Development and Characterization of Glassy Carbon Electrodes for a Bipolar Electrochemical Double Layer Capacitor

Dissertation / Doktorarbeit von Artur Braun

Abstract:

The thermal gasphase oxidation (activation) of Glassy Carbon (GC) and bipolar monolithic electrodes as prepared from thermally activated GC were investigated. The goal of these investigations was to understand the thermal activation process and to find a correlation between process parameters plus GC materials properties and the electrode performance. With this knowledge it should be possible to optimize the performance of GC electrodes.

Thermal activation creates a porous film on the surface of GC. With proceeding activation the film grows into the GC.

The film thickness of various GC plates and sheets was determined with scanning electron microscopy.

The microstructure and mesostructure of the samples were investigated with Xray diffraction and Small Angle X-ray scattering.

The internal surface area of activated GC samples were determined with gas adsorption measurements and Small Angle X-ray scattering.

The electrochemical double layer capacitance and the diffusion resistance of GC samples were determined with electrochemical impedance spectroscopy.

Studies on GC with different temperature of pyrolysis reveal a correlation between pyrolysis temperature and asymptotic active film thickness: The higher the pyrolysis temperature, the smaller the asymptotic film thickness.

X-ray diffraction on GC with various pyrolysis temperatures exhibits a correlation between temperature of pyrolysis as well as intensity and full width at half maximum of prominent diffraction peaks: The higher the pyrolysis temperature, the smaller the defect density in GC.

From the two latter findings, a correlation between asymptotic film thickness and defect density is found. The defect density is a qualitative, macroscopic and global measure for the effective diffusion coefficient Deff., which is an essential for the film growth, but experimentally difficult to measure.

Results from the small angle X-ray scattering reveal that the micropores in GC have a diameter of about 1 ° A to 2 ° A and that the distribution of the diameter is not very sharp.

Considerable part of the micropores has diameters smaller than 7 ° A to 9 ° A, which cannot be wetted by electrolytes and therefore do not contribute to the capacitance. The diffusive resistance has a minimum (at about 50 m.cm2) after a specific activation time (around 30 to 60 minutes). For larger activation times the resistance increases monotonously.

8 Due to the activation the pore size distribution is shifted towards somewhat larger diameters, with the result that an additional part of the previously smaller pores becomes accessible for the electrolyte and the capacitance increases. Impedance measurements in combination with film thickness determination prove that the volumetric capacitance increases during activation towards a value of approximately 100 F/cm3.

During activation the X-ray diffraction (002)-peak shifts towards larger 2Θ angles, which indicates a densification of the graphite skeleton structure of GC. As a further result it was found that the internal surface area of the active film decreases upon activation, probably because of growth and coalescence of pores on the cost of smaller pores. The two latter results are fully in line with the recently proposed falling cards model for hard carbons by W. Xing et alii.

The exponent of decay of the scattered small angle X-ray intensity for large Q values was found to be very close to 3.0 for fully activated GC (non-activated GC: ≈2.5), which indicates that the internal surface area and the GC are extended through the whole sample volume.

Deviations from all findings above were found for GC activated only weakly (short oxidation at low oxidant concentration or low oxidation temperature):

(i) Evolution of capacitance (and therefore probably also film thickness) follows an exponential law, (ii) graphite skeleton density decreases, (iii) exponent of decay tends towards a minimum of 2.0, (iv) diffusion resistance exceeds the minimum resistance by several orders of magnitude.

The weakly activated GC obviously represents a transition state of oxidized GC. During activation the overall sample experiences a burn-off, which is controlled by chemical reaction, and a film growth, which is controlled by diffusion of reactants.

The superposition of both processes complicates the mathematical formulation of the system considerably.

A mathematical model for the thermal activation of GC was proposed and an exact analytical solution for the evolution of film thickness on GC with plane geometry (monolithic electrodes) and spherical geometry (spheres for powder electrodes) was found.

