Fundamental structure of a SOFC
The basic components of a SOFC and the net reactions at each electrode are given in Figure 1. The gaseous fuel diffuses into the porous structure of the anode, and is oxidized with the help of an oxygen ion from the electrolyte to release electrons. The electrons next transport through the electronically conducting phase in the anode to the external circuit and to the cathode. There, molecular oxygen is reduced into oxygen anions.
Figure 1. Simple drawing of a SOFC with the net reactions at each electrode .
The overall cell reaction may be:
H2 + ½ O2 s H2O (1)
CO + ½ O2 s CO2 (2)
CH4 + 2O2 s CO2 + 2H2O (3)
The SOFC electrolyte is an oxygen-anion conductor. The operating temperature of the SOFC is mostly set by the requirement for high ionic conductivity of the electrolyte. So, for example, a temperature higher than ~700°C is necessary for yttria-stabilized zirconia electrolyte .
Any part of a SOFC, anode, cathode, or electrolyte, can serve as a mechanical support which is made much thicker than other parts. The favor trend today is to reduce the operating temperature from ~1000°C to 500-800°C in order to reduce the cost of the other parts of SOFC. Therefore, a cell design with a thin electrolyte to lower the ohmic resistance and a thick mechanical support on the anode side is the best choice (Figure 2). The support is usually made of the anode material, but with a coarser microstructure than that of the anode functional layer .
The electrolyte of a SOFC is a solid oxide ion conductor which has to meet certain criteria on the electrochemical, chemical, thermodynamical, thermal and mechanical properties as listed below :
‚ a high ionic conductivity for the oxygen anion (> 10-3 S cm-1), and a low electronic conductivity (to avoid an internal short circuit between the anode and the cathode) over a wide range of oxygen pressures, since the electrolyte is subjected to an oxidizing atmosphere at the cathode side ( PO2 ~1 atm) and to a reducing atmosphere at the anode side ( PO2 ~10-20 atm); ‚ to be chemically stable in relation to the reactant environment and contacting electrode materials under SOFC operation as well as fabrication conditions; ‚ a thermal expansion compatible with the other parts; ‚ to be dense enough to separate the fuel and the air compartments; ‚ to be thermodynamically stable over a wide range of temperature and PO2 .
where jo and E are factors depending on the electrolyte materials, T is the electrolyte temperature, and R is the ideal gas constant. The ionic conductivity will increase by increasing the operating temperature, or by refining the crystal structure by doping methods.
Four groups of material have been used as SOFC electrolyte: doped ZrO2 and CeO2, LaGaO3-based perovskites, and apatites . The first two groups are most widely employed and are thus discussed further.
A high dopant concentration leads to the introduction of more vacancies into the lattice, as well as more interactions (associations) between oxygen vacancies and dopant cations, which reduce the free vacancy concentration . So, the conductivity as a function of the dopant concentration will reach a maximum at relatively low additions of dopant. The composition with 8 mol% Y2O3 (8YSZ) has traditionally dominated because the ionic conductivity exhibits a maximum at that yttria content.
Both doped and undoped ceria possess a mixed ionic and electronic conductivity at low oxygen pressure. The conductivity depends on the activation energies of the oxygen ion migration via the oxygen vacancies and of the defect complex association. The contribution of the electronic conductivity becomes significant under reducing atmosphere at high temperature.
Trivalent rare earth oxides dopants such as Gd2O3, Sm2O3, Y2O3 can induce a much higher conductivity and stability with CO2 and H2O than bivalent ones because they have an ionic radius closer to that of the host ion . The oxide ion conductivity is a function of temperature, dopant concentration and type as read from Table 2.
The maximum ionic conductivity occurs at ~10-20 mol% for most dopants. Ce0.9Gd0.1O1.95 (10CGO) is the most studied ceria based electrolyte . The inconsistence of the conductivity peaks for different Gd contents shown in Figure 4 might come from impurities inside the starting materials, synthesis processes, and sintering conditions which create different microstructures with different grain boundaries and bulk resistances. The impurities may come from the original ores as SiO2 is ubiquitous in minerals, from furnace refractories during the sintering procedure or from glass-ware used for the precursor fabrication .
The ionic conductivity also depends on surrounding atmosphere. In general, there exist always exchanges of oxygen between the solid (O2- ions), at least at the surface, and the gaseous atmosphere (oxygen molecules), followed by the formation of oxygen vacancies. The two electrons of O2- ion remain on the oxygen site. They are strongly attracted by the nearby cations and easily delocalised on all the cationic sites of the solid. The nature of majority defects (doubly ionized oxygen vacancies Vo¨ and electrons) depends on the temperature and on the partial pressure of oxygen. In a reducing atmosphere ( PO2 <10-18 bar), Vo¨ and electrons are in majority
(n-type electronic conduction), while in an oxidizing atmosphere ( PO2 >10-10 bar), cationic vacancies are numerous and capture the electrons released (p-type electronic conduction) .
