dec. 2005. W. Liang, J. Lei, and C. R. Martin, “Effect of synthesis temperature on the structure, doping level and charge-transport properties of polypyrrole,” Synthetic Metals, vol. 52, no. 2, pp. 227-239, Oct. 1992. M. Ue, K. Ida, and S. Mori, “Electrochemical Properties of Organic Liquid Electrolytes Based on Quaternary Onium Salts for Electrical Double-Layer Capacitors,” Journal of The Electrochemical Society, vol. 141, no. 11, pp. 29892996, Nov. 1994. M. Yamaura, T. Hagiwara, M. Hirasaka, T. Demura, and K. Iwata, “Structure and properties of biaxially stretched polypyrrole films,” Synthetic Metals, vol. 28, no. 1-2, pp. 157-164, Jan. 1989. A. Izadi-Najafabadi, D. T. H. Tan, and J. D. Madden, “Towards High Power Polypyrrole/Carbon Capacitors,” Synthetic Metals, vol. 152, no. 1-3, pp. 129-132, Sep. 2005. K. Jurewicz, S. Delpeux, V. Bertagna, F. Beguin, and E. Frackowiak, “Supercapacitors from nanotubes/polypyrrole composites,” Chemical Physics Letters, vol. 347, no. 1-3, pp. 36-40, Oct. 2001. J.-H. Kim, A. K. Sharma, and Y.-S. Lee, “Synthesis of polypyrrole and carbon nano-fiber composite for the electrode of electrochemical capacitors,” Materials 65 [33] [34] [35] [36] [37] Letters, vol. 60, no. 13-14, pp. 1697-1701, Jun. 2006. K. Lota, V. Khomenko, and E. Frackowiak, “Capacitance properties of poly(3,4ethylenedioxythiophene)/carbon nanotubes composites,” Journal of Physics and Chemistry of Solids, vol. 65, no. 2-3, pp. 295-301, Mar. 2004. C. Kim and S. Zhang, “Conductivity of carbon nanofiber/polypyrrole conducting nanocomposites,” Journal of Mechanical Science and Technology, vol. 23, no. 1, pp. 75-80, Jan. 2009. J. Jang, J. Bae, M. Choi, and S.-H. Yoon, “Fabrication and characterization of polyaniline coated carbon nanofiber for supercapacitor,” Carbon, vol. 43, no. 13, pp. 2730-2736, Nov. 2005. G. Che, B. B. Lakshmi, C. R. Martin, E. R. Fisher, and R. S. Ruoff, “Chemical Vapor Deposition Based Synthesis of Carbon Nanotubes and Nanofibers Using a Template Method,” Chemistry of Materials, vol. 10, no. 1, pp. 260-267, Jan. 1998. E. J. Ra, K. H. An, K. K. Kim, S. Y. Jeong, and Y. H. Lee, “Anisotropic electrical conductivity of MWCNT/PAN nanofiber paper,” Chemical Physics Letters, vol. 413, no. 1-3, pp. 188-193, Sep. 2005. 66 Appendix A Cyclic Voltammetry Measurements Without Filtering 67 Appendix B MATLAB Code %% plots of NF/PPy samples: vs. density ionic cond vs. density; tau close all; clear all; clc; cavrg=[0.38 0.29 1.63 2.57]; density=[1.67e-1 8.66e-1 9.56e-1 1.24]; density2=[1.5 8.66e-1 9.56e-1 1.24]; ppymass1=[0 2.6e-3 1.09e-2 1.53e-2]*1000; ppymass2=[0.0032 2.6e-3 1.09e-2 1.53e-2]*1000; thickness=[ 7.00E-04 2.75E-03 3.34E-03 5.64E-03]; condplus=[2.8e-7 5.3e-6 6.8e-6 9.0e-6]; condminus=[8.5e-7 1.2e-5 1.3e-5 1.5e-5]; cvol=[1.7e2 5.1e1 1.1e2 1.6e2]; tauplus=[3e2 7.2e1 1.8e2 2.5e2]; tauminus=[9.8e1 3.2e1 9.5e1 1.5e2]; figure; set(gca,'FontSize',16) errorbar(ppymass1,density,0.26*density,'r*','LineWidth',2,' MarkerSize',12);hold on; ylabel('Density (g/cm^3)','FontSize',16) xlabel('PPy mass (mg)','FontSize',16) title('Density vs PPy mass','FontSize',16) %legend('Data','Filtered Data','FontSize',16) figure; set(gca,'FontSize',16) errorbar(density2,condplus,0.15*condplus,'r*','LineWidth',2 ,'MarkerSize',12);hold on; errorbar(density2,condminus,0.15*condminus,'bs','LineWidth' ,2,'MarkerSize',12);hold on; ylabel('\kappa (S/cm)') xlabel('Density (g/cm^3)') title('Ionic Conductivity (\kappa) vs Density') legend('+ Scan','- Scan'); figure; set(gca,'FontSize',16) 68 errorbar(density2,tauplus,0.26*tauplus,'r*','LineWidth',2,' MarkerSize',12);hold on; errorbar(density2,tauminus,0.26*tauminus,'bs','LineWidth',2 ,'MarkerSize',12);hold on; ylabel('\tau (s)') xlabel('Density (g/cm^3)') title('Time Constant (\tau) vs Density') legend('+ Scan','- Scan'); 69 %sample 19.m = sample 4h close all; clear all; clc; vsample19=xlsread('sample19.xlsx','sample19','A212347:A2359 32'); isample19=xlsread('sample19.xlsx','sample19','B212347:B2359 32'); area=0.7*1.5; L=55e-4; effp=isample19(isample19>0); effn=isample19(isample19<0); qp=(sum(effp)/length(effp)/10); qn=(sum(effn)/length(effn)/10); eff=abs(qn)/qp; figure; set(gca,'FontSize',16); plot(vsample19,isample19);hold on; axis([ -0.85 0.65 -0.001 0.001]) ylabel('I(A)','FontSize') xlabel('V(V)','FontSize') title('CV at 1mV/s, Sample 10','FontSize') isample19=isample19/area; temp=100; a = 1; b = ones(1,temp)*1/temp; ifilter = filter(b,a,isample19); ifilterpos=ifilter(1:11784); ifilterneg=ifilter(11785:end); vpos=vsample19(1:11784); vneg=vsample19(11785:end); ppymass=0.00276; cond_electrodepos=5.3e-6; C_volpos=63.5; % volumetric capacitance in F/cm^3 L=55e-4; % thickness of electrode in cm taupos=C_volpos*L^2/cond_electrodepos cond_electrodeneg=12e-6; % solution conductivity in electrode according to Bruggeman's relation C_volneg=38; % volumetric capacitance in F/cm^3 