Get Complete Project Material File(s) Now! »
Context of high temperature Solid Oxide Cells
This chapter gives an outline of the fundamentals upon which this work is constructed. In particular, the operating principle, the materials and the basics of solid-state electrochemistry in SOCs are presented. As the electrode microstructure is closely linked to the different physical phenomena within this electrochemical device, a special attention has been paid throughout this chapter to highlight this dependence. The limitations of this technology, mainly linked to long-term degradation, are discussed afterwards. A strategy based on temperature reducing accompanied by an adapted microstructural optimization is proposed. Finally, an overview of the state-of-the-art of advances in this regard is provided with a detailed methodology to reach this goal.
SOC operating principle
The SOCs are high temperature electrochemical energy conversion devices composed of ceramic and metallic components. The same system can alternatively be operated in Solid Oxide Fuel Cell (SOFC) mode to produce electricity by electrochemical oxidation of hydrogen, or in Solid Oxide Electrolysis Cell (SOEC) mode to produce hydrogen from the reduction of steam under an electrical current.
A SOC is a three layers setup made of a dense electrolyte sandwiched between two porous electrodes. The so-called hydrogen electrode is the cell layer in which hydrogen is oxidized in fuel cell mode or produced in electrolysis mode according to the reversible reaction (I.2a). The redox-reaction of O2 takes place in the so-called oxygen electrode following the global reaction (I.2b).
The electrolyte ensures the conduction of the oxygen ions produced by the previous half-cell reactions from one electrode to the other. In SOFC mode, gaseous oxygen molecules are reduced at the oxygen electrode into O2- ions that migrate through the dense electrolyte to be oxidized at the hydrogen electrode into steam and electrons. In SOEC mode, the water molecules are dissociated to form hydrogen gas and oxygen ions that diffuse through the dense electrolyte to be oxidized at the O2 electrode [Ni2008]. The schematic operating principle of a SOC, in fuel cell and electrolysis modes, is shown on Fig. I.1.
Figure I.1 Schematic operating principle of a SOC in (a) SOFC mode and in (b) SOEC with the half reactions for each electrode.
Operating conditions – Typical SOC operating temperatures are between 500°C and 1000°C. This high temperature level is crucial as the ionic conductivity of the solid electrolyte depends exponentially on temperature. In SOFC mode, this allows flexibility to run directly on hydrocarbon fuels such as methane via internal reforming [Vernoux2000, Zhu2003, Tu2004, Morel2005, Laurencin2008]. In SOEC mode, carbon dioxide and steam can simultaneously be co-electrolyzed to produce a syngas of carbon monoxide and hydrogen [O’Brien2008, Aicart2014].
Besides, conversely to low temperature technologies, like Polymer Electrolyte Membrane (PEM) fuel cells, no expensive catalysts like Platinum are required for the activation of the electrochemical reactions [Kiburakaran2009, Mekhilef2012]. Indeed, the high operating temperature promotes the efficiency of the electrochemical reactions by increasing kinetic rates, resulting in high performances. In addition, from a thermodynamic point of view, working at high temperature can also allow reducing the cost of hydrogen production if a heat source is available (nuclear power or renewable energy for example). Indeed, the electrical demand for the water splitting (Gibbs energy, ΔG) is reduced when increasing the operating temperature. As illustrated in Fig. I.2, the decrease in ΔG is then balanced by an increase in the thermal need (entropy, TΔS) while the enthalpy ΔH is only slightly affected by the operating temperature [Hino2004, Ni2008, Laguna2012].
Application areas – The most prominent application area for SOC-based systems is stationary power plants. At this relatively large scale (several kW to MW), the high temperatures and the associated structural problems during heating up and cooling down can be handled most effectively. This technology is particularly suitable to deal with the renewable energies intermittence (such as solar or wind energies). Indeed, the flexibility of this system between electrical energy storage and production allows matching the electricity consumption with the fluctuating energy production.
