Table of contents

Mixed ionic-electronic conducting (MIEC) materials are of great interest for a variety of high-temperature applications, such as solid oxide fuel cell (SOFC) cathodes or dense ceramic membranes for gas separation. Several MIEC perovskite oxides, namely of the composition A_xSr_1-xCo_yFe_1-yO_3-d (A = La, Ba), exhibit excellent oxygen-ionic and electronic transport properties. Ba_0.5Sr_0.5Co_0.8Fe_0.2O_3-δ (BSCF), La_0.58Sr_0.4Co_0.2Fe_0.8O_3-δ (LSCF), or La_0.6Sr_0.4CoO_3-δ (LSC) are amongst these state-of-the-art high-flux materials and, hence, very promising candidates for a high-permeation oxygen transport membrane (OTM).

In order to assess their applicability for OTMs under operating conditions, it is essential, though, to determine their electrochemical transport properties not only under oxidizing conditions (pure oxygen, air), but also as a function of oxygen partial pressure pO₂ down to low values of, e.g., 10^-6 bar. This can be achieved in a closed tubular zirconia "oxygen pump" setup [1,2] facilitating precise and continuouspO₂ control in the entire range between pO₂ = 10^-20...1 bar above 700 °C. By performing fast pO₂ step changes and applying the electrical conductivity relaxation (ECR) method to selected MIEC compositions, the oxygen equilibration kinetics of dense ceramic bulk samples of BSCF, LSCF, or LSC have been studied, yielding chemical diffusion coefficients, D^δ, and surface exchange coefficients, k^δ, as f(T,pO₂) for 700 ≤ T / °C ≤ 900 and 10^-5 ≤ pO₂ / bar ≤ 0.21. Furthermore, chemical stability issues of MIEC oxides are also addressed over a wide pO₂ range using this "oxygen pump".

By reducing the thickness of an OTM, the oxygen flux will be increased. However, upon reducing the thickness below a "characteristic" membrane thickness [3], given by the ratio of D^δ and k^δ, surface exchange becomes rate-determining for thin membranes. No enhancement of the oxygen flux is achieved by thinning the membrane any further if one cannot enhance the surface oxygen exchange. This, however, is possible by modifying the membrane surfaces with a porous functional/catalytic layer that provides more surface area for the exchange of oxygen between the gas phase and the membrane material. This becomes especially important for the permeate-side OTM surface where k^δ values are reduced as a result of the low pO₂.

This concept has already been successfully applied to OTMs [4], facilitating unprecedented oxygen fluxes in the case of a dense BSCF membrane covered with a microporous BSCF functional layer [5], but also to SOFCs where nanoscaled LSC thin-film cathodes with a nanoporous microstructure [6] led to a large enhancement of oxygen surface reduction and, in combination with the occurrence of chemical hetero-interfaces [7], thus resulted in the best performance of SOFC cathodes reported so far in literature.

With the help of a 3D FEM OTM transport model [8] the interplay of transport parameters (D^δ and k^δ) and functional-layer microstructure (thickness, porosity, particle sizes) can be readily assessed. In our present study it is shown that the choice of materials in an OTM depends not only on their pO₂- and temperature-dependent transport parameters, but also on phase stability, chemical compatibility and the optimum functional-layer microstructure. To this end high-resolution structural analyses (TEM) are also required.

References

[1] C. Niedrig et al., J. Electrochem. Soc. 160 (2013), F135.

[2] C. Niedrig et al., manuscript in preparation (2013).

[3] H. J. M. Bouwmeester and A. J. Burggraaf, in P. J. Gellings and H. J. M. Bouwmeester (Eds.), The CRC Handbook of Solid State Electrochemistry, CRC Press, Boca Raton FL (1997), p. 481.

[4] S. Baumann et al., J. Eur. Ceram. Soc. 33 (2013), 1251.

[5] S. Baumann et al., J. Membrane Sci.377 (2011), 198.

[6] J. Hayd et al., J. Power Sources196 (2011), 7263.

[7] J. Hayd et al., J. Electrochem. Soc. 160 (2013), F351.

[8] A. Häffelin et al., ECS Trans. 57 (2013), 2543.

The tungsten - tungsten trioxide (W-WO3) materials system has recently shown excellent promise as a medium for grid level energy storage and for solar energy conversion. In both these applications the thermal energy transfer is accompanied by a redox reaction in the W-WO3 couple. For viability of this system as an energy storage medium, it is important to have facile redox kinetics that allows us to rapidly store and release energy. Facile kinetics are also required in the utilization of this system in solar syngas production. In this paper we show that the addition of an ionic conducting phase, namely yttria stabilized zirconia (YSZ) greatly enhances the redox kinetics of the W-WO3 couple. Specifically, the reduction kinetics of the WO3-W reaction are facilitiated by the addition of YSZ. As a control experiment, a redox experiment with a filler material that is not an ionic conductor did not show such an enhancement in the kinetics. Chemical reaction rates were determined by transient thermogravimetry, and the data were analyzed using relevant kinetic and transport theory.

Recently, interest in solid-oxide electrolysis cells (SOECs) with platinum or nickel cermet electrodes capable of reducing combustion effluents to synthesis gas has driven the search for mixed ionic-electronic conductors with suitable catalytic activity for reduction of carbonaceous species. There remains a need for discovering chemically stable electrode materials combining electronic and oxygen-ionic conductivities with suitable catalytic activity for oxidation or reduction. Our research group has focused upon identifying ambivalent transition-metal dopants capable of imparting both electrochemical and catalytic properties to perovskite electrolyte materials. In this study, the authors present a combined electrochemical, catalytic and permeation investigation of iron-doped barium zirconium perovskite in the context of the electrochemical reduction of combustion gas (CO₂, H₂O) to synthesis gas (CO, H₂).

Powders of BaZr_0.90Fe_0.10O_3-δ (BZF10) were synthesized via solid-state reaction and characterized by synchrotron X-ray diffraction (XRD) and transmission electron microscopy (TEM) coupled with energy dispersive X-ray spectroscopy (EDS). The powders were pressed and sintered into 13mm dense button cells, and a porous Pt electrode and a dense Pt blocking electrode were attached to either face of the pellet. The resulting system was used to conduct electrical conductivity relaxation (ECR), electrochemical impedance spectroscopy (EIS), and reaction analysis over a wide range of dry O₂, CO₂, H₂O, and humidified CO, H₂ partial pressures at temperatures of 600 – 800^oC. Analysis of equilibrium electrical conductivities (σ_∞) vs. temperature and partial pressures provides critical insight into the underlying electrochemical mechanisms for surface reaction. Non-linear regression of conductivity vs. time data allows deduction of surface electrochemical reaction rates and solid-state diffusivities of charge carriers, and probing of surface oxygen-acceptor/donor kinetics. BZF10 was verified as an electrochemical stable, mixed ionic-electronic conducting ceramic suitable for SOEC application. The σ_∞ vs. O₂ partial pressure indicates a high oxygen-ion acceptor activity for Ba_0.9Fe_0.1ZrO_3+δ. Equivalent experiments conducted over a range of CO₂ partial pressures (pCO₂) indicate a net decrease in σ_∞ with respect to pCO₂, coinciding with up to a 16% conversion of CO₂ to CO and O₂ at steady-state. Reduction of the Fe-doped perovskite suggests the formation of a CO₃^-2 intermediate during the overall conversion of CO₂ to O₂ and probing of CO₂ reduction activity indicates significant catalytic activity, which is attributed to the oxygen acceptor-donor activity of the Fe-doped perovskite. Experimental results indicate that the material behaves as a mixed oxygen ionic-electronic conductor, with significant catalytic activity for CO₂ reduction.

Hydrogen is an important raw material for production of ammonia, methanol, liquid hydrocarbons, etc. The application of membrane technology is expected to considerably reduce the capital and energy cost in hydrogen production. Composite membranes consisting of BaCeO₃-based proton conductor and electronic conductor (e.g. nickel) have been developed for this application. However, these membranes (e.g. Ni–BaZr_0.8–xCe_xY_0.2O_3–δ (Ni-BZCY), 0.4 ≤ x ≤ 0.8) suffered serious performance loss in CO₂-containing environment at 900 ^oC due to the reaction between BaCeO₃ and CO₂.^{[1, 2]} In order to avoid the chemical stability issue of BaCeO₃, CO₂-tolerant hydrogen membranes have been developed, e.g., RE₆WO_12-δ (RE: rare earth metal), Ca-doped LaNbO₄, Ce_0.8Sm_0.2O_2-δ, and Ni-La_0.4875Ca_0.0125Ce_0.5O_2-δ. However, the permeation fluxes of those membranes are significantly lower than that of Ni-BZCY membranes due to their low proton conductivity and/or electronic conductivity. Among the proton conductors that are tolerant to CO₂, BaZr_0.8Y_0.2O_3–δ (BZY) possesses the highest bulk proton conductivity.^[3]However, single phase BZY membrane shows very low hydrogen flux due to the relatively poor electronic conductivity. The low flux may be resolved in a similar way to Ni-BZCY membranes: combining BZY with highly electronic-conducting Ni to form Ni-BZY composite membrane. Ni-BZY membrane is expected to possess both high hydrogen permeation flux and chemical stability, which are the key factors for successful adoption of Ni-BZY hydrogen permeation membrane for practical applications. However, there has been no report on Ni-BZY composite membrane for hydrogen permeation study, probably due to the difficulty in obtaining dense membrane with large BZY grains.

The sintering of Ni-BZY membrane needs to be performed below the melting point of Ni (~1453 ^oC). Unfortunately, due to the highly refractory nature, BZY samples prepared through the traditional solid state reaction method needs to be sintered at extremely high temperatures (1700-2100 ^oC) for a long time (24 h) to reach relatively high density. At a sintering temperature of 1400 ^oC, conventional solid state reaction method can only produce BZY with low relative density and small grain size. Moreover, BZY has a low grain boundary proton conductivity due to the blocking effect of space charge layer.^[4]Small grain size and large number of grain boundaries will greatly limit the total proton conductivity of BZY.

In this work, the sintering behavior, microstructure, and phase composition of the Ni-BZY from different methods were investigated. Dense Ni-BZY membranes with large BZY grains were successfully achieved through a two-step solid state reaction method.

Fig. 1 shows the surface and cross-section SEM images of sintered Ni-BZY membrane. The membrane is very dense. The size of BZY grains is about 1 μm. The size of Ni particles is rather large, ~ 5-10 μm. The membrane is conductive at room temperature, suggesting a connective network of Ni is formed. The flux of a 0.40-mm-thick Ni-BZY membrane at 900 ^oC in wet 20 and 40% H₂ are 3.4 and 4.3*10^-8 mol/cm²s, respectively. These values are highest among all non-BaCeO₃-based hydrogen membranes, suggesting the Ni-BZY is very promising in the application for hydrogen separation.

Acknowledgements

We gratefully acknowledge the financial support from the HeteroFoaM Center, an Energy Frontier Research Center funded by the U.S. Department of Energy, Office of Science, Basic Energy Sciences under Award # DESC0001061, and the DOE Office of Nuclear Energy's Nuclear Energy University Programs.

References

[1] C. D. Zuo, T. H. Lee, S. J. Song, L. Chen, S. E. Dorris, U. Balachandran, M. L. Liu, Electrochem. Solid-State lett.2005, 8, J35.

[2] C. D. Zuo, S. E. Dorris, U. Balachandran, M. L. Liu, Chem. Mater.2006, 18, 4647.