tauneg=C_volneg*L^2/cond_electrodeneg 70 Cpos=C_volpos*area*L; Cneg=C_volneg*area*L; Cspec=[ Cpos/ppymass Cneg/ppymass]; k=1e-3; % voltage scan rate V/s Tmaxpos=2370/taupos; Tmaxneg=2390/tauneg; % cond_electrode*tau*k/L; T=[0:0.1:Tmaxpos]; % T=t/tau tmp=zeros(1,length(T)); gamma=1; for m=0:2000 Am=taupos*k*(32*(-1)^m/(pi*(2*m+1))^3); tmp=tmp-(Am*pi*(2*m+1)/2*exp(pi^2/4*(2*m+1)^2*T)*sin(pi/2*gamma*(2*m+1))); end j=cond_electrodepos/L*(taupos*k+tmp); % Negative Scan T2=[0:0.1:Tmaxneg]; % T=t/tau tmp2=zeros(1,length(T2)); gamma=1; for m=0:2000 Am=-1*tauneg*k*(32*(-1)^m/(pi*(2*m+1))^3); tmp2=tmp2-(Am*pi*(2*m+1)/2*exp(pi^2/4*(2*m+1)^2*T2)*sin(pi/2*gamma*(2*m+1))); end j2=cond_electrodeneg/L*(-tauneg*k+tmp2); % plot([T*tau (Tmax+T2)*tau],[j j2]) V1=[-0.785:(1.185/(10*2370/taupos)):0.4]; V2=[0.395:(-1.195/(10*2390/tauneg)):-0.8]; %% filtering window=3699; toend=length(ifilterpos)-window; ifilterposdiff=ifilterpos(toend:end)-j(end); ifilterneg(100:100+window)=ifilterneg(100:100+window)flipud(ifilterposdiff); vposdiff=vpos(toend:end); %% figure; plot(vsample19(100:end),ifilter(100:end),vneg(100:end),ifil terneg(100:end),'LineWidth',2);hold on; set(gca,'FontSize',24) plot([V1],[j],'-.r','LineWidth',2);hold on; 71 plot([V2],[j2],'-.k','LineWidth',2);hold on; axis([ -0.85 0.45 -0.0007 0.0007]) ylabel('J(A/cm^2)') xlabel('V(V)') legend('Filtered Data','Data with Compensation','+ Scan Fit','- Scan Fit'); title('CV at 1mV/s, Sample 4h') 72 %% Sample 16.m = Sample 12h in Thesis clear all;close all;clc; %cycle5v=xlsread('april6withcomparison.xlsx','secondcv','A6 883:A20630'); %cycle6v=xlsread('april6withcomparison.xlsx','secondcv','A2 0630:A34377'); %cycle7v=xlsread('april6withcomparison.xlsx','secondcv','A3 4377:A48124'); vsample16=xlsread('april6withcomparison.xlsx','secondcv','A 48100:A61871'); area=1.6*1.4; %cycle5i=xlsread('april6withcomparison.xlsx','secondcv','C6 883:C20630'); %cycle6i=xlsread('april6withcomparison.xlsx','secondcv','C2 0630:C34377'); %cycle7i=xlsread('april6withcomparison.xlsx','secondcv','C3 4377:C48124'); isample16=xlsread('april6withcomparison.xlsx','secondcv','C 48100:C61871'); effp=isample16(isample16>0); effn=isample16(isample16<0); qp=(sum(effp)/length(effp)/10); qn=(sum(effn)/length(effn)/10); eff=abs(qn)/qp; figure; plot(vsample16,isample16);hold on; axis([ -0.85 0.65 -0.003 0.003]) ylabel('I(A)','FontSize',16) xlabel('V(V)','FontSize',16) set(gca,'FontSize',16) title('CV at 1mV/s, Sample 8','FontSize',16) isample16=isample16/area; temp=100; a = 1; b = ones(1,temp)*1/temp; ifilter = filter(b,a,isample16); ifilterpos=ifilter(6876:end); ifilterneg=ifilter(1:6875); vpos=vsample16(6876:end); vneg=vsample16(1:6875); ppymass=0.0109; 73 %%% % defining physical parameters % cond_free=2e-4; % free solution conductivity, 0.65M TEAB (6.1e-3 for 1M TEAP) in PC from Ue et. al. Journal of Electrochemical Society, Vol. 141, No. 11, 1994 % cond_electrodepos=2.6e-7; % solution conductivity in electrode according to Bruggeman's relation cond_electrodepos=6.8e-6; C_volpos=113; % volumetric capacitance in F/cm^3 L=66.8e-4; % thickness of electrode in cm taupos=C_volpos*L^2/cond_electrodepos cond_electrodeneg=12.8e-6; % solution conductivity in electrode according to Bruggeman's relation C_volneg=104; % volumetric capacitance in F/cm^3 tauneg=C_volneg*L^2/cond_electrodeneg Cpos=C_volpos*area*L; Cneg=C_volneg*area*L; Cspec=[ Cpos/ppymass Cneg/ppymass]; % Positive Scan k=1e-3; % voltage scan rate V/s Tmaxpos=2760/taupos; Tmaxneg=2770/tauneg; % cond_electrode*tau*k/L; T=[0:0.1:Tmaxpos]; % T=t/tau tmp=zeros(1,length(T)); gamma=1; for m=0:2000 Am=taupos*k*(32*(-1)^m/(pi*(2*m+1))^3); tmp=tmp-(Am*pi*(2*m+1)/2*exp(pi^2/4*(2*m+1)^2*T)*sin(pi/2*gamma*(2*m+1))); end j=cond_electrodepos/L*(taupos*k+tmp); % Negative Scan T2=[0:0.1:Tmaxneg]; % T=t/tau tmp2=zeros(1,length(T2)); gamma=1; for m=0:2000 Am=-1*tauneg*k*(32*(-1)^m/(pi*(2*m+1))^3); tmp2=tmp2-(Am*pi*(2*m+1)/2*exp(pi^2/4*(2*m+1)^2*T2)*sin(pi/2*gamma*(2*m+1))); end 74 j2=cond_electrodeneg/L*(-tauneg*k+tmp2); % plot([T*tau (Tmax+T2)*tau],[j j2]) V1=[-0.78:(1.38/(10*2760/taupos)):0.6]; V2=[0.585:(-1.385/(10*2770/tauneg)):-0.8]; window=1299; toend=length(ifilterpos)-window; ifilterposdiff=ifilterpos(toend:end)-j(end); ifilterneg(100:100+window)=ifilterneg(100:100+window)flipud(ifilterposdiff); vposdiff=vpos(toend:end); %cv figure; plot(vsample16(100:end),ifilter(100:end),vneg(100:end),ifil terneg(100:end),'Linewidth',2);hold on; plot([V1],[j],'-.r');hold on; plot([V2],[j2],'-.k');hold on; axis([ -0.85 0.65 -0.003 0.003]) ylabel('J(A/cm^2)','FontSize',24) xlabel('V(V)','FontSize',24) % title('CV at 1mV/s, Sample 16','FontSize',16) legend('Filtered Data','Data with Compensation','+ Scan Fit','- Scan Fit','FontSize',24); set(gca,'FontSize',24) title('CV at 1mV/s, Sample 12h','FontSize',24) 75 %sample 23.m = Sample 16h in thesis close all;clear