SOC materials and manufacturing
The choice of materials is of central importance for the cell reliability. Each SOC component (i.e. electrolyte, H2 and O2 electrodes) needs to fulfill several requirements that are sometimes conflicting, a compromise is then to find between these different constraints. They should allow highly efficient electrode reactions (Eqs. I.2a and I.2b) as well as ensure fast reactant transport. Hence, the electrodes must have a high catalytic activity and high electronic and ionic conductivity, combined with sufficient reaction sites and good transport paths for the different species. The later point emphasizes the importance of an optimized electrode microstructure. A special attention is given throughout this work to address the microstructural requirements. Besides, the materials should be stable at high temperature either under oxidizing or reducing atmospheres at the O2 or H2 electrodes, respectively. The electrolyte material ought to exhibit a good ionic conduction and a negligible electronic one to prevent short circuits. However, each component cannot be optimized individually. Indeed, the whole materials assembly should also own compatible thermal expansion behavior to prevent the cracking and delamination that can result from the mismatch in thermal expansion coefficients. Finally, materials should not be reactive with other cell components during manufacturing or operation and to be stable over long-term operation.
The numerous requirements have promoted extensive material research in the course of the last decades to optimize the components of SOCs (i.e. electrolyte, H2 and O2 electrodes). Several comprehensive overviews over this subject can be found in the literature [Minh1993, Yamamoto2000, Tietz2002, Brandon2004, Haile2003, Weber2004]. The most prominent materials used in state-of-the-art SOC components will be briefly presented in the following subsections.
The main function of the electrolyte is to separate the cell electrodes. It must be gas-tight and conduct oxide ions O2- between the O2 and the H2 electrodes. Hence, it should provide significant ionic conductivity and negligible electronic conduction. A conventional choice of electrolyte material in SOCs is stabilized zirconia (YSZ, Y2xZr1-2xO2-x, where x is the concentration of the dopant in mol.% Y2O3 in ZrO2). The doping of ZrO2 with trivalent yttrium ions forms oxygen ion vacancies in the host lattice. This allows vacancy-mediated conduction of oxygen ions and stabilizes the cubic phase of zirconia at higher temperatures. The dopant concentration has a significant impact on the ionic conductivity of the electrolyte that ranges from 0.05 S.cm-1 to 0.14 S.cm-1 at 1000°C for a doping ratio between 5 %mol and 12 %mol [Yamamoto2000]. The electrolyte is classically doped with 8 %mol Y2O3 (denoted as 8YSZ in the following). Besides, 3YSZ (3 mol% Y2O3 in ZrO2) has drawn attention due to its higher mechanical stability. But it is not suitable for intermediate or low temperature usage because of its lower ionic conductivity.
In order to decrease the operating temperature, other ceria-based oxide materials are under investigation. Among these materials gadolinium-doped ceria (CGO, Ce1-2xGd2xO2-x-δ, x=dopant concentration of Gd2O3 in CeO2) is a promising candidate. However, one drawback is that for low oxygen partial pressures (particularly on the H2 electrode side), they become electronically conductive, especially at temperatures above 600°C [Dalslet2006]. Nevertheless, these materials are still of interest for other applications (like LSCF-CGO composite O2 electrode).
The role of the H2 electrode is to provide an appropriate medium for the hydrogen reduction/oxidation reactions. Hence, the main material requirements are high catalytic activity, high electronic and ionic conductivities, compatible thermal expansion and chemical stability. For an H2-electrode-supported cell, this electrode provides additionally the mechanical support to the cell. It is usually composed of two layers; (i) a 300-1500 µm thick highly porous substrate which provides the mechanical stability and the transport of gases from and towards the second layer which is (ii) a 5-10 µm thin electrode functional layer where the reactions take place.