[3] K. D. Kreuer, Ann. Rev. Mater. Res.2003, 33, 333.

[4] S. M. Haile, Y. Yamazaki, R. Hernandez-Sanchez, Chem. Mater.2009, 21, 2755.

Caption

Figure 1. (a) surface SEM image of sintered Ni-BZY membrane after thermal etching, (b) cross-section SEM image of sintered Ni-BZY membrane.

While commercialization of PEFCs has been started in residential application and is planned in automotive application, detailed understanding of degradation phenomena is essential. A large part of PEFC degradation, especially in automotive applications, is known to be related to carbon support degradation, mainly caused by carbon corrosion and related phenomena. During the fuel cell start-stop cycles, the potential of the cathode can reach up to 1.5 V (versus reversible hydrogen electrode, RHE). Fluctuation of cell voltage up to such higher potentials can cause oxidation-induced carbon support corrosion especially for cathode electrocatalysts [1-3].

In this tutorial talk, after summarizing typical degradation mechanisms with the state-of-the-art carbon black support, various efforts to develop alternative electrocatalyst support materials are reviewed. Table 1 compiles possible electrocatalyst support materials considered recently. There are three possible strategic directions in developing alternative durable supports as shown in Fig. 1: (1) graphitized carbon black, (2) alternative carbon support, and (3) alternative non-carbon support.

(1) Graphitized carbon black

While carbon is thermochemical unstable under the PEFC cathode condition, carbon corrosion can be kinetically slow for graphitized carbon black [4]. As graphite is relatively stable, so that higher temperature heat-treatment to increase the degree of graphitization is a useful procedure to ensure longer-term durability.

(2) Alternative carbon support

Various nanostructured carbon-based materials, such as mesoporous carbon [5] and graphene [6] may be applicable to design nanostructure of electrocatalysts. Generally speaking, graphene-like surface is stable against carbon corrosion, but Pt particles on the graphene surface tend to agglomerate each other [7]. More defective surfaces are suitable to stabilize Pt particles but such surface defects could be a site where carbon corrosion may start.

(3) Alternative non-carbon support such as oxides

As alternatives to the conventional carbon black electrocatalyst support, conductive oxides are mainly considered as carbon-free catalyst support materials [8-10]. Thermochemical calculations have revealed, as shown in Fig. 2 [8], that several oxides could be stable under the PEFC cathode conditions.

As an example, Figure 3 shows the I-V curves before and after such start-stop cycle tests for an MEA with the cathode electrocatalyst layer consisting of the Pt/Sn_0.98Nb_0.02O₂ mixed with 5wt % vapor grown carbon fibers (VGCF-H) as the electron-conductive filler. Start-stop cycle degradation was almost negligible up to 60,000 cycles. This result confirms that Nb-doped SnO₂ can be an alternative electrocatalyst support material for e.g. fuel cell vehicles (FCVs) which are suffered from the voltage cycling up to a high potential. Materials design principles for PEFC electrocatalyst supports will be discussed.

Acknowledgement

Financial support by the Grant-in-Aid for Scientific Research (S) (No. 23226015), JSPS Japan, is gratefully acknowledged.

References

[1] C. A. Reiser, L. Bregoli, T. W. Patterson, J. S. Yi, J. D. Yang, M. L. Perry, and T. D. Jarvi, Electrochem.Solid-State Lett., 8, A273 (2005).

[2] L. M. Roen, C. H. Paik, and T. D. Jarvi, Electrochem. Solid-State Lett., 7, A19 (2004).

[3] A. Ohma, K. Shinohara, A. Iiyama, T. Yoshida, and A. Daimaru, ECS Trans., 41(1), 775 (2011).

[4] X.-J. Zhao, A. Hayashi, Z. Noda, K. Kimijima, I. Yagi, K. Sasaki, Electrochim. Acta, 97, 33 (2013).

[5] A. Hayashi, H. Notsu, K. Kimijima, J. Miyamoto, I. Yagi, Electrochim. Acta, 53 (21), 6117 (2008).

[6] J.-F. Liu, K. Sasaki, and S. M. Lyth, ECS Trans., 58 (1), 1751 (2013).

[7] K. Sasaki, K. Shinya, S. Tanaka, Y. Kawazoe, T. Kuroki, K. Takata, H. Kusaba, and Y. Teraoka, Mater. Res. Soc. Symp. Proc., 835, 241 (2005).

[8] K. Sasaki, F. Takasaki, Z. Noda, S. Hayashi, Y. Shiratori, and K. Ito, ECS Trans., 33(1), 473 (2010).

[9] A. Masao, S. Noda, F. Takasaki, K. Ito, and K. Sasaki, Electrochem. Solid-State Lett., 12 (9), B119 (2009).

[10] F. Takasaki, S. Matsuie, Y. Takabatake, Z. Noda, A. Hayashi, Y. Shiratori, K. Ito, and K. Sasaki, J. Electrochem. Soc., 158 (10), B1270 (2011).

Introduction

Solid oxide fuel cell (SOFC) is the promising type of fuel cells with e.g. high electric efficiency and fuel flexibility. Durability and reliability are the most important technological issues on such an early commercialization stage. In our research group, we are focusing on the chemical degradation of SOFCs, determining the lifetime of SOFCs operating at high temperatures using various kinds of practical fuels. In this paper, various extrinsic and intrinsic degradation phenomena are systematically classified and their degradation mechanisms are discussed [1-3].

Extrinsic chemical degradation phenomena

While the advantage of SOFCs is their fuel flexibility, impurity species can often flow into the SOFC system, causing degradation phenomena. Due to their high operational temperatures, species with a high vapor pressure can also evaporate and flow into the SOFC stacks from system components. In addition, the use of low-purity raw materials could cause impurity poisoning by the contaminants in such raw materials. Possible extrinsic degradation phenomena are summarized in Fig. 1, including mechanisms associated with surface adsorption/desorption in the case of sulfur at a relatively low concentration, but associated with accumulation and reaction product formation for other impurities.

Intrinsic chemical degradation phenomena

External impurities can be, in principle, removed by using a getter as a desulfurization reactor or by using fuels with better purity. However, chemical degradation associated with diffusion from neighboring components may become important in long-term operation. As compiled in Fig. 2, various intrinsic degradation phenomena have been revealed by evaluating SOFC samples after long-term tests beyond one thousand or several thousand hours. Future perspectives are also presented including the importance of long-term / cycle testing followed by detailed microstructural analysis.

Acknowledgements: We thank for the financial support by METI to establish Next-Generation Fuel Cell Research Center. Financial supported by the NEDO SOFC project is gratefully acknowledged.

References

[1] K. Sasaki et al., J. Power Sources, 196[22], 9130 (2011).

[2] K. Sasaki et al., J. Electrochem. Soc., 153 [11], A2023 (2006).

[3] K. Sasaki et al., ECS Trans., 57 [1], 315 (2013).

Over the past two decades, considerable work has been reported on solid oxide fuel cells (SOFC) [1-3]. While progress has been made, there has not been significant further gain in performance (at 800^oC), beyond that achieved about fifteen years ago (~2 Wcm^-2), despite using many different cathodes. Thus in many cases, the cathode may not be limiting the performance. To achieve further improvements in cell performance beyond what has been discussed in previous studies [4,5], it will require accurate identification of various polarization losses, their sources, and dependence on material and microstructural parameters, atmosphere and temperature. The majority of the modeling studies do not provide the required guidance as they rely on a number of often not so easily measurable parameters thus necessitating a somewhat arbitrary selection of such parameters.

The objective of this work is to estimate various polarization losses based on out-of-cell measurements made using micro-fabricated electrodes, microstructural measurements on cells, and cell electrochemical performance measurements using a parametric model. Such an approach may identify where the dominant losses are and how to further improve cell performance.

The minimum number of independent parameters necessary to model cell performance for an anode-supported SOFC with five distinct layers are 9 and they are: D^eff1_O2-N2 and D^eff2_O2-N2 , the effective binary diffusivities through the cathode current collector and the cathode functional layer, respectively; D^eff1_H2-H2Oand D^eff2_H2-H2O , the effective binary diffusivities through the anode support and the anode functional layer, respectively; cathode exchange current density (i_0^c); anode exchange current density ((i_0^a), cathode transfer coefficient (α_c), anode transfer coefficient (α_a), and ohmic resistance R_i. The parametric model is given by equation (1).

V(i)=E0-iR_i-(RT/2F)arcsinh(i/2i₀^c)-(RT/2F)arcsinh(i/2i₀^a)+(RT/2F)ln(p_H2'(i)p_H2O^o/(p_H2^op_H2O')+(RT/4F)ln(p_O2'(i)/p_O2^o) (1)

where p_O2^o and p_O2'(i) are the oxygen partial pressures just outside the cathode current collector and close to the cathode functional layer/electrolyte interface, respectively; p_H2^o and p_H2'(i) are the hydrogen partial pressures in the fuel just outside the anode support and close to the anode functional layer/electrolyte interface, respectively; p_H2O^o and p_H2O' are the water vapor partial pressures in the fuel just outside the anode and close to the anode functional layer/electrolyte interface, respectively. In equation (1), both transfer coefficients are assumed to be 0.5.

As seen in Fig. 1, at low current density, the most dominant contribution is cathodic activation polarization for a cathode grain size of 2 mm. However, beyond about 1.3 A/cm², the ohmic contribution is the dominant one, and underscores its importance.

In Figs. 2a to 2b, the cathode functional layer grain size is varied from 2 μm to 0.2 μm. This increases the maximum power density by 0.3 W/cm². However, if the ohmic resistance can be somehow reduced from 0.1 to 0.05 Ωcm², the maximum power density increases by 1.2 W/cm² (Figs. 2b and 2c) even in a thin electrolyte, anode-supported cell. This shows the profound role of ohmic contribution on cell performance. By contrast, a change from LSM/YSZ to Pt/YSZ only leads to a modest increase in maximum power density of about 0.4 W/cm² (Figs. 2c and 2d). This is consistent with the observation that cells with several very different cathodes exhibit similar performance. Further work is thus required in lowering the ohmic contribution. Many cathodes that have been developed over the years appear quite satisfactory and not limiting cell performance for operation at 800^oC.

Funded by DOE EFRC Grant Number DE-SC0001061 as a flow-through from the University of South Carolina.

References:

1. A. V. Virkar, J. Power Sources, 1478-31 (2005).

2. A. V. Virkar, J. Power Sources, 154324-325 (2006).

3. R. Radhakrishnan, A. V. Virkar, and S. C. Singhal, J. Electrochem. Soc., 152 (1) A210-A218 (2005).

4. J-W. Kim, A. V. Virkar, K-Z. Fung, K. Mehta, and A. V. Virkar, J. Electrochem. Soc., 146(1) 69-78 (1999).

5. F. Zhao and A. V. Virkar, J. Power Sources, 14179-95 (2005).

Figure lengends:

Fig. 1: Calculated polarizations as a function of current density for cathode grain size (d) = 2 μm, ohmic resistance (Ri) = 0.1 Ωcm², anodic effective charge transfer resistance = 0.015 Ωcm², and LSM/YSZ as the cathode material.

Fig. 2: calculated voltage and power density as a function of current density for : (a) d = 2 μm, Ri = 0.1 Ωcm², cathode LSM/ l, (b) d = 0. 2 μm, Ri = 0.1 Ωcm², cathode LSM/YSZ, (c) d = 0.2 μm, Ri = 0.05 Ωcm², and cathode LSM/YSZ, (d) d = 0.2 μm, Ri = 0.05 Ωcm², and cathode Pt/YSZ. Anodic charge transfer resistance is fixed as 0.015 Ωcm² for all of the four plots.