all; volt=0.001; area=1.5*1.45; vsample23=xlsread('sample23.xlsx','1stcv','A47200:A70780'); isample23=xlsread('sample23.xlsx','1stcv','B47200:B70780'); effp=isample23(isample23>0); effn=isample23(isample23<0); qp=(sum(effp)/length(effp)/10); qn=(sum(effn)/length(effn)/10); eff=abs(qn)/qp; figure; plot(vsample23,isample23);hold on; axis([ -0.85 0.45 -0.004 0.004]) ylabel('I(A)','FontSize',16) xlabel('V(V)','FontSize',16) set(gca,'FontSize',16) title('CV at 1mV/s, Sample 9','FontSize',16) isample23=isample23/area; ppymass=0.0153; %%% % defining physical parameters % cond_free=2e-4; % free solution conductivity, 0.65M TEAB (6.1e-3 for 1M TEAP) in PC from Ue et. al. Journal of Electrochemical Society, Vol. 141, No. 11, 1994 % cond_electrodepos=2.6e-7; % solution conductivity in electrode according to Bruggeman's relation cond_electrodepos=9e-6; % solution conductivity in electrode according to Bruggeman's relation C_volpos=173; % volumetric capacitance in F/cm^3 L=74.9e-4; % nf thickness of electrode in cm taupos=C_volpos*L^2/cond_electrodepos cond_electrodeneg=1.5e-5; % solution conductivity in electrode according to Bruggeman's relation C_volneg=142; % volumetric capacitance in F/cm^3 tauneg=C_volneg*L^2/cond_electrodeneg Cpos=C_volpos*area*L; Cneg=C_volneg*area*L; Cspec=[ Cpos/ppymass Cneg/ppymass]; temp=100; a = 1; 76 b = ones(1,temp)*1/temp; ifilter = filter(b,a,isample23); ifilterpos=ifilter(1:11800); ifilterneg=ifilter(11801:end); vpos=vsample23(1:11800); vneg=vsample23(11801:end); % Positive Scan k=1e-3; % voltage scan rate V/s Tmaxpos=2368/taupos; Tmaxneg=2374/tauneg; % cond_electrode*tau*k/L; T=[0:0.01:Tmaxpos]; % T=t/tau tmp=zeros(1,length(T)); gamma=1; for m=0:2000 Am=taupos*k*(32*(-1)^m/(pi*(2*m+1))^3); tmp=tmp-(Am*pi*(2*m+1)/2*exp(pi^2/4*(2*m+1)^2*T)*sin(pi/2*gamma*(2*m+1))); end j=cond_electrodepos/L*(taupos*k+tmp); cond_electrodepos*taupos*k/L; % Negative Scan T2=[0:0.01:Tmaxneg]; % T=t/tau tmp2=zeros(1,length(T2)); gamma=1; for m=0:2000 Am=-1*tauneg*k*(32*(-1)^m/(pi*(2*m+1))^3); tmp2=tmp2-(Am*pi*(2*m+1)/2*exp(pi^2/4*(2*m+1)^2*T2)*sin(pi/2*gamma*(2*m+1))); end j2=cond_electrodeneg/L*(-tauneg*k+tmp2); % plot([T*tau (Tmax+T2)*tau],[j j2]) V1=[-0.784:(1.184/(100*2368/taupos)):0.4]; V2=[0.387:(-1.187/(100*2374/tauneg)):-0.8]; window=999; toend=length(ifilterpos)-window; ifilterposdiff=ifilterpos(toend:end)-j(end); ifilterneg(200:200+window)=ifilterneg(200:200+window)flipud(ifilterposdiff); vposdiff=vpos(toend:end); figure; 77 plot(vsample23(100:end),ifilter(100:end),vneg,ifilterneg,'L inewidth',2);hold on; plot([V1],[j],'-.r');hold on; plot([V2],[j2],'-.k');hold on; axis([ -0.85 0.45 -0.003 0.003]) ylabel('J(A/cm^2)','FontSize',24) xlabel('V(V)','FontSize',24) title('CV at 1mV/s, Sample 16h','FontSize',24) legend('Filtered Data','Data with Compensation','+ Scan Fit','- Scan Fit','FontSize',24); set(gca,'FontSize',24) figure; plot(vsample23,isample23);hold on; axis([ -0.85 0.45 -0.003 0.003]) ylabel('J(A/cm2)','FontSize',16) xlabel('V(V)','FontSize',16) set(gca,'FontSize',16) title('CV at 1mV/s, Sample 10','FontSize',16) 78 % plots of ionic cond and tau and cap vs sample no. clear; clc; close all; cappos1mv=xlsread('results jan17.xlsx','april6','C16:C21'); capneg1mv=xlsread('results jan17.xlsx','april6','C24:C29'); cappos10mv=xlsread('results jan17.xlsx','april6','C53:C58'); capneg10mv=xlsread('results jan17.xlsx','april6','C61:C66'); kpos1mv=xlsread('results jan17.xlsx','april6','L69:L74'); kneg1mv=xlsread('results jan17.xlsx','april6','O69:O74'); tpos1mv=xlsread('results jan17.xlsx','april6','R69:R74'); tneg1mv=xlsread('results jan17.xlsx','april6','U69:U74'); kpos10mv=xlsread('results jan17.xlsx','april6','L77:L82'); kneg10mv=xlsread('results jan17.xlsx','april6','O77:O82'); tpos10mv=xlsread('results jan17.xlsx','april6','R77:R82'); tneg10mv=xlsread('results jan17.xlsx','april6','U77:U82'); density=[1.5 1.4 1.2 1.6 1.4 1.0]; %g/cm3 x=1:6; thickness=[7 10 9.8 9 9.5 18]; %in microns figure; errorbar(cappos1mv,0.05*cappos1mv,'r*','LineWidth',2,'Marke rsize',10);hold on; errorbar(capneg1mv,0.05*capneg1mv,'b*','LineWidth',2,'Marke rsize',10);hold on; errorbar(cappos10mv,0.05*cappos10mv,'ms','LineWidth',2,'Mar kersize',10);hold on; errorbar(capneg10mv,0.05*capneg10mv,'cs','LineWidth',2,'Mar kersize',10);hold on; set(gca,'FontSize',16); xlabel('Sample No.'); ylabel('Capacitance (F)'); legend('+ Scan 1 mV/s','- Scan 1 mV/s','+ Scan 10 mV/s','Scan 10 mV/s'); title('Capacitance vs. Sample No.'); figure; errorbar(kpos1mv,0.15*kpos1mv,'r*','LineWidth',2,'Markersiz e',10);hold on; errorbar(kneg1mv,0.15*kneg1mv,'b*','LineWidth',2,'Markersiz e',10);hold on; errorbar(kpos10mv,0.15*kpos10mv,'ms','LineWidth',2,'Markers ize',10);hold on; 79 errorbar(kneg10mv,0.15*kneg10mv,'cs','LineWidth',2,'Markers ize',10);hold on; set(gca,'FontSize',16); xlabel('Sample No.'); ylabel('\sigma (S/cm)','FontSize',16); legend('+ Scan 1 mV/s','- Scan 1 mV/s','+ Scan 10 mV/s','Scan 