The prominent drawback of ceria is that, at low oxygen partial pressure and high temperature, Ce4+ is easily reduced to Ce3+, resulting in a n-type electronic conduction and a lattice expansion. The electronic conduction will facilitate electronic leakage currents between anode and cathode . The lattice expansion may create cracks and fissures at the electrode/electrolyte interface and a subsequent delamination of the electrode from the electrolyte [5,16].
Moreover, CeO2 will react with YSZ at high temperatures [1,5,7]. So, doped ceria is most suitable for intermediate temperature SOFC (below 800°C) , while doped zirconia is better for high temperature SOFC.
Anode material and three-phase boundary
The material for SOFC anode is a porous cermet (a composite of ceramic and metal) which is a mixed ionic and electronic conductor. The two most commonly applied anodes are Ni-YSZ and Ni-CGO because of their low cost, high conductivities and good catalytic activities . Ni serves as a catalyst for H2 bond breaking, steam reforming of hydrocarbons and as a channel for electron transport. The ceramic phase acts mainly as a framework to retain the dispersion of the metal particles and the porosity during long-term operation, as well as a channel for O2- to diffuse farther into the anode. Moreover, the porosity provides the pathways for the fuel diffusion and products removals.
Oxidation reaction can only occur at triple phase boundaries (TPB), the positions where the ionic conducting phase (for O2-), electronic conducting phase (for Hads, electron) and gas phase are in contact. The effective TPB in fact extends ~10 µm from the electrolyte into the electrode despite the effort to use cermet anode [18,19]. The electrolyte material within the electrode is only effective if its particles are sintered to the electrolyte and linked together. This requires a high sintering temperature of at least 1500K for YSZ particles . It is clear that SOFC performance depends strongly on the microstructure of the anode, in which a fine homogeneous distribution of three phases is important for anode to operate efficiently.
The electrical conductivity of a cermet anode depends on the particle size, size distribution, contiguity of each component, porosity, and ratio of nickel/YSZ content. The percolation threshold for the electronic conductivity is at about 30 vol% nickel as indicated in Figure 5. This value decreases when the NiO particle size is reduced while the YSZ size is increased. It is the average size, not the BET-specific surface that determines the cermet conductivity, since small particles can agglomerate to give a surface area similar to that of big particles . The suggested anode electronic conductivity varies from 1 to 100 S/cm, depending on how well a current collector is connected to the anode .
When operating on hydrocarbon fuels, the anode must be stable against carbon dioxide and sulfide compounds. Ni, however, is sensitive to sulfur and catalyzes the deposition of carbon [1,6], leading to a rapid deterioration of the cell performances. One approach is to replace Ni with Cu which catalyzes nothing and add cerium oxide to act as an oxidation catalyst , another one is to dope Ni/YSZ anode with molybdenum and gold .
The cathode in SOFC is responsible for the reduction of oxygen and the transport of oxygen ions to the electrolyte. So, it also operates based on the triple phase boundaries like the anode. A typical choice is the perovskite (La,Sr)MnO3±d (LSM) [9,14]. The YSZ/LSM composite is used primarily instead of Sr-doped LaFeO3 (LSF) or LaCoO3 (LSC) since YSZ-LSM mixtures can be heated to higher temperatures before undergoing a solid state reaction with YSZ to form a La2Zr2O7 insulating phase [21,22], resulting in the cell degradation.
For lower operation temperatures of SOFCs, La1-xSrxCo1-yFeyO3-d (LSCF) may be used for its fast ion transport, good oxygen reduction kinetics and acceptable electronic conduction .
By conducting semi-empirical models counting all the chemical reactions and diffusion steps with only one or two charge-transfer reactions, the authors supported the hydrogen spillover pathway and the four rate-determining processes1*貸 are: water adsorption/desorption on YSZ, water dissociation on YSZ, surface diffusion of 蛸託只 and hydrogen spillover to oxide ion.
However, the density-functional theory calculations by Rossmeisl et al.  of the surface adsorption energies of the hydrogen atoms, oxygen atoms and hydroxyl radicals on metals showed that the measured conductivity obtained by Setoguchi et al.  is well-correlated with the oxygen binding energies, indicating the dominance of oxygen spillover reaction pathway.
No consensus exists so far concerning the kind of the rate determining reaction step. It is assumed to be the hydrogen adsorption/diffusion process on the surface of Ni particles and a charge transfer process on zirconia electrolyte surface by Jiang et al. . Different findings from the literature may imply that the active mechanism varies with operating conditions and/or sample preparation methods .