Solid oxide fuel cells (SOFC) are connected in series to stacks by metal interconnectors (MIC). Hereby, the potentially high performance of anode-supported cells (ASC) is significantly reduced. [1,2]. Conclusively, gas diffusion polarisation at the mixed conducting cathode contributes close to the sum of all ohmic losses to the overall polarization, both controlling stack performance.

In order to investigate the matter further, a 2D FEM repeat unit model was developed, which accounts to ohmic and polarisation loss on a stack level [3]. In the model the physical processes i) gas diffusion in the porous electrodes, ii) electric /ionic conduction in the electrodes and electrolyte as well as iii) the electrochemical electrode reactions are incorporated. Electrode microstructure is implemented via 3D reconstruction of FIB-tomography images [5,6]. Electrode kinetic parameters are determined via EIS evaluations [4]. Performance predicted by the model was validated with the help of current/voltage-characteristics measured on high performance anode supported cells.

The model calculations demonstrate the interdependence of cathode material parameters (thickness, porosity, conductivity, electrode kinetics) and MIC-design, and widen the understanding on their influence on stack performance. The results show i) that an optimal MIC-design depends on the cathode material parameters and ii) a well-chosen cathode thickness increases the overall power output.

[1] L. Blum, W. A. Meulenberg, H. Nabielek, R. Steinberger-Wilckens, Worldwide SOFC Technology Overview and Benchmark, International Journal of Applied Ceramic Technology2 pp. 482-492 (2005).

[2] M. Kornely, A. Leonide, A. Weber, E. Ivers-Tiffée, Journal of Power Sources196 (17), pp. 7209-7216 (2011).

[3] H. Geisler, M. Kornely, A. Weber, E. Ivers-Tiffée, ECS Trans. 57, pp. 2871-2881 (2013).

[4] A. Leonide, Y. Apel, E. Ivers-Tiffée, ECS Trans. 19, pp. 81-109 (2009)

[5] J. Joos, T. Carraro, A. Weber, E. Ivers-Tiffée, J. Power Sources 196, pp. 7302-7307 (2011).

[6] J. Joos, M. Ender, I. Rotscholl, N. H. Menzler, E. Ivers-Tiffée, J. Power Sources 246, pp.819-830 (2014)

A solid electrolyte will always possess a finite electronic conductivity, in particular electrolytes like doped ceria that easily get reduced and become mixed ionic and electronic conductors. This given rise too high leak currents through the solid oxide cell (SOC). Especially, problems have been observed for ceria based electrolytes, but also in case of solid oxide electrolyser cells (SOEC) with yttria stabilized zirconia (YSZ) big electronic leak currents have been observed for very high overvoltages on one or both electrodes.

Furthermore, it is important to realize that the potential gradient driving the O^2-ions is not the Fermi potential, which is the potential of the electrons, but the Galvani potential (or inner potential) (1). The concepts of potentials describing the electrical situation of a solid electrolyte is shown i Fig. 1, and an example of the Fermi potential (π) and Galvani potential (φ) profiles are shown in Fig. 2. The Fermi potential will be affected directly by the Galvani potential, whereas the Galvani potential need not necessarily be affected by the Fermi potential because the concentration of "free" electrons may be very low.

The paper gives illustrative examples at various temperatures and operation conditions. Furthermore, the situation within cells based on gadolinia doped ceria (CGO) and on YSZ electrolytes are compared. Finally, it is discussed how the Fermi potential and electron conductivity will be affected by the various parameters including operation conditions.

Reference

[1] T. Jacobsen and M. Mogensen, ECS Transactions, 13 (no.26), 259 (2008).

Solid oxide fuel cells (SOFCs) are ideally suited for environmentally benign conversion of chemical energy in hydrocarbons to electricity. While medium scale SOFC power systems in the hundreds of kWe have been demonstrated, larger scale application of such systems have been hobbled by still too high cost. Reducing the manufacturing costs of single cells is an important goal as one step towards commercialization. Previously, we demonstrated a simplified process to manufacture the complex multilayer architecture of anode supported single cells in a single high temperature firing step, after deposition of the anode active layer, electrolyte, cathode active layer, and cathode current collector layers serially by screen printing on to an anode substrate fabrication by a process known as high shear compaction. In this paper, we present a novel manufacturing process in which the basic green-state cell is fabricated using the same steps as before, but in which the high temperature conventional sintering step is replaced by a technique known as flash sintering which reduces time at temperature by up to an order of magnitude, thereby resulting in significant savings in time and electrical energy expended. We present details of this new process, cross-sectional microstructures of cells fabricated using this technique, and compare the power densities of the cells processed by flash sintering with the prior co-fired process.

Strontium-doped lanthanum cobaltite is one of the most promising cathode candidates for intermediate temperature solid oxide fuel cells. Recently, the hetero-interfaces between the perovskite (La,Sr)CoO₃ (LSC₁₁₃) and the Ruddlesden-Popper (La,Sr)₂CoO₄ (LSC₂₁₄) phases have shown highly enhanced oxygen exchange kinetics. This observation offers the potential for a new class of cathode structure design, either via multilayering or by forming vertical heteroepitaxial nanocomposites, with high densities of these special interfaces. In this work, nanoscale multilayers and vertically aligned nanocomposites of LSC_113/214 have been prepared by pulsed laser deposition. Their surface chemistry was characterized by Auger Electron Spectroscopy with high spatial resolution, and their surface electronic structure was probed by scanning tunneling microscopy at elevated temperatures and in oxygen gas. The intimate contact of LSC₂₁₄ to the LSC₁₁₃ phase, rendered the charge transfer on the LSC₂₁₄ surface to be more facile, contributing to enhanced oxygen reduction reactivity near the LSC_113/214 interfacial regions intersecting the surface. The enhancement of the ORR kinetics is also confirmed by EIS measurement at relatively low temperatures below 400 ^oC. The surface chemistry at higher temperatures ages by time and hides the improvements that arise from the coupling of these two phases through their interface. These observations lead to a novel means for SOFC cathode fabrication with enhanced performance and provide guidelines for identifying other promising cathode candidates.

Surface chemistry and reactivity of solid oxide fuel cell (SOFC) and electrolyzer (SOEC) electrodes play a key role for transport and exchange processes, such as the oxygen reduction reaction in SOFC cathodes, and drive electrochemical performance and durability. In order to identify drivers and problems, it is necessary to probe electron and ion transfer and transport processes at electrolyte/electrode interfaces and surfaces and build broad fundamental understanding. This knowledge can then be used to drive the innovation of novel, more powerful electrode materials and cell designs.

We used a combination of spatially resolved scanning photoelectron microscopy and electrochemical measurements to study in situ the surface chemistry of cathode perovskite catalysts and electrolyte in electrochemical model cells during high temperature operation. The electrodes were made of screen-printed (La1-xSrx)MeO3 with Mn, Fe, Co, Ni as transition metal Me and x = 0 – 0.4 and covered a wide range in ionic and electric conductivity from pure ion to mixed conduction. The electrolyte was a thin yttrium-stabilized zirconia sheet. We studied the surface chemistry of oxide catalyst and electrolyte at temperatures between 400 and 700°C, oxygen pressure (10-7bar – 1mbar) and also in humidity and/or hydrocarbon containing environment and demonstrated systematic changes in the perovskite surface termination with oxygen chemical potential. Oxidizing environment was found to promote transition metal termination and presence of various oxygen surface species, while reducing environment drove segregation of strontium to the surface and suppressed the level of surface oxygen.

We also showed that this simple segregation behavior was dramatically changed in presence of impurities. Silica segregated to the surface, forming a glassy, insulating surface layer that slowed down the oxygen incorporation reaction and produced a strong degradation of the cell performance. While the degradation of Ni/NiO electrodes could be reversed by an applied cell voltage, no such remedy was observed for perovskite electrodes.

In our in-situ set up, oxygen in- or excorporation could be driven by an applied electric field across the cell. Oxygen ion flux and electric field caused dynamic changes of catalyst and electrolyte surface chemistry, including redox reaction, changes in surface segregation and long range surface diffusion at the electrode surface. The electrochemical response of the model cells was interpreted in terms of reactions and reaction steps that matched the spectroscopic observations, thus constructing new understanding of the processes in operating electrodes.

Introduction

Mixed ionic-electronic conductors as LSCF (La_0.58Sr_0.4Co_0.2Fe_0.8O_3-δ) and cermets like Ni/8YSZ (8 mol% Y₂O₃ stabilized ZrO₂) are well-established cathode and anode materials for solid oxide fuel cells. At operating temperatures of 600 °C or below, electrode polarization losses as well as ohmic losses in the solid electrolyte rise steeply. Thereby, anode-supported cells (ASC), proving highest performance and excellent durability at temperatures > 700 °C, are unsuitable in their actual "standard-type" design.

In this contribution, the solid-gas electrochemistry at both electrode/electrolyte interfaces of a standard ASC was significantly improved by modifications using nanotechnology:

• A thin-film 8YSZ electrolyte of ~1.5 µm, made by a wet-chemical method, in combination with a dense Ce_0.8Gd_0.2O_1.9 (CGO) thin-film buffer layer efficiently reduced the ohmic loss contribution by 80% [1].

• A nano-scaled La_0.6Sr_0.4CoO_3-δ (LSC) thin-film cathode of 200 nm, deposited by metal organic deposition (MOD), cut the area specific resistance of the two-layer cathode of 30 µm down by two orders of magnitude, compared to µm-scaled cathode structures [2].

• A nano-scaled Ni/8YSZ anode layer of 10 to 100nm, grown in operando by a reverse current treatment (RCT) [3], lowered the area specific resistance of a standard µm scaled anode by up to 40% [4].

Experimental

Anode-supported half cells were manufactured by Forschungszentrum Jülich. They consist of a regular anode support (thickness 0.5 mm, NiO/8YSZ), with a µm-scaled anode functional layer (AFL, thickness 7 µm, NiO/8YSZ) and an ultra-thin-film electrolyte (thickness 1.5 µm, 8YSZ). Onto this electrolyte, a dense CGO layer (thickness 850 nm) was deposited by physical vapor deposition (PVD) to prevent secondary phase formation. At IWE, the nano- scaled LSC thin-film cathode (thickness 200 nm) was deposited by spin-coating of a metal-organic solution on top. This required a modified fabrication route, including anode reduction at 800 °C prior to cathode deposition and two consecutive heat treatments (170 °C and 650 °C) of the LSC thin film. Subsequently, the cathode was completed by screen printing a µm-scaled LSC current collector on top. The nano-scaled anode/electrolyte interface was evolved from a reverse current treatment during cell operation at 600 °C (~1% H₂O in H₂).

Results and Discussion

The nanostructures and phase compositions of both electrode/electrolyte interfaces were studied by high-resolution and scanning electron microscopy methods, among others. MOD derived LSC thin-film cathodes typically consist of a heterointerface made of the perowskite La_0.6Sr_0.4CoO_3-δ and the Ruddlesden-Popper type phase (La,Sr)₂CoO_4±δ , which, in combination with the nanoscaled microstructure, leads to an extremely enhanced oxygen surface exchange [2]. The Ni/YSZ anode had undergone a complete rearrangement in the close vicinity to the electrolyte, resulting in an enormous increase of triple phase boundaries on the nanoscale [3]. A reaction model explaining the solid-solid transformations during the reverse current treatment, which was set up in model experiments, is presented. Analysis by electrochemical impedance spectroscopy (EIS) and current/voltage (C/V) curves revealed an area specific resistance (ASR) of 359 mΩcm² at 600 °C and 60% H₂O in H₂and a power density of 900 mW/cm² at 700 mV in dry hydrogen atmosphere. This performance is by 400% better than measured for a standard-type ASC with a LSCF cathode, which underlines the potential of introducing nanotechnology for mixed conducting electrodes in SOFC.