10 mV/s'); title('Ionic Conductivity (\sigma) vs. Sample No.'); figure; errorbar(tpos1mv,0.25*tpos1mv,'r*','LineWidth',2,'Markersiz e',10);hold on; errorbar(tneg1mv,0.25*tneg1mv,'b*','LineWidth',2,'Markersiz e',10);hold on; errorbar(tpos10mv,0.25*tpos10mv,'ms','LineWidth',2,'Markers ize',10);hold on; errorbar(tneg10mv,0.25*tneg10mv,'cs','LineWidth',2,'Markers ize',10);hold on; set(gca,'FontSize',16); xlabel('Sample No.'); ylabel('\tau (s)','FontSize',16); legend('+ Scan 1 mV/s','- Scan 1 mV/s','+ Scan 10 mV/s','Scan 10 mV/s'); title('Time Constant (\tau) vs. Sample No.'); %% sample 1.m clear all; close all; clc; volt=0.001; mass4h30=0.0032; vol4h30=pi()*1.25*(7.25e-4); vol4h30active=pi()*1*(7.25e-4); den4h30=mass4h30/vol4h30; mass4h30=den4h30*vol4h30active; %time4h30=xlsread('cv.xlsx','fourhourminusthirty1mv','D2752 0:D55052'); i4h30=xlsread('4h-30 cv.xlsx','1mv before','B27520:B55052'); v4h30=xlsread('4h-30 cv.xlsx','1mv before','A27520:A55052'); i4h302=xlsread('4h-30 cv.xlsx','1mv after','B27520:B55052'); 80 v4h302=xlsread('4h-30 cv.xlsx','1mv after','A27520:A55052'); %%%% looking for degradation %%%% figure; set(gca,'FontSize',16) plot(v4h30,i4h30,'b','LineWidth',2); hold on; plot(v4h302,i4h302,'g','LineWidth',2);hold on; %plot(v4h30(100:end),i4h30filter(100:end),'m'); hold on; axis([ -0.85 0.65 -0.001 0.001]) ylabel('I(A)') xlabel('V(V)') title('CV at 1 mV/s, Sample 1') legend('Beginning','End') %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% i4h30=i4h30/3.14; i4h302=i4h302/3.14; %%%% filtering %%%% temp=100; a = 1; b = ones(1,temp)*1/temp; i4h30filter = filter(b,a,i4h30); i4h30filterpos=i4h30filter(1:13785); i4h30filterneg=i4h30filter(13786:27533); v4h30pos=v4h30(1:13785); v4h30neg=v4h30(13786:27533); %%%%%%%%%%%%%%%%%%% % defining physical parameters % cond_free=2e-4; % free solution conductivity, 0.65M TEAB (6.1e-3 for 1M TEAP) in PC from Ue et. al. Journal of Electrochemical Society, Vol. 141, No. 11, 1994 cond_electrodepos=2.8e-7; % solution conductivity in electrode according to Bruggeman's relation C_volpos=200; % volumetric capacitance in F/cm^3 L=7e-4; % thickness of electrode in cm taupos=C_volpos*L^2/cond_electrodepos; cond_electrodeneg=8.5e-7; % solution conductivity in electrode according to Bruggeman's relation C_volneg=140; % volumetric capacitance in F/cm^3 tauneg=C_volneg*L^2/cond_electrodeneg; Rpos=L/(cond_electrodepos*3.14); 81 Rneg=L/(cond_electrodeneg*3.14); Cpos=C_volpos*L*3.14; Cneg=C_volneg*L*3.14; % Positive Scan k=1e-3; % voltage scan rate V/s Tmaxpos=2766/taupos; Tmaxneg=2784/tauneg; % cond_electrode*tau*k/L; T=[0:0.1:Tmaxpos]; % T=t/tau tmp1=zeros(1,length(T)); Am1=zeros(1,length(T)); gamma=1; for m=0:20000 Am1=taupos*k*(32*(-1)^m/(pi*(2*m+1))^3); tmp1=tmp1-(Am1*pi*(2*m+1)/2*exp(pi^2/4*(2*m+1)^2*T)*sin(pi/2*gamma*(2*m+1))); end j=cond_electrodepos/L*(taupos*k+tmp1); % Negative Scan T2=[0:0.1:Tmaxneg]; % T=t/tau tmp2=zeros(1,length(T2)); for m=0:20000 Am2=-1*tauneg*k*(32*(-1)^m/(pi*(2*m+1))^3); tmp2=tmp2-(Am2*pi*(2*m+1)/2*exp(pi^2/4*(2*m+1)^2*T2)*sin(pi/2*(2*m+1))); end; j2=cond_electrodeneg/L*(-tauneg*k+tmp2); % plot([T*tau (Tmax+T2)*tau],[j j2]) V1=[-0.783:(1.383/(10*2766/taupos)):0.6]; V2=[0.592:(-1.392/(10*2784/tauneg)):-0.8]; %%%% compensation at +0.6V %%%% window=3999; toend=length(i4h30filterpos)-window; i4h30filterposdiff=i4h30filterpos(toend:end)-j(end); i4h30filterneg(200:200+window)=i4h30filterneg(200:200+windo w)-flipud(i4h30filterposdiff); v4h30posdiff=v4h30pos(toend:end); mean(i4h30filterposdiff) toend2=length(i4h30filterneg)-window; i4h30filternegdiff=i4h30filterneg(toend2:end)-j2(end); 82 %i4h30filterneg(200:200+window)=i4h30filterneg(200:200+wind ow)-flipud(i4h30filterposdiff); v4h30negdiff=v4h30neg(toend2:end); mean(i4h30filternegdiff) %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% figure; plot(v4h30(100:end),i4h30filter(100:end),v4h30neg,i4h30filt erneg,'Linewidth',2);hold on; plot([V1],[j],'-.r','Linewidth',2);hold on; plot([V2],[j2],'-.k','Linewidth',2);hold on; set(gca,'FontSize',25) axis([ -0.85 0.65 -0.0004 0.0004]) ylabel('J(A/cm^2)') xlabel('V(V)') title('CV at 1mV/s, Sample 1') %title('CV at 1mV/s, -30 degrees 0.125mA/cm^2','FontSize',16) legend('Filtered Data','Data with Compensation','+ Scan Fit','- Scan Fit'); % + scan fit 2.35e-7S/cm 200F/cm3 % - scan Fit 9e-7S/cm 140F/cm3 eff4h30p=i4h30(i4h30>0); eff4h30n=i4h30(i4h30<0); q4h30p=(sum(eff4h30p)/length(eff4h30p)/10); q4h30n=(sum(eff4h30n)/length(eff4h30n)/10); eff4h30=abs(q4h30n)/q4h30p; eff4h30p2=i4h302(i4h302>0); eff4h30n2=i4h302(i4h302<0); q4h30p2=(sum(eff4h30p2)/length(eff4h30p2)/10); q4h30n2=(sum(eff4h30n2)/length(eff4h30n2)/10); eff4h302=abs(q4h30n2)/q4h30p2; 83 %sample 2.m clear all; close all; clc; volt=0.001; %time2h30=xlsread('cv.xlsx','twohourminusthirty1mv','D27537 :D55052'); %time1h30=xlsread('cv.xlsx','onehourminusthirty1mv','D210:D 55052'); i2h30=xlsread('cv.xlsx','twohourminusthirty1mv','B27537:B55 