Table of contents :
CHAPTER 1 LITERATURE SURVEY
2. FUNDAMENTAL STRUCTURE OF A SOFC
2.1.1. Doped zirconia
2.1.2. Doped ceria
2.2. ANODE MATERIAL AND THREE-PHASE BOUNDARY
3. OXIDATION MECHANISM ON SOFC ANODE
4. SOFC ELECTRODE POLARIZATION
5. EFFECTS OF SULFIDE POLLUTANTS
5.1.MAJOR COMPONENTS OF BIOGAS
5.2.MINOR COMPONENTS OF BIOGAS
5.3. EFFECTS OF SULFIDE COMPOUNDS ON SOFC
5.4. LONG-TERM BEHAVIOR OF A SOFC UNDER H2S
CHAPTER 2 EXPERIMENTAL METHODS AND PROCEDURES
2. RAMAN SPECTROSCOPY
3. IMPEDANCE SPECTROSCOPY
3.1. PRINCIPLE OF MEASURE AND ANALYSIS
3.2. THE CAPACITIVE DOUBLE LAYER
3.3. ORIGIN OF INDUCTIVE ELEMENTS
4. SCANNING ELECTRON MICROSCOPE (SEM)
5. X-RAY DIFFRACTION (XRD)
6.1. GAS FLOW CONTROL
6.2. HOME-MADE IN SITU CELL (LEPMI)
6.3. INVESTIGATIONS OF H2S AND NI REACTION
6.3.1. Ni pellet making
6.3.2. Contact with H2S at a working temperature
6.3.3. Contact with H2S during the heating process
6.4. INVESTIGATIONS OF H2S AND NI-CGO REACTION
6.4.1. Powder mixing
6.4.2. Ni-CGO pellet making
6.4.3. Ni-CGO pellet characterizations
188.8.131.52. Raman spectrum of doped CeO2 from literature
184.108.40.206. Raman spectra of Ni-CGO
220.127.116.11. Morphology of Ni-CGO pellet
6.4.4. Investigation procedure for H2S and Ni-CGO reaction
6.5. HALF-CELL NI-YSZ/YSZ
6.5.1. Sample construction
6.5.2. Sample installation
6.5.3. Experimental procedure
CHAPTER 3 EFFECTS OF H2S ON ANODE MATERIALS
2. RAMAN SPECTRA OF NICKEL SULFIDE COMPOUNDS
2.3. THERMAL DECOMPOSITION OF NIS AND NI3S2
2.4. OTHER NICKEL SULFIDES
3. IMPACTS OF H2S ON NI PELLET
3.1. IDENTIFICATION OF THE REACTION KINETICS AND PRODUCTS
3.1.1. In situ Raman spectroscopy
3.1.2. Phase identifications by X-ray diffraction
3.1.3. Conclusion on the reactivity of H2S on Ni with temperature
3.2. SURFACE MORPHOLOGY CHANGES
3.2.1. In situ optical imagery monitor
3.2.2. Ex situ investigations by Scanning Electron Microscopy
3.3. IMPACTS OF H2S ON NI PELLET DURING THE HEATING PROCESS
4. IMPACTS OF H2S ON NI-CGO ANODE MATERIAL
4.1. AT 715°C AND ABOVE
4.1.1. Formation of nickel sulfide crystals at 715°C
4.1.2. Disappearances of nickel sulfide crystals at higher than 715°C
4.1.3. Morphological changes under H2S at above 715°C
4.2. AT 500°C
4.3. AT 200°C
5. REMOVAL OF NICKEL SULFIDES
5.1. AT 850°C IN AR
5.2. AT 715°C IN 3%H2/AR
CHAPTER 4 EFFECT OF H2S ON ELECTROCHEMICAL PROPERTIES OF SOFC ANODE
2. REVIEW OF IMPEDANCE STUDIES ON THE EFFECTS OF H2S ON SOFCS
3. GENERAL ANALYSIS OF IMPEDANCE SPECTRA OBTAINED AT 500°C
3.1. TYPICAL SHAPES OF IMPEDANCE SPECTRA
3.2. STRUCTURE AND SHAPE OF CONCENTRATION IMPEDANCE
3.3. PROPOSED EQUIVALENT CIRCUIT
4. CHARACTERIZATION OF ANODE INITIAL STATE AT 500°C IN CLEAN FUEL
5. EFFECT OF H2S ON 500 MV-POLARIZING CELL (500MV-CELL) AT 500°C
5.1. AGING BEHAVIOR IN CLEAN FUEL
5.2. EFFECT OF H2S ON THE ELECTRICAL PROPERTIES
6. EFFECT OF H2S ON CELL IN OPEN CIRCUIT CONDITION (OCP-CELL) AT 500°C
6.1. AGING BEHAVIOR IN CLEAN FUEL
6.2. EFFECTS OF H2S ON ELECTRICAL PROPERTIES
7. CORRELATION BETWEEN NICKEL SULFIDE QUANTITY AND ELECTRICAL CHANGES
8. EFFECT OF H2S ON MORPHOLOGY CHANGE
GENERAL CONCLUSION & PERSPECTIVES