Conclusions

This study has created a better understanding, how custom-tailored nanostructures potentially support the performance of mixed-conducting electrodes for SOFC at ≤ 600 °C. Despite alternative approaches reported in literature [5, 6], quite simple nanotechnology strategies were applied, which are also applicable on the larger scale, so that we see great potential in this "nano" ASC design.

References

[1] F. Han, R. Mücke, T. Van Gestel, A. Leonide, N. H. Menzler, H. P. Buchkremer, D. Stöver, J. Power Sources 218 (2012) 157.

[2] J. Hayd, E. Ivers-Tiffée, J. Electrochem. Soc. 160 (2013) F1197.

[3] J. Szász, D. Klotz, H. Störmer, D. Gerthsen, E. Ivers-Tiffée, ECS Trans. 57 (2013) 1469.

[4] D. Klotz, B. Butz, A. Leonide, J. Hayd, D. Gerthsen, E. Ivers-Tiffée, J. Electrochem. Soc. 158 (2011) B587.

[5] A. Evans, A. Bieberle-Hütter, J. L. M. Rupp, L. J. Gauckler, J. Power Sources 194 (2009) 119.

[6] Y. Takagi, S. Adam, S. Ramanathan, J. Power Sources 217 (2012) 543.

Mixed oxide-ion and electron conducting ceramic material La₂NiO_4+d(LNO) have been reported to exhibit impressive oxide ion and p-type electronic conductivity, and high catalytic activity which make them an excellent component of some electrochemical devices such as fuel cell cathode, oxygen permeation membrane. The hyperstoichiometric LNO oxides show high oxygen diffusion coefficient and high oxygen ion conductivity, which are attributes of their crystal structure.

The extent of oxygen nonstoichiometry significantly affects the electrochemical properties of LNO systems. The nonstoichiometric oxygen content may be varied by aliovalent doping or by substituting the host cation with smaller size cations[1,2]. The conduction mechanism in LNO-based system has been explained on the basis of both, the band conduction by delocalized electrons and polaron-hopping by localized electrons[3,4].

The objective of this work is to study the effect of a donor cation (Al³⁺) on oxygen nonstoichiometry and resultant thermodynamic properties of LNO system. In this work we have prepared a non-transition metal cation Al³⁺-doped LNO system, La₂Ni_0.95Al_0.05O_4.025+d, and presented an equilibrium defect model for this system. The variation of oxygen nonstoichiometry was measured by coulometric titration and the absolute value of oxygen nonstoichiometry was calculated from thermogravimetric analysis. Various thermodynamic quantities for LNAO were calculated from the experimental data and were compared with those for LNO in our previous work.

Fig.1 shows pO₂ dependence of oxygen nonstoichiometry at different temperature. As can be seen, the values of δ increased with decreasing temperature and with increasing pO₂. The inset of Fig. 1 shows the relationship of log δ vs. log pO₂, as can be seen, at isothermal condition the dependence of δ on the pO₂exponent (m) is slightly less than 1/6. It is clear from the above observation that the oxygen exponent decreases with increasing oxygen activity, suggesting a positive deviation of the LNAO system from the ideal solution behavior[4].

The best estimated values of N_v obtained from fitting the Eq. (1) with fitting parameters K_ox, K_f, and N_v to the δ vs. log(pO2) data in Fig.(1). The density of states of the valence band varies from 1.87x10²⁰ to 2.55x10²⁰ in 800-1000 ^oC range. And , the effective mass m_h^*of holes is 1.02 - 1.21 times the rest mass m_o. This indicates the occurrence of band-like conduction and allows the effect of conduction by a small degree of polaron hopping to be ignored. The behavior of holes starts deviates from that of ideal solution at a very low δ value and r_h^• increases with the increasing δ and reaches to ~7 at r_h^•≈0.08 and T=900°C. The calculation of excess partial molar quantities showed that incorporation of interstitial oxygen is less favorable in La₂Ni_0.95Al_0.05O_{4.025 + δ} in comparison to La₂NiO_4 + δ.

Reference

• Nakamura T, Yashiro K, Sato K, Mizusaki, J. Solid State Ionics, 180, 368 (2009).

• Naumovich EN, Patrakeev MV, Kharton VV, Yaremchenko AA, Logvinovich DI, Marques FMB. Solid State Sciences, 7,1353 (2005).

• H. S. Kim, and H. I. Yoo, Solid State Ionics, 232, 129 (2013).

• H. S. Kim, and H. I. Yoo, Phys. Chem. Chem. Phys., 12, 4704 (2010).

Many oxides exhibit advantageous properties in energy conversion and storage applications, such as solid oxide fuel cells (SOFCs), heterogeneous catalysis, photoelectrochemical cells, and batteries, due in part to their high temperature stability, corrosion resistance, and dopant cation solubility, respectively. In particular, mixed ionic and electronic conducting oxides have been demonstrated to aid in increasing reaction kinetics at electrodes in SOFCs due to the simultaneous presence of electrons and oxygen vacancies. With changes in oxygen partial pressure and temperature during normal operation operation, the defect concentration and hence transport characteristics of MIEC oxides inevitably change. In addition, defects are often coupled to a significant lattice dilation, known as chemical expansion, which under some circumstances leads to mechanical failure. Furthermore, as length scales are reduced, the energetics for electron and vacancy formation have been shown to be modified. In this presentation, recent work examining the the interplay of electro-chemo-mechanical properties of bulk ceramic to thin film size scales will be discussed. This work was facilitated by using the model material system, (Pr,Ce)O_2-δ (PCO), which, due to the ease of reduction of Pr from 4+ to 3+ valence, exhibits significant MIEC behavior at relatively high oxygen pressures (e.g. air). Impedance spectroscopy was used to study both the oxygen exchange resistance and oxygen content (via chemical capacitance) of cells consisting of PCO thin films deposited on yttria stabilized zirconia substrates. When compared to defect formation energies derived from a detailed thermogravimetric and conductivity model for bulk PCO, films were found to exhibit enhanced reducibility. This phenomenon was further verified using an in situ optical transmission measurement of films, coupled with chemical capacitance, which correlated the amount of absorbing Pr⁴⁺ centers to oxygen vacancy concentration. Additionally, oxygen exchange probed by optical transmission relaxation measurements on "bare" films exhibited slower kinetics as compared to corresponding electrochemical measurements. The role of metal electrodes as well as impurities in modifying oxygen exchange rate will be discussed, as well as new means to improve oxygen exchange of the aged "bare" film surface. Lastly, thermo-chemical expansion measurements of bulk and thin film PCO samples will be presented with discussion related to the atomistic role of individual point defects and defect association.

La_0.6Sr_0.4CoO₃ has been studied as an SOFC cathode because of its high mixed ionic electronic conductivity. The mixed conductivity makes the zone of oxygen reduction reaction extended from triple-phase boundary to two-phase boundary. The rate limiting step of the cathode reaction is oxygen reduction reaction on two-phase boundary (1). Sase et al. reported that the oxygen reduction is activated at hetero-interface of La_0.6Sr_0.4CoO₃ and La_1.5Sr_0.5CoO₄ (2). The hetero- interface effect leads to development of higher performance electrode. However, the detail mechanism is still not clear.

This study aims to elucidate the mechanism of the hetero-interface effect by the relationship between the electrode performance and the surface chemistry.

Thin-film electrodes were fabricated on the electrolyte substrate of Ce_0.9Gd_0.1O_1.95 (10GDC) by pulse laser deposition (PLD) method. The two-layer electrodes were made: the first layer was La_0.6Sr_0.4CoO₃ (LSC40); the second layer, La_1.5Sr_0.5CoO₄(LSC155), was prepared exactly onto the first layer. Additionally single layer electrodes of LSC155 and LSC40 were also formed on the same 10GDC electrolyte to evaluate the hetero- interface effect. Several samples were prepared with changing thickness of LSC155.

A micro probe apparatus equipped with a potentio-galvano stat and a frequency response analyzer was employed for the electrochemical measurement. The measured electrode could be selected by manipulating probes of the micro prober. This setup enables to measure four electrodes simultaneously without changing the sample and its thermal history. The thin-film electrodes, the porous Pt on the side surface, and the porous Pt on the opposite side of the sample were used as working, reference, and counter electrodes, respectively. The measurement conditions were at 873, 973, and 1073 K, in air.

The electrode reaction resistances of the hetero-interface electrodes were measured by an impedance measurement. It was expected that the resistance of the double layer electrodes should become larger with increase in thickness of LSC155 due to longer ionic diffusion pathway. However, the measured electrode resistances did not depend much on LSC155 thickness. Electrode resistances of most of the double-layered electrodes were smaller than those of the LSC155 single-phase electrode in spite of LSC155 thickness. This phenomenon implies the electrochemical properties of LSC155 film on LSC40 might be modified and different from those of LSC155 film on a substrate

Focusing on the LSC40 single-phase electrode, the electrode resistance increased from 873 to 973 K. It may be the influence of rearrangement of cations, which can occur over 923 K: the deposition temperature of the electrodes. The resistances of the layered electrodes were smaller than that of LSC40 single-phase electrode. Element distribution was evaluated by EDX analysis. The enrichment of strontium was observed at the surface grains of the degraded LSC40. The concentrations of lanthanum and cobalt were lower at the surface grains. This means the surface composition changed from that of bulk. Therefore, the segregation of strontium on the surface of LSC40 may degrade the electrode performance.

This research was supported by JST-PREST and JST-CREST project.

References

1. T. Kawada, K. Masuda, J. Suzuki, A. Kaimai, K. Kawamura, Y. Nigara, J. Mizusaki, H. Yugami, H. Arashi, N. Sakai, H. Yokokawa, Solid State Ionics, 271–279, 121 (1999).

2. M. Sase, K. Yashiro, K. Sato, J. Mizusaki, T. Kawada, N. Sakai, K. Yamaji, T. Horita, H. Yokokawa, Solid State Ionics, 1843–1852, 178 (2008).

While impressive solid oxide fuel cell (SOFC) performance has been achieved, durability under "real world" conditions is still an issue for commercial deployment. In particular cathode exposure to H₂O and CO₂ results in long-term performance degradation issues. Therefore, we have embarked on a multi-faceted fundamental investigation of the effect of these contaminants on cathode degradation mechanisms in order to establish cathode composition/structures and operational conditions to enhance cathode durability Using a dual Focused Ion Beam (FIB)/SEM) we are quantifying in 3-D the microstructural changes of cathode before and after the onset of cathode performance degradation. This includes changes in TPB density, phase-connectivity, and tortuosity, as well as tertiary phase formation. This is linked to heterogeneous catalysis methods to elucidate the cathode oxygen reduction reaction (ORR) mechanism to determine how H₂O and CO₂ affect the ORR as a function of temperature, time, and composition. By use of in-situ¹⁸O-isotope exchange of labeled contaminants we will determine whether oxygen incorporated in the lattice of LSM and LSCF, and their composites with YSZ, originated from ambient O₂ or the contaminant as well as intermediate adsorbed species and mechanisms that lead to degradation. The results will be used to develop a cohesive and overarching theory that explains the microstructural and compositional cathode performance degradation mechanisms.