052'); v2h30=xlsread('cv.xlsx','twohourminusthirty1mv','A27537:A55 052'); i2h302=xlsread('2h-30 cv.xlsx','2h-30 1mv2','B27500:B55052'); v2h302=xlsread('2h-30 cv.xlsx','2h-30 1mv2','A27500:A55052'); %compare the effects of cycling figure; set(gca,'FontSize',16) plot(v2h30,i2h30,'b','LineWidth',2); hold on; plot(v2h302,i2h302,'g','LineWidth',2);hold on; axis([ -0.85 0.65 -0.001 0.001]) ylabel('I(A)','FontSize',16) xlabel('V(V)','FontSize',16) title('CV at 1mV/s, Sample 2') legend('Beginning','End') i2h30=i2h30/3.14; temp=100; a = 1; b = ones(1,temp)*1/temp; i2h30filter = filter(b,a,i2h30); i2h30filterpos=i2h30filter(1:13675); i2h30filterneg=i2h30filter(13676:27516); v2h30pos=v2h30(1:13675); v2h30neg=v2h30(13676:27516); %%% % defining physical parameters 84 % cond_free=2e-4; % free solution conductivity, 0.65M TEAB (6.1e-3 for 1M TEAP) in PC from Ue et. al. Journal of Electrochemical Society, Vol. 141, No. 11, 1994 cond_electrodepos=6e-7; % solution conductivity in electrode according to Bruggeman's relation C_volpos=147 ; % volumetric capacitance in F/cm^3 L=10e-4; % thickness of electrode in cm taupos=C_volpos*L^2/cond_electrodepos cond_electrodeneg=9e-7; % solution conductivity in electrode according to Bruggeman's relation C_volneg=107; % volumetric capacitance in F/cm^3 tauneg=C_volneg*L^2/cond_electrodeneg; Rpos=L/(cond_electrodepos*3.14); Rneg=L/(cond_electrodeneg*3.14); Cpos=C_volpos*L*3.14; Cneg=C_volneg*L*3.14; % Positive Scan k=1e-3; % voltage scan rate V/s Tmaxpos=2773/taupos; Tmaxneg=2792/tauneg; % cond_electrode*tau*k/L; T=[0:0.1:Tmaxpos]; % T=t/tau tmp=zeros(1,length(T)); gamma=1; for m=0:2000 Am=taupos*k*(32*(-1)^m/(pi*(2*m+1))^3); tmp=tmp-(Am*pi*(2*m+1)/2*exp(pi^2/4*(2*m+1)^2*T)*sin(pi/2*gamma*(2*m+1))); end j=cond_electrodepos/L*(taupos*k+tmp); % Negative Scan T2=[0:0.1:Tmaxneg]; % T=t/tau tmp2=zeros(1,length(T2)); %tauneg gamma=1; for m=0:2000 Am2=-1*tauneg*k*(32*(-1)^m/(pi*(2*m+1))^3); tmp2=tmp2-(Am2*pi*(2*m+1)/2*exp(pi^2/4*(2*m+1)^2*T2)*sin(pi/2*gamma*(2*m+1))); end j2=cond_electrodeneg/L*(-tauneg*k+tmp2); 85 % plot([T*tau (Tmax+T2)*tau],[j j2]) V1=[-0.786:(1.386/(10*2772/taupos)):0.6]; V2=[0.596:(-1.396/(10*2792/tauneg)):-0.8]; window=5090; toend=length(i2h30filterpos)-window; i2h30filterposdiff=i2h30filterpos(toend:end)-j(end); i2h30filterneg(200:200+window)=i2h30filterneg(200:200+windo w)-flipud(i2h30filterposdiff); %cv figure; plot(v2h30(100:end),i2h30filter(100:end),v2h30neg,i2h30filt erneg,'Linewidth',2);hold on; plot([V1],[j],'-.r','Linewidth',2);hold on; plot([V2], [j2],'-.k','Linewidth',2);hold on; set(gca,'FontSize',25); legend('Filtered Data','Data with Compensation','+ Scan Fit','- Scan Fit','FontSize'); axis([ -0.85 0.65 -0.0004 0.0004]) ylabel('J(A/cm^2)') xlabel('V(V)') title('CV at 1mV/s, Sample 2') eff2h30p=i2h30(i2h30>0); eff2h30n=i2h30(i2h30<0); q2h30p=(sum(eff2h30p)*13692/10); q2h30n=(sum(eff2h30n)*13820/10); eff2h30=abs(q2h30n)/q2h30p; 86 %sample 3.m clear all; close all; clc; volt=0.001; mass1h30=0.0038; % vol1h30=pi()*1.25*(9.833e-4); % vol1h30active=pi()*1*(9.833e-4); % den1h30=mass1h30/vol1h30; % mass1h30=den1h30*vol1h30active; time1h30=xlsread('cv.xlsx','onehourminusthirty1mv','D27520: D55052'); i1h30=xlsread('cv.xlsx','onehourminusthirty1mv','B27520:B55 052'); i1h30=i1h30/3.14; v1h30=xlsread('cv.xlsx','onehourminusthirty1mv','A27520:A55 052'); temp=100; a = 1; b = ones(1,temp)*1/temp; i1h30filter = filter(b,a,i1h30); i1h30filterpos=i1h30filter(1:13786); i1h30filterneg=i1h30filter(13787:27302); v1h30pos=v1h30(1:13786); v1h30neg=v1h30(13787:27302); %%% % defining physical parameters % cond_free=2e-4; % free solution conductivity, 0.65M TEAB (6.1e-3 for 1M TEAP) in PC from Ue et. al. Journal of Electrochemical Society, Vol. 141, No. 11, 1994 %cond_electrodepos=4e-7; % solution conductivity in electrode according to Bruggeman's relation cond_electrodepos=4.5e-7; % solution conductivity in electrode according to Bruggeman's relation C_volpos=125; % volumetric capacitance in F/cm^3 L=10e-4; % thickness of electrode in cm taupos=C_volpos*L^2/cond_electrodepos; cond_electrodeneg2=11e-7; % solution conductivity in electrode according to Bruggeman's relation 87 C_volneg2=110; % volumetric capacitance in F/cm^3 tauneg2=C_volneg2*L^2/cond_electrodeneg2; Rpos=L/(cond_electrodepos*3.14); Rneg=L/(cond_electrodeneg2*3.14); Cpos=C_volpos*L*3.14; Cneg2=C_volneg2*L*3.14; % Positive Scan k=1e-3; % voltage scan rate V/s Tmaxpos=2772/taupos; Tmaxneg2=2790/tauneg2; T=[0:0.1:Tmaxpos]; % T=t/tau tmp=zeros(1,length(T)); gamma=1; for m=0:2800 Am=taupos*k*(32*(-1)^m/(pi*(2*m+1))^3); tmp=tmp-(Am*pi*(2*m+1)/2*exp(pi^2/4*(2*m+1)^2*T)*sin(pi/2*gamma*(2*m+1))); end j=cond_electrodepos/L*(taupos*k+tmp); % Negative Scan T3=[0:0.1:Tmaxneg2]; % T=t/tau tmp3=zeros(1,length(T3)); %tauneg2 gamma=1; for m=0:2800 