Thanks to its high electrical conductivity (e.g. σel>105 S.m-1 under 1000°C [Ormerod2003]) and high electrochemical activity, the Nickel is considered as a standard material for the H2 electrode. However, using the Ni alone limits the reactions (Eqs. I.2a and I.2b) to the Triple Phase Boundary lengths (TPBls) in contact with the electrolyte. A porous composite of Ni and YSZ is classically used to extend the reactions into the volume of the electrode. Adding YSZ to Ni not only allows extending the reactions into the thickness of the electrode, but also leads to reduce the thermal expansion coefficient mismatch between the electrode and the electrolyte. The composition of this ceramic-metallic composite called ‘cermet’ is then a parameter that needs to be adjusted for an optimal electrode performance and robustness [Ni2008]. The Ni content usually varies between 30 and 50 %vol for ideal properties [Aruna1998].
It should be noticed that the H2 electrode microstructure plays a huge impact on the cell performance. Indeed, while a fine microstructure is requested for the active layer to maximize the reaction sites (especially the TPBls density), a coarse microstructure is praised for the substrate to favor the gases diffusion. Both Ni-YSZ layers manufacturing involves classically the use of relatively cheap ceramic processing techniques like tape casting or screen printing [Tietz2000, Moon2008]. For detailed reviews about these ‘classical’ manufacturing processes, the reader is invited to refer to the following references [Thorel2015, moon2008, Somalu2017], and for less common processes [Tietz2002]. Moreover, in order to reach the high porosity levels expected for the substrate, pore-formers are usually mixed to the materials powder prior to sintering [Sanson2008]. The pore-formers consist in organic or polymer compounds that are burned before sintering. This operation leads to the creation of macropores in the final microstructure that facilitate the gases diffusion through the H2 electrode.
The components of this electrode are working under a highly oxidizing atmosphere at high temperature. Therefore, pure oxide ceramics are usually used in this case. The first designs for this electrode consisted of a composite of a good electronic conductor mixed with an electrolyte material for the ionic conduction. The microstructure of such a composite was similar to the hydrogen Ni-YSZ electrode with reactions taking place at the TPB lengths. A classical O2 electrode was La1-xSrxMnO3-delta – YSZ (LSM-YSZ). A second-generation class of materials is the ABO3-type perovskite manganites, cobaltates, and ferrates (A=La, Sr, Ca, Ba; B=Mn, Co, Fe) [Joos2017]. Depending on the configuration, these perovskites exhibit a Mixed Ionic and Electronic Conduction (MIEC). Their main advantage is that the entire MIEC surface contributes to the electrochemical reactions (Eqs. I.2a and I.2b) as discussed by Monaco et al. [Monaco2018]. La1-xSrxCoO3-delta (LSC) and La1-xSrCoyFe1-yO3-delta (LSCF) perovskites are the most used MIEC materials. These compounds exhibit improved electrochemical performances compared to the LSM-YSZ composite [Tietz2006]. However, the main drawback remains their chemical instability and their reactivity with YSZ (cf. Section I.4.2). In order to mitigate this problem, a thin CGO barrier layer of around 5 µm is commonly placed between the electrolyte and the MIEC. Finally, it should be noted that the addition of highly ionic conducting materials like Ceria doped Gadolinium Oxide (CGO) to LSCF or LSC has attracted much attention for improved O2 electrode [Gao2016, Dusaster1999, Hwang2005]. Indeed, compared to the single solid-phase materials, the LSCF-CGO or LSC-CGO composite exhibits higher performances at intermediate temperatures and presents a better mechanical compatibility with the classical electrolyte in Yttria Stabilized Zirconia (YSZ) [Kim17a, Kim17b, Leng08, Dusatre99]. For this reason, a special attention is given to the LSCF-CGO composite in the following.
Physical phenomena in SOCs
During operation, irreversible processes occur in the system, resulting in a drop of cell performances. In order to characterize these physical phenomena, classical techniques involve steady-state polarization measurements (i-V function) and electrochemical impedance spectroscopy (EIS). The i-V curve plots the cell voltage evolution as a function of the current density, Vcell=f(i). This dependence gives relevant information about the cell performances. Regarding the EIS plot, this technique evaluates the transient behavior of the cell with a decomposition in the frequency range by applying a sinusoidal ‘small’ perturbation iac(t) and measuring its answer. The plot of the impedance Z(f) provides information on the reaction mechanisms.