High operational temperatures needed for solid oxide fuel cells (SOFCs) are required by the large activation energy of the oxide diffusion through the electrolyte. These high temperatures, in the range of 800-1000 °C, accelerate cell degradation and severely reduce the number of materials suitable for construction. Employing newly developed, more highly conducting electrolytes such as scandia-stabilized zirconia (ScSZ) enable SOFCs to operate at lower temperatures without sacrificing power (1). Intermediate temperature (IT) solid oxide fuel cells, operating in the 500-750 °C range, permit non-ceramic and non-metallic components to be used in the fuel cell stack, reducing costs and enabling a larger variety of stack geometries (2). However, the high catalytic activity of nickel still causes unwanted carbon deposition on the anode surface (3). This carbon buildup can cause premature cell degradation as well as reduce the power density of the fuel cell.

Chronopotentiometry, a technique that couples real time electrochemical monitoring of cell voltage with synchronous in situRaman spectroscopy, provides insight into the mechanisms of carbon deposition and removal as well as the effect deposited carbon has on fuel cell performance (3). This technique can quantify the carbon growth with < 2 second temporal resolution and 1 μm spatial resolution across the anode surface.

Figure 1: Raman "G" peak growth at 675 °C observed in situfrom a cell operating under methane with a ScSZ electrolyte supported membrane electrode assembly.

In this work, we examine qualitative differences in carbon buildup (Fig. 1) between yittria-stabilized zirconia (YSZ) and ScSZ supported solid oxide fuel cells with a Ni-YSZ cermet anode exposed to methane at 675 °C and 725 °C. We also discuss the differences in the linear sweep voltammetry (LSV) between cells running under methane and hydrogen. (Fig. 2)

Figure 2: (Top) Comparison of carbon growth at different polarizations of YSZ and ScSZ SOFCs operating at 725 °C under methane fuel. (a) ScSZ OCV (b) ScSZ 75% I_Max (c) YSZ OCV (d) YSZ 75% I_Max (Bottom) Comparison of LSV traces running under H₂ and CH₄. The "swing" in the CH₄ trace, indicating a change in carbon removal mechanism can be seen at 150 mA (4).

Observed differences between the amounts of carbon formed for YSZ and ScSZ fuel cells at high and low temperatures and different polarizations (Fig. 2) implicate a role played by the electrolyte in the formation of carbon on the anode. Different carbon removal mechanisms have fundamentally different Nernst potentials, and when coupled with Raman spectroscopy, allow us to identify not only the relative amounts, but also the type of carbon formed.

References

1. S.P.S. Badwal, F.T. Ciacchi and D. Milosevic, Solid State Ionics, 136-137 91 (2000)

2. Brett et al., Chemical Society Reviews, 1568-1578 37 (2008)

3. Kirtley et al., Anal. Chem., 9745-53 84 (2012)

4. A.J. Bard, and L.R. Faulkner. Electrochemical Methods, p. 226-47 John Wiley & Sons, Inc., New York (2000).

Abstract:

New perovskite oxides La_0.75Sr_0.25Mn_0.5Cr_0.5-xCu_xO_3-δ (LSMCr_0.5-xCu_x, x=0, 0.05, 0.10, 0.20) have been developed as both cathode and anode for symmetric single solid oxide fuel cells (SOFCs). The as-prepared LSMCr_0.5-xCu_x powders are rhombohedral perovskite phase, and their properties are remarkably affected by Cu content (x). For example, the particle sizes and electrical conductivities of LSMCr_0.5-xCu_x increase with x both in air and in 5% H₂-Ar atmosphere. Their p-type conductivity shows lower values under 5% H₂-Ar reducing atmosphere than in air due to the decrease of the ratio of Mn⁴⁺/Mn³⁺ cation pairs through the reduction of Mn⁴⁺. Among of the oxides, LSMCr_0.45Cu_0.05 shows the best anodic and cathodic performance in symmetrical half cells. The maximum power density of a LSMCr_0.45Cu_0.05 / ScSZ / LSMCr_0.45Cu_0.05 symmetric single SOFC reaches as high as 451.2 mW·cm^-2 in wet H₂ as fuel at 900 ^oC.

Keywords: solid oxide fuel cell; symmetric cell; LSCM; Cu substitution

Acknowledgements:

This research was financially supported by the Natural Science Foundation of China (21173147), 973 Program of China (2014CB239700). Thanks were also given to the support of National Key Laboratory of Metallic Matrix Composite Material.

*Corresponding author. Tel.: +86 21 34206255; Fax: +86 21 54741297.

E-mail address: yimei@sjtu.edu.cn (Y.-M. Yin).

Electrochemical promotion of catalysis (EPOC) is a promising method for enhancing catalytic activity through the application of a small electrical stimulus between the catalyst-working and counter electrode deposited on a solid electrolyte ¹. The electronic properties of the catalyst can be modified resulting in a change in catalytic activity. In the case of yttria-stabilized zirconia (YSZ) as a solid electrolyte, the addition or removal of O^2- species on the catalyst surface can be controlled in situ depending on the specified reaction conditions. Fully reversible and "permanent" or "persistent" EPOC has been reported for more than 70 various catalytic systems ¹. In reversible EPOC experiments, the reaction rate returns to its initial value after the electrical stimulus is interrupted. For permanent EPOC (P-EPOC), the reaction rate remains at a higher value than the initial open circuit value ^2,3. Despite receiving much attention, this phenomenon has not yet reached commercial application. One of the main technical factors preventing such development is the use of thick film catalysts with low surface areas and high material costs ⁴.

Ceria, CeO₂, is a mixed ionic-electronic conducting (MIEC) material that conducts O^2- due to oxygen vacancies in the crystallographic structure in addition to conducting electrons at elevated temperatures. Furthermore, due to its non-stoichiometry, CeO₂ has the ability to undergo conversion between Ce⁴⁺ and Ce³⁺ quite easily ⁵. These properties make the use of ceria-containing catalysts of interest for many applications. In heterogeneous catalysis, Pt group metals deposited on CeO₂ show a metal-support interaction (MSI) effect associated with charge transfer between the two solids that are in contact. In EPOC studies, using a MIEC can also ensure electrical connectivity between highly dispersed nanoparticle catalysts ⁶.

In this study, electrochemical enhancement of catalytic activity of a low particle size (1.9 nm) ruthenium nanoparticles catalyst for ethylene oxidation was investigated. Ru nanoparticles, synthesized using a modified polyol reduction method, were supported on CeO₂ resulting in a 1 wt% Ru loading (RuNPs/CeO₂) (i.e., typical in heterogeneous catalysis studies).The highly dispersed RuNPs/CeO₂ catalyst powder was supported on a YSZ solid electrolyte in order to apply polarization. The discussion of this study includes the effect of the partial pressure of ethylene with constant partial pressure of oxygen, temperature, and applied positive and negative current on the catalytic activity of the RuNPs/CeO₂ catalyst as well as the role of the cerium redox state in the observed persistent effect. In addition, the catalytic properties of the RuNPs/CeO₂ catalyst is compared to that of larger Ru particles supported on CeO₂ (TD-Ru/CeO₂) (same metal loading) and blank CeO₂, both supported on a YSZ solid electrolyte.

Characterizations of the catalysts were carried out using TEM and SEM. In addition, XPS analysis was done for the RuNPs/CeO₂catalyst as prepared and "spent" (after the reaction).

Overall, it was observed that only the RuNPs/CeO₂ catalyst could be catalytically enhanced, showing a pronounced enhancement (up to 2.5 times) of the catalytic rate for negative polarization. The opposite effect was observed for positive polarization. This effect of both positive and negative polarization is illustrated by Fig.1.

Fig. 1.Transient effect of current application for C₂H₄ oxidation over Ru/CeO₂ on YSZ electrolyte at 350°C for an applied current of -2 μA for 4 hours and +2 μA for 6 hours (0.012 kPa C₂H₄).

Apparent Faradaic efficiencies up to 100 were also determined, indicating a non-Faradaic effect. In addition, a persistent effect was observed, showing stability up to 16 hours after current interruption. The modification of the cerium oxidation state (i.e., reduction from Ce⁴⁺ to Ce³⁺) is proposed to enhancethe catalytic performance of the Ru nanoparticles. This is due to the presence of more oxygen vacancies in the ceria interlayer causing a stronger metal-support interaction.These results demonstrate the feasibility of in-situ modification of the metal support-interaction between Ru nanoparticles and CeO₂catalytic support.

References

1. C. G. Vayenas, S. Bebelis, C. Pliangos, S. Brosda, and D. Tsiplakides, Electrochemical Activation of Catalysis: Promotion, Electrochemical Promotion, and Metal-Support Interactions, Kluwer Academic/Plenum Publishers, New York, (2001).

2. J. Nicole and C. Comninellis, J. Appl. Electrochem., 28, 223–226 (1998).

3. S. Wodiunig, V. Patsis, and C. Comninellis, Solid State Ionics, 137, 813–817 (2000).

4. D. Tsiplakides and S. Balomenou, Catal. Today, 146, 312–318 (2009).

5. A. Trovarelli, Ed., Catalysis by Ceria and Related Materials, Imperial College Press, London, (2002).

6. A. Kambolis et al., Electrochem. commun., 19, 5–8 (2012).

The fuel flexibility of solid oxide fuel cells (SOFC) is highly attractive for numerous technical applications that convert natural gas, kerosene or diesel reformate into electrical energy. For the operation with hydrocarbon fuels, there is a great deal of knowledge about the catalytic reforming chemistry in the SOFC anodes, whereas little is known about the solid/gas electrochemistry. This work reports a comprehensive understanding of the reaction mechanisms and transport pathways that determine the performance of syngas fueled Ni/YSZ anodes.

Based on our initial work, in which the individual loss mechanisms for syngas fueled Ni/YSZ anodes were identified via electrochemical impedance spectroscopy (EIS) [1], a schematic model of the reaction mechanisms and transport pathways is developed (Fig. 1). Since exclusively H₂ is electrochemically oxidized at the triple-phase boundary, the CO within the fuel is subsequently oxidized via the water-gas shift reaction [1]. The resulting gas transport properties within the anode substrate feature two transport pathways (H₂/H₂O and CO/CO₂), which are coupled by the water-gas shift reaction.

This schematic model is implemented in a transient finite element method (FEM) simulation [2]. The isothermal model represents the electrochemical fuel oxidation and the heterogeneous reforming chemistry on the catalytically active Ni-sites in terms of global kinetics. The multi component gas transport of the fuel gas species through the porous electrode structure is represented by the Maxwell-Stefan equation. Coherently calculating the complex species transport phenomena and the kinetics of the reforming chemistry, the model is capable to reproduce the multiple semicircles of the measured gas transport impedance. With the help of the FEM model, the characteristics of the measured impedance are explained in detail.

With the validation of the presented reaction diffusion network, an overall understanding of the loss mechanisms for syngas fueled Ni/YSZ anodes has been reported for the first time in literature [3]. With this knowledge, the impact of sulfur-poisoning on (i) the electrochemical fuel oxidation and on (ii) the heterogeneous reforming chemistry within syngas fueled Ni/YSZ anodes can be monitored separately [4].

References

[1] A. Kromp, A. Leonide, A. Weber, E. Ivers-Tiffée, J. Electrochem. Soc., 158 (8), B980-B986 (2011).

[2] A. Kromp, H. Geisler, A. Weber, E. Ivers-Tiffée, Electrochim. Acta 106, 418-424 (2013).