Am3=-1*tauneg2*k*(32*(-1)^m/(pi*(2*m+1))^3); tmp3=tmp3-(Am3*pi*(2*m+1)/2*exp(pi^2/4*(2*m+1)^2*T3)*sin(pi/2*gamma*(2*m+1))); end j3=cond_electrodeneg2/L*(-tauneg2*k+tmp3); V1=[-0.786:(1.386/(10*2772/taupos)):0.6]; V3=[0.595:(-1.395/(10*2790/tauneg2)):-0.8]; %cv window=6990; toend=length(i1h30filterpos)-window; i1h30filterposdiff=i1h30filterpos(toend:end)-j(end); i1h30filterneg(200:200+window)=i1h30filterneg(200:200+windo w)-flipud(i1h30filterposdiff); 88 plot(v1h30(80:end),i1h30filter(80:end),v1h30neg,i1h30filter neg,'Linewidth',2);hold on; plot([V1],[j],'-.r','Linewidth',2);hold on; plot([V3],[j3],'-.k','Linewidth',2);hold on; set(gca,'FontSize',25); axis([ -0.85 0.65 -0.0004 0.0004]) ylabel('J(A/cm^2)') xlabel('V(V)') title('CV at 1mV/s, Sample 3') %title('CV at 1mV/s, -30 degrees 0.5mA/cm^2') legend('Filtered Data','Data with Compensation','+ Scan Fit','- Scan Fit'); eff1h30p=i1h30(i1h30>0); eff1h30n=i1h30(i1h30<0); q1h30p=(sum(eff1h30p)/length(eff1h30p)/10); q1h30n=(sum(eff1h30n)/length(eff1h30n)/10); eff1h30=abs(q1h30n)/q1h30p; % eff1h30=abs(q1h30n)/q1h30p 89 % Sample 4 clear all; close all; clc; volt=0.001; % mass4hroom=0.0044; % vol4hroom=pi()*1.25*(9.125e-4); % vol4hroomactive=pi()*1*(9.125e-4); % den4hroom=mass4hroom/vol4hroom; % mass4hroom=den4hroom*vol4hroomactive; i4hroom=xlsread('cv.xlsx','fourhourroom1mv','B13770:B41280' ); v4hroom=xlsread('cv.xlsx','fourhourroom1mv','A13770:A41280' ); i4hroom=i4hroom/3.14; time4hroom=xlsread('cv.xlsx','fourhourroom1mv','D13770:D412 80'); temp=100; a = 1; b = ones(1,temp)*1/temp; ifilter = filter(b,a,i4hroom); ifilterpos=ifilter(13754:end); ifilterneg=ifilter(1:13753); vpos=v4hroom(13754:end); vneg=v4hroom(1:13753); %%% % defining physical parameters % cond_free=2e-4; % free solution conductivity, 0.65M TEAB (6.1e-3 for 1M TEAP) in PC from Ue et. al. Journal of Electrochemical Society, Vol. 141, No. 11, 1994 cond_electrodepos=3.5e-7; % solution conductivity in electrode according to Bruggeman's relation C_volpos=175; % volumetric capacitance in F/cm^3 L=9e-4; % thickness of electrode in cm taupos=C_volpos*L^2/cond_electrodepos cond_electrodeneg=10e-7; % solution conductivity in electrode according to Bruggeman's relation C_volneg=175; % volumetric capacitance in F/cm^3 tauneg=C_volneg*L^2/cond_electrodeneg; 90 cond_electrodeneg2=10e-7; % solution conductivity in electrode according to Bruggeman's relation C_volneg2=125; % volumetric capacitance in F/cm^3 tauneg2=C_volneg2*L^2/cond_electrodeneg2; Rpos=L/(cond_electrodepos*3.14); Rneg2=L/(cond_electrodeneg2*3.14); Cpos=C_volpos*L*3.14; Cneg2=C_volneg2*L*3.14; % Positive Scan k=1e-3; % voltage scan rate V/s Tmaxpos=2784/taupos; Tmaxneg=2800/tauneg; Tmaxneg2=2784/tauneg2; % cond_electrode*tau*k/L; T=[0:0.1:Tmaxpos]; % T=t/tau tmp=zeros(1,length(T)); gamma=1; for m=0:2000 Am=taupos*k*(32*(-1)^m/(pi*(2*m+1))^3); tmp=tmp-(Am*pi*(2*m+1)/2*exp(pi^2/4*(2*m+1)^2*T)*sin(pi/2*gamma*(2*m+1))); end j=cond_electrodepos/L*(taupos*k+tmp); % Negative Scan T2=[0:0.1:Tmaxneg]; % T=t/tau T3=[0:0.1:Tmaxneg2]; tmp2=zeros(1,length(T2)); tmp3=zeros(1,length(T3)); gamma=1; for m=0:2000 Am=-1*tauneg*k*(32*(-1)^m/(pi*(2*m+1))^3); tmp2=tmp2-(Am*pi*(2*m+1)/2*exp(pi^2/4*(2*m+1)^2*T2)*sin(pi/2*gamma*(2*m+1))); Am3=-1*tauneg2*k*(32*(-1)^m/(pi*(2*m+1))^3); tmp3=tmp3-(Am3*pi*(2*m+1)/2*exp(pi^2/4*(2*m+1)^2*T3)*sin(pi/2*(2*m+1))); end j3=cond_electrodeneg2/L*(-tauneg2*k+tmp3); 91 % plot([T*tau (Tmax+T2)*tau],[j j2]) V1=[-0.792:(1.392/(10*2784/taupos)):0.6]; V3=[0.592:(-1.392/(10*2784/tauneg2)):-0.8]; window=4999; toend=length(ifilterpos)-window; ifilterposdiff=ifilterpos(toend:end)-j(end); ifilterneg(200:200+window)=ifilterneg(200:200+window)flipud(ifilterposdiff); vposdiff=vpos(toend:end); %cv figure; plot(v4hroom(100:end),ifilter(100:end),vneg(50:end),ifilter neg(50:end),'Linewidth',2);hold on; set(gca,'FontSize',25); plot([V1],[j],'-.r','Linewidth',2);hold on; plot([V3],[j3],'-.k','Linewidth',2);hold on; axis([ -0.85 0.65 -0.0004 0.0004]) ylabel('J(A/cm^2)') xlabel('V(V)') title('CV at 1mV/s, Sample 4') %title('CV at 1mV/s, +25 Degrees 0.125mA/cm^2','FontSize',16) legend('Filtered Data','Data with Compensation','+ Scan Fit','- Scan Fit'); eff4hroomp=i4hroom(i4hroom>0); eff4hroomn=i4hroom(i4hroom<0); q4hroomp=(sum(eff4hroomp)/length(eff4hroomp)/10); q4hroomn=(sum(eff4hroomn)/length(eff4hroomn)/10); eff4hroom=abs(q4hroomn)/q4hroomp; 92 %Sample5 clear all; close all; clc; volt=0.001; % mass2hroom=0.0041; % vol2hroom=pi()*1.25*(9.66e-4); % vol2hroomactive=pi()*1*(9.66e-4); % den2hroom=mass2hroom/vol2hroom; % mass2hroom=den2hroom*vol2hroomactive; time2hroom=xlsread('cv.xlsx','twohourroom1mv','D27520:D5505 2'); i2hroom=xlsread('cv.xlsx','twohourroom1mv','B27520:B55052') ; % i2hroompavrg=xlsread('cv.xlsx','twohourroom1mv','B7735:B798 5'); % i2hroomnavrg=xlsread('cv.xlsx','twohourroom1mv','B19525:B19 775'); % i2hroomnaverage=sum(i2hroomnavrg)/251 % i2hroompaverage=sum(i2hroompavrg)/251 