Table of contents :
Layout of the manuscript
Chapter I: Context of high temperature Solid Oxide Cells
I.1 SOC operating principle
I.2 SOC materials and manufacturing
I.2.2 Hydrogen electrode
I.2.3 Oxygen electrode
I.3 Physical phenomena in SOCs
I.3.1 Open circuit voltage (OCV)
I.3.2 Ohmic overpotential
I.3.3 Concentration overpotential
I.3.4 Activation overpotential
I.4 SOC durability and degradation mechanisms
I.4.1 SOC durability
I.4.1 Degradation mechanisms
I.5 Importance of microstructure
I.6 State-of-the-art review of the investigated fields
I.6.1 Overview on microstructure stochastic geometrical models
I.6.2 State of the art of microstructural correlations
I.6.3 Overview on LSCF-CGO reaction mechanisms
I.6.4 State of the art of LSCF-CGO microstructural optimization
I.7 Conclusion and methodology
Chapter II: 3D reconstruction and characterization of SOC electrodes microstructures
II.1 Presentation of the investigated cells
II.1.1 Cell-A: commercial complete cell in LSCF and Ni-YSZ
II.1.2 Cell-B: commercial complete cell in LSC and Ni-YSZ
II.1.3 Cell-C: in-house symmetrical cell in LSCF-CGO
II.2 Three-dimensional reconstruction of SOC electrodes
II.2.1 3D reconstruction via FIB-SEM tomography
II.2.2 3D reconstruction via synchrotron X-ray holotomography
II.3 Raw image processing and segmentation
II.4 Macropores separation from the micropores network in the diffusion layer
II.5 Microstructural quantitative characterization: properties measurement
II.5.1 Morphological parameters
II.5.2 Physical parameters
II.6 Representative Volume Element (RVE)
II.7 Results and discussion
II.7.1 Results for Cell-A
II.7.2 Results for Cell-B
II.7.3 Results for Cell-C
Chapter III: Stochastic geometrical models and validation
III.1 Random Field model description and validation
III.1.1 Random Field model description for SOC electrodes
III.1.2 Validation of the random field model on 3D reconstructions
III.1.3 Discussion: model flexibility to various electrode microstructures
III.2 Particle-based model description and validation
III.2.1 Particle-based model description for SOC electrodes
III.2.2 Validation of the sphere-packing model on 3D reconstructions
III.2.3 Model flexibility: illustration for a diffusion layer with a bimodal pore distribution
III.3 Concluding remarks: models comparison
Chapter IV: Microstructural correlations
IV.2 Theoretical development and methodology
IV.2.1 Theoretical development: analytical expression for Si/j and Si
IV.2.3 Methodology for calibrating and validating the microstructural correlations
IV.3. Calibration of the microstructural correlations
IV.3.1 Correlations for the electronic/pores and ionic/pores interfacial specific surface area
IV.3.2 Correlation for the electronic/ionic interfacial specific surface area
IV.3.3 Correlation for the density of TPBl
IV.4. Discussion and validation
IV.4.1 Model prediction with respect to the RF and SP microstructures
IV.4.2 Model validation on diverse electrodes reconstructions
IV.4.3 Model validation on the loss of TPBl after Ni agglomeration
Chapter V: Impact of microstructure on the LSCF-CGO composite electrode performance
V.1 Experimental conditions, modeling and methodology
V.1.1 Testing conditions
V.1.2 Electrochemical model description
V.2. Experimental results and models validation
V.2.1 Electrochemical measurements
V.2.2 Calibration and validation of the electrochemical model
V.3. Role of the LSCF-CGO microstructure on the reaction mechanism
V.3.1 Reaction mechanism at 700°C and impact on the electrode performances
V.3.2 Role of microstructure on the electrode response: effect of composition, porosity
and particle size
Conclusions & Perspectives
A.1 Expression of the correlation rX(h) as a function of the weight function ω(h)
A.2 Relation between CX(h), rX(h), and lX
A.3 Correlations for the specific surface area between the electronic and pore phases
A.4 Correlations for the Triple Phase Boundary lengths TPBls density