The film thickness on plane electrodes is described by a so-called Generalized LambertW-Function G. For large activation times (asymptotic behaviour) the film thickness is a constant, which is in given by the ratio of the effective diffusion coefficient Deff. of GC and its reaction rate k.

The time dependence of the film thickness evolution is expressed by G(t). An increase of the oxidant concentration accelerates the film growth, but the asymptotic film thickness is not affected by the concentration. The function G(t) was fitted 9 to experimental data. Experimental data and the model exhibit reasonable agreement.

The model and its solutions are universal and may be extended and also appliedto other systems having reaction controlled changes of sample dimensions anddiffusion controlled film growth.

Table of Contents:

1.	Introduction	21
2.	Electrochemical Double Layer Capacitors	25
2.1	Classification of Electrical and Electrochemical Power Sources	25
2.2	Fundamentals	27
2.2.1	The Electrochemical Double Layer	27
2.2.2	Working Principle of Supercapacitors	28
2.2.3	Design of EDLC with Bipolar Electrodes	31
2.3	Applications for Supercapacitors	33
2.4	Capacitor World Market	34
3.	Glassy Carbon	37
3.1	On Activated Glassy Carbon	40
3.1.1	Thermal Activation of Glassy Carbon	43
4.	Experimental	45
4.1	Sample Preparation	45
4.2	Thickness Determination	46
4.2.1	Determination of the Sample Thickness	46
4.2.2	Determination of the Film Thickness	47
4.3	X-ray Diffraction	48
4.4	Small Angle X-ray and Neutron Scattering	50
4.4.1	Small Angle X-ray Scattering	50
4.4.2	Small-Angle Neutron Scattering	53
4.5	Electrochemical Characterization	53
4.5.1	Cell Setup	53
4.6	Nitrogen Gas Adsorption	54
5.	Data Analysis	55
5.1	X-ray Diffraction	55
5.2	Small Angle Scattering	60
5.3	Electrochemical Characterization	72
5.3.1	Cyclic Voltametry	72
5.3.2	Electrochemical Impedance Spectroscpy	74
5.3.3	Diffusive Resistance of Porous Electrodes	80
6.	Results and Discussions	83
6.1	Thickness and Mass Changes during Activation	83
6.1.1	Weight Loss and Thickness Changes	85
6.1.2	Film Thickness	98
6.2	A Model for the Film Growth	99
6.2.1	Introduction	101
6.2.2	Establishing the Model	102
6.2.3	Experimental	109
6.2.4	Results and Discussion	109
6.2.5	Conclusions	115
6.3	Influence of Activation Parameters on Electrode Performance	117
6.3.1	Influence of Reaction Time and Temperature	117
6.3.2	Influence of O2-Concentration	139
6.3.3	Capacitance after Weak Activation	140
6.3.4	Capacitance of Thermally Activated GC with Different HTT	143
6.3.5	Capacitance of Thermally Activated GC from Capton_Foils	157
6.3.6	Impact of Reduction on Electrode Performance	159
6.3.7	Estimation of Capacitance by Geometrical Considerations	162
6.3.8	Influence of Electrolyte Temperature on Capacitance and Resistance	164
6.4	Comparison with Electrochemically Activated Glassy Carbon	169
6.4.1	General Remarks	169
6.4.2	Electrochemically Treated Thin GC Sheets	171
6.5	Structure and Structural Changes in GC	180
6.5.1	Structural Differences in GC	180
6.5.2	Structural Changes During Activation	187
6.6	Small Angle X-ray Scattering on Glassy Carbon	192
6.6.1	SAXS on Various Non-Activated GC	192
6.6.2	SAXS on Thermally Activated GC	198
6.6.3	Consideration of Plausibility	211
7.	Conclusions	215
A.	Activation of Glassy Carbon Powder	219
A.1	Experimental	219
A.1.1	Oxidation of GC Powder	220
A.2	Evolution of Surface Area during Activation	221
A.2.1	Introduction of the Model	222
A.2.2	Results and Discussion	225
B.	Film Growth Model for Spherical Particles	235