[3] A. Kromp, A. Weber, E. Ivers-Tiffée, ECS. Trans. 57 (1), 3063-3075 (2013).

[4] A. Kromp, S. Dierickx, A. Leonide, A. Weber, E. Ivers-Tiffée, J. Electrochem. Soc., 159 (5), B597-B601 (2012).

Introduction

Mixed Ionic and Electronic Conducting (MIEC) multi-cation oxide nano-particles have been used to enhance the performance of SOFC electrodes [1]. These SOFC MIEC nano-particles are often produced by thermally decomposing gelled MIEC precursor solutions infiltrated into porous ionic conducting (IC) scaffolds. Average SOFC nano-particle sizes produced by this thermal decomposition technique are typically 55 nm in diameter [1, 2], and result in SOFC operating temperatures (defined as the lowest temperature where the electrode polarization resistance, (R_P), is equal to 0.1 Ωcm²) in excess of 600°C. Since MIEC particle size directly correlates with the surface area available for oxygen incorporation (and hence SOFC electrode performance), the objective of this study was to determine whether it was possible to reduce MIEC nano-particle size through desiccation and/or sequential infiltration.

Experimental Methods

Heavily-infiltrated nano-composite La_0.6Sr_0.4Co_0.8Fe_0.2O₃ (LSCF) on Ce_0.9Gd_0.1O_1.95 (GDC) and LSCF-GDC on GDC nano-composite cathodes were prepared through multiple nitrate solution infiltrations into porous GDC scaffolds. For the LSCF on GDC samples, precursor LSCF solutions were gelled inside a GDC scaffold, desiccated, and fired at 700°C. For the LSCF-GDC on GDC samples, GDC precursor solutions were gelled inside a GDC scaffold and fired at 700°C prior to LSCF precursor solution infiltration, gelation, and firing at 700°C.

Results

As shown in Figure 1, average nano-particle sizes of 20 nm ± 5 nm (determined through scanning electron microscopy) were achieved through both desiccation (using CaCl₂ as a desiccant) and sequential infiltration (using 8.0 vol % of pre-existing nano-GDC as a decomposition catalyst). By altering the amount of desiccation it is possible to controllably adjust the LSCF nano-particle size from 20 nm ± 5 nm ( note the ± refers to the standard deviation of the actual particle sizes and is not a reflection of the scanning electron microscopy (SEM) resolution, which is ± 1 nm) to 60 nm ± 18 nm. Similarly, sequential infiltration can be used to adjust average nano-particle size from 20 nm ± 5 nm to 50 nm ± 18 nm.

As shown in Figures 2 & 3, the desiccation and sequential infiltration approaches progressively change precursor solution decomposition temperature peak positions. Figure 2 shows that desiccation shifts the strontium nitrate peak position to lower temperatures with stronger desiccants. Figure 3 demonstrates that as the nano-GDC loading level increases, the decomposition peak positions shift from 800°C with 0 vol % nano-GDC to 200°C with 50 vol % nano-GDC. (Volume percentage is the volume % of the IC scaffold porosity occupied by nano-particles). Through the adjustment of precursor gel decomposition peak position to lower temperatures smaller nano-particle sizes are achieved.

As shown in Figure 4, NCCs infiltrated with LSCF with 20 nm nano-particle diameters have R_P values of 0.1 Ωcm² at 565°C using drying/desiccation and NCCs using LSCF + GDC with 20 nm nano-particle diameters have R_P values of 0.1 Ωcm² at 540°C using sequential infiltration.

Conclusions

Two different approaches have been developed to influence precursor solution decomposition temperature and reduce average nano-particle sizes. Average MIEC nano-particle size were reduced to 20 nm and SOFC operating temperatures (i.e. the temperature at which R_P ≤ 0.1 Ωcm²) are 565°C with desiccation and 540°C with sequential infiltration.

Acknowledgements

This work was made possible through a Michigan State University faculty startup grant.

References

[1] J.D. Nicholas, S.A. Barnett, J. Electrochem. Soc, 157, B536-B541, 2010.

[2] Z. Zhan, D. Han, T. Wu, X. Ye, S. Wang, T. Wen, S. Cho, S.A. Barnett, R. Soc. Chem. Adv, 2, 4075-4078, 2012.

[3] J.D. Nicholas, L. Wang, A.V. Call, S.A. Barnett, Phys. Chem. Chem. Phys, 14, 15379-15392, 2012.

Losses from cathode polarization in solid oxide fuel cells (SOFC) limit their performance at intermediate temperatures. A fundamental understanding of the oxygen reduction mechanisms on SOFC cathodes is essential for the development of lower temperature SOFCs with high performance. In order to separate the various contributions to cathode polarization, different techniques and experiments have been carried out to gain an understanding of the kinetics occurring at the gas-solid interface. Isotope exchange is a powerful tool for analyzing the oxygen exchange process (1, 2, 3, 4, 5), however, most isotope tracer experiments are limited in their scope. Experiments typically take one of two approaches, a focus on the kinetics of oxygen adsorption on the surface (SSIKTA), or the diffusion of oxygen through the lattice (SIMS). For SOFC cathodes, we are concerned with both the dissociation of oxygen on the surface and the ability of oxygen atoms to incorporate into and diffuse through the lattice.

In this study, we attempt to determine the ORR mechanism and extract fundamental kinetics rates for La0.6Sr0.4Co0.2Fe0.8O3-x (LSCF) and (La0.8Sr0.2)0.95MnO3-x (LSM) using isothermal isotope exchange (IIE) (6,7). Studies of isotope exchange can greatly contribute to the elucidation of the interactions between molecular oxygen and the cathode surface. Isotope exchange can help characterize the dissociation of oxygen molecules at the oxide surface as well as the conduction of oxygen ions within the oxide. IIE experiments will be conducted at various temperatures and oxygen partial pressures.

To extract fundamental kinetic rates from in-situ isotope exchange experiments on LSM and LSCF, a model has been developed based on a two-step reaction mechanism across the heterogeneous gas-solid interface. The coupled reactions can be shown in two elementary steps: dissociative adsorption and incorporation. The relation between the fundamental kinetics rates and surface exchange coefficient (k_ex) (8, 9) will also be linked.

Acknowledgments

This work was supported by the Department of Energy under contract DEFE0009084. The authors would like to thank Dr. Dongxia Liu for reviewing our model.

References

• M.W. den Otter, B.A. Boukamp, H.J.M. Bouwmeester, Solid State Ionics 139 (2001) 89–94

• Henny J. M. Bouwmeester,* Chunlin Song, Jianjun Zhu, Jianxin Yi, Martin van Sint Annaland and Bernard A. Boukamp, Phys. Chem. Chem. Phys., 2009, 11, 9640–9643

• S. Lacombe, H. Zanthoff, and C. Mirodatos, Jounal of Catalysis, 155, 106-116(1995)

• M.W. den Otter, B.A. Boukamp, H.J.M. Bouwmeester, Solid State Ionics 139 (2001) 89–94

• Schohn L. Shannon and James G. Goodwin, Jr.*, Chem. Rev. 1995, 677-695

• C.C. Kan, E.D. Wachsman, Solid State Ionics 181 (2010) 338–347

• C.C. Kan, H. H. Kan, F. M. Van Assche IV, E. N. Armstrong, and E. D. Wachsman, Journal of The Electrochemical Society, 155 (10) B985-B993 (2008

• E. N. Armstrong, K. L. Duncan, D. J. Oh, J. F. Weaver, and E. D. Wachsman, Journal of The Electrochemical Society, 158 (5) B492-B499 (2011)

• S. Wang, P.A.W. van der Heide, C. Chavez, A.J. Jacobson, S.B. Adler, Solid State Ionics 156 (2003) 201–208

Solid Oxide Fuel Cells (SOFC) cathodes must exhibit high catalytic activity for oxygen reduction, as well as good thermal and chemical compatibility with solid electrolytes in order to minimize resistance and promote stabile operation [1]. The Ln₂NiO_4+δ (Ln = La, Nd, Pr, for example) nickelates, belonging to the Ruddlesden-Popper series, have been proposed as cathode materials. These oxides have the aptitude to accommodate oxygen ions into interstitial sites in the LnO_x layer, which gives rise to a high ionic conductivity by interstitial oxygen diffusion [2]. In addition, the Ln₂NiO_4+δ materials present Thermal Expansion Coefficients (TEC) similar to the most widely used SOFC or IT-SOFC electrolytes: 8mol% Y₂O₃ stabilized ZrO₂ (YSZ), Ce_0.9Gd_0.1O_1.96 (CGO) and La_0.9Sr_0.1Ga_0.8Mg_0.2O_2.85 (LSGM) [3-4].

In this work, we present a study of the reaction mechanism and the kinetic parameters of porous Nd₂NiO_4+δelectrodes. Three different microstructures have been obtained using different preparation methods. The conventional Solid State Reaction (SSR) and Pechini methods, and a soft chemical route based on a sol-gel method. Details of these syntheses are described in [5]. These materials have been deposited by spin coating on dense LSGM electrolytes and were fired at 1000°C. The resulting microstructures were quantified by 3D focused ion beam-scanning electron microscope (FIB-SEM) tomography. Figure 1 shows example 2D images from the 3D FIB-SEM data sets showing markedly different microstructures.

Electrochemical Impedance Spectroscopy (EIS) was used to measure the cells for temperatures ranging from 600 to 700°C and oxygen partial pressures (pO₂) between 1 atm and 610^-4 atm in order to identify possible rate-limiting steps of the oxygen reduction reaction. The EIS spectra, which showed a low and a high frequency response, were fit using an electrical equivalent circuit model composed of two elements – the high frequency contribution was fit using a Gerischer-type element, whereas a parallel resistor/capacitor (RC) element was used for the low frequency arc. The low frequency contribution was only detected at low pO₂ and temperatures, with its size increasing with increasing particle size. The Gerischer element suggests a co-limitation by surface exchange kinetics and oxygen diffusion. The variation of the resistance (R_G) and relaxation time (τ_G) of the Gerischer-type contribution are plotted in Figure 2.

The 3D FIB-SEM data (porosity, specific surface area, tortuosity) was used in the ALS model [6] to find the characteristic kinetic parameters, oxygen surface exchange (k) and oxygen chemical diffusion coefficient (D_chem) that correctly predict the EIS-measured R_G and τ_G values. The results will be compared with reported values for Nd₂NiO_4+δ. Note that the main change in microstructure, the changes in particle size and specific surface area (Fig. 1), agree with the measured R_G values according to the ALS model. That is, high surface area (smaller particle size) led to lower R_G. The microstructure plays an important role in determining the mechanism of the reaction. Thus, the lower pO₂ dependence for R_G and τ_G and the high polarization resistance suggest that the kinetics of the oxygen reduction reaction is mainly limited by a bulk transport path for the SSR sample.

References

[1] S. B. S. Adler, Chem. Rev. 104, (2004) 4791.

[2] F. Mauvy, J. M. Bassat, E. Boehm, P. Dordor, J. C. Grenier, and J. P. Loup, J. Eur. Ceram. Soc. 24, (2004) 1265.

[3] Kharton VV, Viskup AP, Kovalevsky AV, Naumovich E, Marques FMB. Solid State Ionics 143 (2001); 337–53.

[4] Minervini L, Grimes RW, Kilner JA, Sickafus KE. Journal of Materials Chemistry 10, (2000), 2349–54.

[5] Montenegro-Hernández A, Mogni L, Caneiro A. International Journal of Hydrogen Energy 37(23) (2012) 18290-18301.