v2hroom=xlsread('cv.xlsx','twohourroom1mv','A27520:A55052') ; i2hroom=i2hroom/3.14; temp=100; a = 1; b = ones(1,temp)*1/temp; ifilter = filter(b,a,i2hroom); ifilterpos=ifilter(1:13775); ifilterneg=ifilter(13776:end); vpos=v2hroom(1:13775); vneg=v2hroom(13776:end); %%% % defining physical parameters % cond_free=2e-4; % free solution conductivity, 0.65M TEAB (6.1e-3 for 1M TEAP) in PC from Ue et. al. Journal of Electrochemical Society, Vol. 141, No. 11, 1994 93 cond_electrodepos=3.5e-7; % solution conductivity in electrode according to Bruggeman's relation C_volpos=185; % volumetric capacitance in F/cm^3 L=9.5e-4; % thickness of electrode in cm taupos=C_volpos*L^2/cond_electrodepos cond_electrodeneg=9.5e-7; % solution conductivity in electrode according to Bruggeman's relation C_volneg=140; % volumetric capacitance in F/cm^3 tauneg=C_volneg*L^2/cond_electrodeneg Rpos=L/(cond_electrodepos*3.14); Rneg=L/(cond_electrodeneg*3.14); Cpos=C_volpos*L*3.14; Cneg=C_volneg*L*3.14; % Positive Scan k=1e-3; % voltage scan rate V/s Tmaxpos=2740/taupos; Tmaxneg=2786/tauneg; % cond_electrode*tau*k/L; T=[0:0.1:Tmaxpos]; % T=t/tau tmp=zeros(1,length(T)); gamma=1; for m=0:2000 Am=taupos*k*(32*(-1)^m/(pi*(2*m+1))^3); tmp=tmp-(Am*pi*(2*m+1)/2*exp(pi^2/4*(2*m+1)^2*T)*sin(pi/2*gamma*(2*m+1))); end j=cond_electrodepos/L*(taupos*k+tmp); % Negative Scan T2=[0:0.1:Tmaxneg]; % T=t/tau tmp2=zeros(1,length(T2)); gamma=1; for m=0:2000 Am=-1*tauneg*k*(32*(-1)^m/(pi*(2*m+1))^3); tmp2=tmp2-(Am*pi*(2*m+1)/2*exp(pi^2/4*(2*m+1)^2*T2)*sin(pi/2*gamma*(2*m+1))); end j2=cond_electrodeneg/L*(-tauneg*k+tmp2); % plot([T*tau (Tmax+T2)*tau],[j j2]) V1=[-0.77:(1.37/(10*2740/taupos)):0.6]; V2=[0.593:(-1.393/(10*2786/tauneg)):-0.8]; 94 window=5299; toend=length(ifilterpos)-window; ifilterposdiff=ifilterpos(toend:end)-j(end); ifilterneg(200:200+window)=ifilterneg(200:200+window)flipud(ifilterposdiff); vposdiff=vpos(toend:end); figure; plot(v2hroom(100:end),ifilter(100:end),vneg,ifilterneg,'Lin ewidth',2);hold on; plot([V1],[j],'-.r','Linewidth',2);hold on; plot([V2],[j2],'-.k','Linewidth',2);hold on; set(gca,'FontSize',25); axis([ -0.85 0.65 -0.0004 0.0004]); ylabel('J(A/cm^2)'); xlabel('V(V)'); title('CV at 1mV/s, Sample 5') %title('CV at 1mV/s, +25 Degrees 0.25mA/cm^2') legend('Filtered Data','Data with Compensation','+ Scan Fit','- Scan Fit'); %cv eff2hroomp=i2hroom(i2hroom>0); eff2hroomn=i2hroom(i2hroom<0); q2hroomp=(sum(eff2hroomp)/length(eff2hroomp)/10); q2hroomn=(sum(eff2hroomn)/length(eff2hroomn)/10); eff2hroom=abs(q2hroomn)/q2hroomp 95 % Sample 6 clear all; close all; clc; volt=0.001; % % % % % mass1hroom=0.0049; vol1hroom=pi()*1.25*(18.125e-4); vol1hroomactive=pi()*1*(18.125e-4); den1hroom=mass1hroom/vol1hroom; mass1hroom=den1hroom*vol1hroomactive; time1hroom=xlsread('cv.xlsx','onehourroom1mv','D27528:D5505 2'); i1hroom=xlsread('cv.xlsx','onehourroom1mv','B27528:B55052') ; v1hroom=xlsread('cv.xlsx','onehourroom1mv','A27528:A55052') ; i1hroom=i1hroom/3.14; temp=100; a = 1; b = ones(1,temp)*1/temp; ifilter = filter(b,a,i1hroom); ifilterpos=ifilter(1:13772); ifilterneg=ifilter(13773:end); vpos=v1hroom(1:13772); vneg=v1hroom(13773:end); %%% % defining physical parameters % cond_free=2e-4; % free solution conductivity, 0.65M TEAB (6.1e-3 for 1M TEAP) in PC from Ue et. al. Journal of Electrochemical Society, Vol. 141, No. 11, 1994 cond_electrodepos=2.2e-7; % solution conductivity in electrode according to Bruggeman's relation C_volpos=95; % volumetric capacitance in F/cm^3 L=18e-4; % thickness of electrode in cm taupos=C_volpos*L^2/cond_electrodepos; cond_electrodeneg=10e-7; % solution conductivity in electrode according to Bruggeman's relation C_volneg=40; % volumetric capacitance in F/cm^3 96 tauneg=C_volneg*L^2/cond_electrodeneg; Rpos=L/(cond_electrodepos*3.14); Rneg=L/(cond_electrodeneg*3.14); Cpos=C_volpos*L*3.14 Cneg=C_volneg*L*3.14 % Positive Scan k=1e-3; % voltage scan rate V/s Tmaxpos=2756/taupos; Tmaxneg=2766/tauneg; % cond_electrode*tau*k/L; T=[0:0.1:Tmaxpos]; % T=t/tau tmp=zeros(1,length(T)); gamma=1; for m=0:2000 Am=taupos*k*(32*(-1)^m/(pi*(2*m+1))^3); tmp=tmp-(Am*pi*(2*m+1)/2*exp(pi^2/4*(2*m+1)^2*T)*sin(pi/2*gamma*(2*m+1))); end j=cond_electrodepos/L*(taupos*k+tmp); % Negative Scan T2=[0:0.1:Tmaxneg]; % T=t/tau tmp2=zeros(1,length(T2)); gamma=1; for m=0:2000 Am=-1*tauneg*k*(32*(-1)^m/(pi*(2*m+1))^3); tmp2=tmp2-(Am*pi*(2*m+1)/2*exp(pi^2/4*(2*m+1)^2*T2)*sin(pi/2*gamma*(2*m+1))); end j2=cond_electrodeneg/L*(-tauneg*k+tmp2); % plot([T*tau (Tmax+T2)*tau],[j j2]) V1=[-0.778:(1.378/(10*2756/taupos)):0.6]; V2=[0.583:(-1.383/(10*2766/tauneg)):-0.8]; window=4399; toend=length(ifilterpos)-window; ifilterposdiff=ifilterpos(toend:end)-j(end); ifilterneg(200:200+window)=ifilterneg(200:200+window)flipud(ifilterposdiff); vposdiff=vpos(toend:end); figure; 97 plot(v1hroom(100:end),ifilter(100:end),vneg,ifilterneg,'Lin ewidth',2);hold on; plot([V1],[j],'-.r','Linewidth',2);hold on; plot([V2],[j2],'-.k','Linewidth',2);hold on; set(gca,'FontSize',25); axis([ -0.85 0.65 -0.0004 0.0004]); ylabel('J(A/cm^2)') xlabel('V(V)') title('CV at 1mV/s, Sample 6') legend('Filtered Data','Data with Compensation','+ Scan Fit','- Scan Fit'); eff1hroomp=i1hroom(i1hroom>0); eff1hroomn=i1hroom(i1hroom<0); q1hroomp=(sum(eff1hroomp)/length(eff1hroomp)/10); q1hroomn=(sum(eff1hroomn)/length(eff1hroomn)/10); eff1hroom=abs(q1hroomn)/q1hroomp 98