[6] Y. Lu, C. Kreller, and S. B. Adler, J. Electrochem. Soc. 156 (2009), B513.

Figure 1. Example 2D FIB-SEM images from the 3D data sets taken from epoxy-infiltrated Nd₂NiO_4+δ cathodes. The cathodes were prepared from powders synthesized by a) Solid State Reaction, b) Pechini and c) Sol-Gel methods. The electrolyte is on the right in each image.

Figure 2. Dependence of the Gerischer-type a) resistance (R_G) and b) relaxation time (τ_G) with pO₂.

The limited durability of solid oxide fuel cells (SOFCs) in practical applications has impeded the commercial adoption of these devices. A primary component of this degradation occurs within the cathode upon long-term exposure to various contaminants, including H₂O, CO₂, and Cr vapor. These contaminants cause significant microstructural and compositional changes within the cathode that adversely affect activation, polarization mechanisms, and ionic and electronic conductivities. Previous works (Gostovic et al. and Smith et al.) have demonstrated that a number of quantifiable microstructural characteristics can be directly related to SOFC performance, the most important of these being triple phase boundary length (L_TPB) and pore surface area. These parameters have not been examined during cell degradation, and further analysis under these conditions will provide insight into specific cell degradation mechanisms, informing future fabrication and operation criteria.

In this work, we present a three-dimensional quantification of porous SOFC cathode materials that have been aged in both clean and contaminated atmospheres. Symmetric cathode cells were produced by standard screen printing methods using an yttria-stabilized zirconia (Y₂O₃-ZrO₂, YSZ) electrolyte, and a composite cathode. The composite cathode consisted of 50% lanthanum strontium manganite (La_1−xSr_xMnO₃, LSM) and 50% YSZ by weight. We report here the quantification of and differences between three YSZ/LSM-YSZ symmetric cells: unaged, aged for 480 hours in a clean air environment, and aged for 480 hours under H₂O atmosphere. This quantification was performed using a dual-beam focused ion beam/scanning electron microscope (FIB/SEM) to serially mill 30nm slices of the cell, taking an image after each slice. In this way, a series of 2D images was acquired, aligned, and reconstructed into a fully quantifiable 3D volume rendering of the cell near the cathode/electrolyte interface, as shown in Figure 1.

After developing a 3D model for each cell, the volumes were quantified by a number of parameters, including L_TPB, particle/pore size and distribution, surface area coverage, porosity, phase volume fraction, and tortuosity. Additional information relating to phase connectivity was calculated by generating a "skeleton" network for each phase, and was compared for each cell. The microstructure was then related to cell performance data acquired from the same samples. Furthermore, the detailed microstructure, compositional changes, and the formation of impurity phases due to the contaminants was characterized by transmission electron microscopy (TEM), utilizing both energy dispersive X-ray spectroscopy (EDS) and electron energy loss spectroscopy (EELS).

1. Gostovic, D., Vito, N. J., O'Hara, K. A., Jones, K. S., & Wachsman, E. D. J. Am. Cer. Soc., 94(2), 620–627, (2011).

2. Smith, J. R., Chen, A., Gostovic, D., Hickey, D., Kundinger, D. P., Duncan, K. L., & Wachsman, E. D. (2009). Solid State Ionics, 180(1), 90–98.

Figure 1: Serial 2D slices of YSZ electrolyte (left) and composite LSM-YSZ cathode (right) acquired in FIB/SEM (a) are reconstructed into a representative 3D volume surface model (b) for further quantification. Data is from unaged symmetric cathode button cell.

In recent years, direct carbon fuel cells (DCFCs) have received significant attention owing to an increasing need for efficient power generation from coals and other carbon-based solid fuels. In a DCFC with a solid oxide electrolyte (e.g., yttria-stabilized zirconia, YSZ), carbon conversion would take place over three-phase boundaries. There is, however, very little interaction between the solid carbon and the Ni-YSZ anode. A hybrid DCFC design combines the advantages of the solid oxide and molten carbonate fuel cell technologies, where the molten carbonate and solid oxide electrolytes are in direct physical contact. Oxide ions are transported through the solid oxide electrolyte toward the molten carbonate holding the dispersed carbon particles inside the anode compartment. Although promising performances of hybrid DCFCs have been reported in the literature, the mechanistic understanding of the carbon conversion is incomplete. In fact, the reactions at the anode are quite complex and involve multistep transfer of electrons in a sequence of elementary reactions. In this paper, we study the electrochemistry of carbon conversion in a hybrid DCFC using Ni-YSZ anodes with various porous structures. Ni-YSZ anodes are fabricated using PMMA spheres with different diameters as a pore-former to tailor their porous structures (pore volume, size, distribution, and surface area). The polarization behaviors of the DCFC are measured as a function of the porous structure. The results show that among the parameters studied, the surface area plays the most critical role in determining the anodic polarization. The electrochemical carbon conversion reaction is discussed in detail, on the basis of the experimental data obtained from mercury porosimetry, polarization and impedance measurements, and gas chromatography.

Direct methanol fuel cells (DMFCs) are attractive power sources for portable electronic devices owing to their advantages of low operating temperature, easy transportation and fuel storage, high-energy efficiency, low exhaustion and fast start-up. The alkalines DMFCs have drawn particular attention in this area for the advantage of high alcohol oxidation kinetics and of the potential use of cheaper non-Pt catalysts. Nickel hydroxide has been extensively studies as electro-catalysts for methanol oxidation. In this present work alpha nickel hydroxide was synthesized and incorporated into carbon ceramic electrodes. The nickel hydroxide was obtained by mixing nickel acetate and glycerin with heating at 50 ° C under constant agitation. The obtained solution was mixed with a solution of NaOH dissolved in butanol, the mixture obtained was stirred for 6 hours and then allowed to stand for 15 days. Carbon ceramic (CC) was obtained by the Sol-Gel process, methyltrimethoxysilane (MTMS) was mixed with ethanol and the catalyst used was NaOH after partial hydrolysis of MTMS graphite was added to the mixture, the material obtained was allowed to stand for 48 h. The ceramic-carbon (CC) was mixed with nickel hydroxide in the proportion of 5, 10, 30, 50 and 70% (w/w) Ni(OH)₂/CC. The mixture was homogenized and packaged in glass tubes of 0.03 cm² and 1 cm in height, and nickel-chromium wire was inserted as electrical contact. The material obtained was designated as ECC/α-Ni(OH)₂ and was used as an electrode for electro-oxidation of methanol in alkaline medium. The x-ray diffraction confirmed the presence of lamellar graphite. The images of SEM/EDS confirmed the presence of nickel hydroxide dispersed irregularly in the sample. Initial studies indicated that the electrodes with the higher concentrations of nickel hydroxide produce higher current intensities at higher anode potential. In studies of the electro - oxidation of methanol was chosen electrode produced with 10 % of Ni(OH)₂, because it has the lowest potential anodic peak. The ECC has no catalytic activity for the electro-oxidation of methanol, but the ECC/α-Ni(OH)₂ promotes the electro-oxidation of methanol potential of 0.53 V vs. Ag/AgCl, yielding a current density of about 100 mA/cm². The highest oxidation current was obtained with 1.90 mol/L of methanol. The ECC has no signs of poisoning, because there is no decrease in the intensity of the anodic current after 1000 cycles. The electrode showed promising thermal behavior, with no loss of electrochemical activity until the temperature of 200^oC. All results indicate that the ECC/α-Ni(OH)₂ is a promising material for electro-oxidation of methanol in alkaline medium, to obtain high current densities.

Crystalline titanium oxide (TiO₂), which exists in the three ubiquitous phases: rutile, anatase and brookite, possesses a direct wide band gap (E_g ~ 3.3 eV at 300 K) similar to that of GaN (E_g ~ 3.4 eV at 300 K), is under consideration as an alternative material to replace GaN for optoelectronic applications. It's known that reduced titanium dioxides (TiO_2-x and TiO) exhibit some unique chemical activities towards water oxidation. In this presentation, we will report our recent studies of using TiO_2-x and TiO as the electrocatalysts to convert carbon dioxide to fuels by incorporating the electrocatalysts into a polymer exchange membrane fuel cell (PEMFC) based full electrochemical cell. Nonstoichiometric Ti₃O₅ was prepared by reducing TiO₂ at high temperature in the presence of high pressure hydrogen. X-ray diffraction and thermogravimetric analysis were used to analyze the phase and nonstoichiometry of TiO and Ti₃O₅. Ti₃O₅ exhibits the highest Faradaic efficiency (~ 12%) towards the formation of formate, comparing to < 2% for both TiO and TiO₂ P25. X-ray photoelectron spectroscopy (XPS) measurements of the Ti₃O₅ gas diffusion electrodes (GDEs) before and after electrolysis also suggest the presence of surface defects such as Ti³⁺ and oxygen vacancies which significantly influence the electrocatalytic activity for CO₂ reduction.

Zirconia-based solid electrolytes are the preferred materials for high-temperature applications due to the combination of their electrical and mechanical properties. Zirconia-8-11 mol% scandia possess higher values of ionic conductivity at temperatures above 600 °C than other zirconia-based electrolytes [1]. At low temperatures the ionic conductivity of this system decreases due to the transition from the cubic fluorite-type phase to the ordered Sc₂Zr₇O₁₇(β) rhombohedral phase [2].

It has been reported that the cubic phase can be stabilized at room temperature by adding minor amounts of e.g. Y₂O₃, CeO₂ and Gd₂O₃[3].

In this work, small amounts of Dy₂O₃ was added to zirconia-10 mol% scandia to evaluate the effect of the additive on the phase composition and the ionic conductivity.

Zirconia-10 mol% scandia (DKKK) and Dy₂O₃(Alfa Aesar) were the starting materials. Dysprosia was added in the 1-2 mol% contents to the zirconia-scandia solid electrolyte. Cylindrical pellets were prepared by pressing followed by sintering at 1500 °C. Sintered pellets were characterized by X-ray diffraction, apparent density measurements, scanning electron microscopy observations and ionic conductivity evaluation by impedance spectroscopy.

Table 1 summarizes the relative densities, ρ_R, of sintered samples with different additive contents, x.

Table 1. Dy₂O₃ content (x, in mol%) and relative density (ρ_R) of sintered samples.

The sintered density decreases with increasing Dy₂O₃ content. The relative density is higher than 92% for sintering experiments at 1500 °C for 5 h for all compositions.

The X-ray diffraction patterns of sintered pellets show that for the additive content of 1 mol% the cubic phase was partially stabilized at room temperature. A small amount of the rhombohedral phase coexists for this additive content. For dysprosia additions of at least 2 mol% the rhombohedral phase was suppressed, as shown in Fig. 1. The X-ray diffraction pattern of Fig. 1 corresponds to that of cubic zirconia-based solid electrolytes.

The ionic conductivity was determined by impedance spectroscopy in the 500-800 °C range. The Arrhenius plot of the total ionic conductivity is depicted in Fig. 2 for the sample containing 2 mol% dysprosia.

The calculated apparent activation energy is 1.2 eV, similar to that of other zirconia-scandia systems.

Stabilization of the cubic phase was successfully accomplished by small additions of dysprosia to zirconia-10 mol% scandia. The magnitude of the ionic conductivity value decreases with increasing additive content. The activation energy for oxide ion conduction is similar to that of other zirconia-scandia systems containing a second additive.

References

[1] E. C. Subbarao, Solid electrolytes and their applications, Plenum Press, New York, 1980.

[2] S. P. S. Badwal, F. T. Ciacchi, D. Milosevic, Solid State Ionics 136/137(2000) 91.