- Library Home /
- Search Collections /
- Open Collections /
- Browse Collections /
- UBC Theses and Dissertations /
- Influences of growth conditions and porosity on polypyrrole...
Open Collections
UBC Theses and Dissertations
Featured Collection
UBC Theses and Dissertations
Influences of growth conditions and porosity on polypyrrole supercapacitor electrode performance Lam, Joanna Wing Yu 2011
pdf
Page Metadata
Item Metadata
Title | Influences of growth conditions and porosity on polypyrrole supercapacitor electrode performance |
Creator |
Lam, Joanna Wing Yu |
Publisher | University of British Columbia |
Date Issued | 2011 |
Description | Supercapacitors are electrical energy storage devices that offer high power density and high energy density. The current energy density of supercapacitors is, however, not sufficient to meet the requirements of many applications. By using polypyrrole (PPy) as an alternate electrode material, the energy density of supercapacitors can be increased. Approaches for simplifying and speeding the production of PPy electrodes are investigated, as are means of increasing power density. The tradeoffs in performance of PPy are investigated when electrochemical deposition conditions – current density and temperature -- are modified to reduce costs. Although the surface morphology changes according to deposition conditions, PPy’s performance in capacitance and charging time is not greatly affected. The best electrode performance is obtained using electrodeposition conditions in which a current density of 0.125 mA/cm² is used and the temperature is held at -30°C. Higher temperatures and faster deposition rates can lead to more voluminous films which are lower in density, volumetric and specific capacitances. Further work is needed to investigate the impact of growth conditions on cycle and shelf life. To decrease the charging time of PPy the hypothesis is that additional porosity will help by creating channels of high ionic mobility. The porosity is achieved by polymerizing onto carbonized polyacrylonitrile nanofibres (NF). PPy-coated NF samples with a density of 1.2 g/cm³ exhibit similar volumetric (160 F/cm³) and specific capacitances (130 F/g) similar to that of pure PPy. The use of NF can increase the apparent ionic conductivities of PPy, allowing NF/PPy samples to charge just as quickly as pure polypyrrole electrodes that are four times less capacitive. However, based on the current model, the advantages of increasing porosity should be more dramatic, suggesting that other mechanisms such as uncompensated resistances and ion depletion may also influence charging time. As such, further work on NF/PPy is needed to determine and hopefully to mitigate the effects of such mechanisms. |
Genre |
Thesis/Dissertation |
Type |
Text |
Language | eng |
Date Available | 2011-04-26 |
Provider | Vancouver : University of British Columbia Library |
Rights | Attribution-NonCommercial-NoDerivatives 4.0 International |
DOI | 10.14288/1.0103217 |
URI | http://hdl.handle.net/2429/34003 |
Degree |
Master of Applied Science - MASc |
Program |
Electrical and Computer Engineering |
Affiliation |
Applied Science, Faculty of Electrical and Computer Engineering, Department of |
Degree Grantor | University of British Columbia |
Graduation Date | 2011-05 |
Campus |
UBCV |
Scholarly Level | Graduate |
Rights URI | http://creativecommons.org/licenses/by-nc-nd/4.0/ |
Aggregated Source Repository | DSpace |
Download
- Media
- 24-ubc_spring_2011_lam_joanna.pdf [ 3.26MB ]
- Metadata
- JSON: 24-1.0103217.json
- JSON-LD: 24-1.0103217-ld.json
- RDF/XML (Pretty): 24-1.0103217-rdf.xml
- RDF/JSON: 24-1.0103217-rdf.json
- Turtle: 24-1.0103217-turtle.txt
- N-Triples: 24-1.0103217-rdf-ntriples.txt
- Original Record: 24-1.0103217-source.json
- Full Text
- 24-1.0103217-fulltext.txt
- Citation
- 24-1.0103217.ris
Full Text
Cite
Citation Scheme:
Usage Statistics
Share
Embed
Customize your widget with the following options, then copy and paste the code below into the HTML
of your page to embed this item in your website.
<div id="ubcOpenCollectionsWidgetDisplay">
<script id="ubcOpenCollectionsWidget"
src="{[{embed.src}]}"
data-item="{[{embed.item}]}"
data-collection="{[{embed.collection}]}"
data-metadata="{[{embed.showMetadata}]}"
data-width="{[{embed.width}]}"
async >
</script>
</div>
Our image viewer uses the IIIF 2.0 standard.
To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0103217/manifest