[3] O. Yamamoto, Y. Arati, Y. Takeda, N. Imanishi, Y. Mizutani, M. Kawai, Y. Nakamura, Solid State Ionics 79(1995) 137-142.

In recent years, the environmental friendly solid oxide fuel cell (SOFC) has been considered as one of the most efficient power generation systems. During the SOFC operation, the over potential is mainly given by the cathodic oxygen reduction reaction. In this respect, the mechanism of oxygen reduction on the Electrode is important to reveal. Specially, employing ac-impedance spectroscopy which provides an exceptionally is powerful tool for separating the dynamics of several electrode processes with different relaxation times. The present work considers the oxygen reduction reaction on the porous mixed conducting La_0.1Sr_0.9Co_0.8Fe_0.2O_3-δ (LSCF1982) electrode by using ac-impedance spectroscopy. Firstly, to attain a mixed conducting state of LSCF1982, the ac-impedance spectra measured at various temperatures and oxygen partial pressures were quantitatively analyses by employing the theoretical impedance model developed for oxygen reduction on the mixed conducting electrode.

All the measured impedance spectra simply consisted of two depressed arcs which overlapped slightly over the whole T and pO₂ ranges. The analysis of the ac-impedance spectra by employing discrete Fourier relaxation transformation (DFRT) allowed us to successfully differentiate the individual reaction steps. This indicates that the overall cathode reaction proceeds by the following two processes: charge transfer reaction and subsequent migration of oxygen vacancy.

It is also shown, that the magnitude of the high frequency arc decreased significantly with increasing T, but it remained nearly constant regardless of pO₂: In contrast, the low-frequency arc remained almost unchanged in magnitude irrespective of the value of T, while it was highly dependent on the value of pO₂. From the experimental results that it is thus deduced that the high frequency arc is attributed to the charge transfer reaction at the TPBs, and the low-frequency arc is associated with diffusion of adsorbed oxygen atom along the electrode/gas interface.

Figure 1 shows the steady-state current of the oxide anodes as a function of the electrode potential and a_O,int.

According to Wang and Nowick [1], the rate determining step determined to exchange current density value as a function of both temperature and oxygen partial pressure.

i₀ ∝pO₂ⁿ

The n - value in a relation gives information about the type of species involved in the electrode reaction. The n values have 1 if the rate limiting step is diffusional process of molecular oxygen, 1/2 for dissociative adsorption and 1/4 for charge transfer process depending on the conditions for Langmuir adsorption isotherm.

In this study for LSCF1982, the slopes of current exchange density were turned out to be 1/4 at 550 to 650^oC. This indicated that the rate limiting step for the electrochemical oxygen reduction in this temperature range is the charge transfer process.

In the present work, the oxygen reduction reaction on the porous LSCF1982 electrode was examined by means of analysis of the ac-impedance spectra and DC polarization. In addition, we'll discuss the other technic like cathodic potentiostatic current transients and Gerisher impedance modeling.

Reference

[1] Da. Yu. Wang and A. S. Nowick, J. Electochem. Soc., 126, 1155 (1979)

Acknowledgement

This work was supported by Solid Oxide Fuel Cell of New & Renewable Energy R&D Program (20123010030010) under the Korea Ministry of Knowledge Economy (MIKE)

Cerium pyrophosphate based solid state proton conducting materials with ionic conductivity >10^-2 S cm^-1 in 100-250 ^oC range have immense importance in fuel cells.^1-5 In this work, Ce_1-xMn_xP₂O₇ (CMP) (x=0.05, 0.075, 0.1, 0.125 and 0.15) composite electrolytes are synthesized by two-step slow digestion method with different P/(Ce+Mn) molar ratio.⁴ The CMP samples with high phosphate content become denser on sintering. The variation of ionic conductivity with temperature is studied in dry and humidified air for the potential application of CMPs as electrolytes in proton-conducting ceramic electrolyte fuel cells. Among various CMP samples, CMP-100-2.7P (i.e. Ce_0.9Mn_0.1P₂O₇ with P/(Ce+Mn)=2.7) shows maximum conductivity of 6.54X10^-6 S cm^-1 at 450 ^oC in dry air and 1.78X10^-2 S cm^-1 at 170 ^oC in humidified air (pH₂O= 0.12 atm). The ionic conductivity of CMPs increases with the increasing pH₂O and CMP-100-2.7P shows maximum conductivity of 2.24X10^-2 S cm^-1 at 170 ^oC in pH₂O=0.16 atm.

Figure. (a) SEM images of fractured section of 400 ^oC sintered pellets of Ce_0.9Mn_0.1P₂O₇ samples with different initial P/(Ce+Mn) ratio; Temperature dependence of ionic conductivity of various CMP samples with different dopant concentration in (b) dry air and (c) humidified air (pH₂O =0.12 atm) condition.

References

• B. Singh, H. N. Im, J. Y. Park, and S. J. Song, J. Electrochem. Soc.,159, F819 (2012).

• B. Singh, H. N. Im, J. Y. Park, and S. J. Song, J. Phys. Chem. C, 117,2653 (2013).

• B. Singh, H. N. Im, J. Y. Park, and S. J. Song, J.Alloys Compd.,578, 279 (2013).

• B. Singh, S. Y. Jeon, J. H. Kim, J. Y. Park, and S. J. Song, (Submitted in J. Electrochem. Soc.).

• C. Chatzichristodoulou, J. Hallinder, A. Lapina, P. Holtappels, and M. Mogensen, J. Electrochem. Soc., 160, F798 (2013).

Gadolinia-doped ceria has been intensively investigated over the last decades as a promising electrolyte and electrode material for intermediate-temperature solid oxide fuel cells [1]. The sintering behavior of this solid electrolyte in reducing atmosphere is different from that in air, because of the reduction of Ce⁴⁺ to Ce³⁺[2]. In this case, changes in the microstructure and in the related properties are expected to occur, due to variation in the oxygen-vacancy concentration.

In this work, the effects of the sintering atmosphere and the particle size of powders on densification and ionic conductivity of Ce_0.9Gd_0.1O_2-δ were systematically investigated.

Ce_0.9Gd_0.1O_2-δ commercial powders (>99.5%, Fuel Cell Materials) with specific surface area ranging from 7.4 to 210 m².g^-1were used as starting materials. Cylindrical pellets were prepared by uniaxial and isopressing following by sintering at 1250°C for 2 h under different atmospheres.

The average initial particle size of the powders was evaluated by nitrogen adsorption (Quantachrome, NOVA 1200) using the BET method. Characterization of the sintered materials was carried out by density measurements, thermodilatometry (Anter-1161, Unitherm), electronic Raman spectroscopy (InVia Raman microscope, Renishaw), electron paramagnetic resonance (Bruker-EMS), field-emission scanning electron microscopy (FEI, Inspect F50) and impedance spectroscopy (HP4192A). The mean grain size was estimated by the linear intercept method.

The sintered pellets reached relatively high densities (93 to 97% of the theoretical value) independent on the sintering atmosphere. The particle size determined by nitrogen adsorption of the starting powders was: 112, 23 and 4 nm for powders, hereafter named S1, S2 and S3, respectively. The linear shrinkage curves show that good densification of gadolinia-doped ceria is attained depending on the initial size of the powder particles. However, for isothermal treatments the densification is homogenized and all sintered pellets reached high values after 1250°C for 2 h.

All investigated properties changed to some extent with the sintering atmosphere and the initial particle size of the starting powders. The grain conductivity, for example, of Ce_0.9Gd_0.1O_2-δ pellets sintered in air and under inert atmospheres (argon and nitrogen) exhibit the same bulk conductivity not depending on the sintering atmosphere, in spite of the changes observed in the electron paramagnetic signal.

References

[1] H. Yahiro, Y. Eguchi, K. Eguchi, H. Arai, J. Appl. Electrochem. 18 (1988) 527.

[2] Z. He, H. Yuan, J. A. Glasscock, C. Chatzichristodoulou, J. W. Phair, A. Kaiser, S. Ramousse, Acta Mater. 58 (2010) 1866.

Hybrid direct carbon fuel cells (HDCFCs) consisting of a solid carbon (carbon black)-molten carbonate ((62-38 wt% Li-K)₂CO₃) mixtures in the anode chamber of an anode-supported solid oxide fuel cell (SOFC)-type full-cell (NiO-yttria-stablized zirconia (YSZ)|YSZ|lanthanum strontium manganite (LSM)-YSZ/LSM) are tested for their electrochemical performance between 700 and 800°C. Performance was investigated using electrochemical impedance spectroscopy (EIS) and current-potential-power density curves. EIS data is interpreted using a model circuit (R-RQ-RQ-RQ), and the dominant processes revealed by the impedance data as a function of temperature, anode and cathode atmospheres, and their flow rates are discussed. In the anode chamber, catalysts are mixed with the carbon-carbonate mixture. These catalysts include various manganese oxides (MnO₂, Mn₂O₃, and Mn₃O₄, Fig. 1) and doped-ceria (CeO₂, Ce_1-xGd_xO₂, Ce_1-xRE_xO₂ (RE = Pr, Gd, Sm, etc.)), the effectiveness of these families of catalysts are discussed with respect to electrochemical, chemical and post-mortem analysis.

Fig. 1. Current-potential-power density curves acquired for a blank (SiC) and manganese oxide (MnO₂, Mn₂O₃, Mn₃O₄) catalysts suspended in the carbon-carbonate mixture in the anode chamber of an HDCFC. 96-4 vol% N₂-CO₂ (anode), air (cathode), 755°C, 0-600 mA, 50 mA/step. Power density corrected to cathode geometric surface area.

Introduction

As developing countries consume increasing quantities of fossil fuels to elevate their standard of living, there is growing concern about greenhouse gas emissions, and an increased need for mitigating emissions during power generation. It is clear that in the foreseeable future renewable energy technologies will be unable to eliminate fossil fuel dependence, and hence efforts to reduce the environmental impact of carbon-based fuels are required.

Carbon Fuel Cells (CFCs) offer the double benefit of efficient utilization of carbonaceous fuels such as coal or biomass, and the production of a concentrated stream of CO₂ that can be easily stored or sold as a marketable product.

Previous work in our laboratory has demonstrated solid-oxide based CFC for efficient electricity production from various types of carbons [1-3]. Our CFC utilizes a bed of solid carbon fuel at the anode compartment and air at the cathode compartment, which are separated by a dense yttria-stabilized zirconia electrolyte (YSZ) for selective transport of oxide ions. Carbon dioxide in the anode compartment reacts with the carbon to produce CO via the Boudouard reaction.

C + CO₂ → 2CO (1)

As shown in Figure 1, the cell oxidizes CO to generate electricity, producing an outlet stream of CO₂, part of which can be recycled to the anode and the reminder sent for storage.

Modeling the Tubular Carbon Fuel Cell Geometry

In this presentation we provide a comprehensive operational model that couples heat transfer with chemical and electrochemical processes as well as mass transport in a tubular carbon fuel cell. In the carbon bed, the Boudouard gasification reaction of the solid carbon fuel is endothermic, while CO oxidation at the anode surface is exothermic. As the kinetics of both reactions is temperature dependent, it is important to understand the coupled relationship between reaction rates, heat release and local temperature. Thus, consideration of heat transfer effects is necessary to develop a realistic understanding of the overall cell operation.

The operation of the cell results in an anode exhaust containing largely CO₂ with the remaining balance of unreacted CO. Any CO in the exhaust is carbon fuel still capable of undergoing oxidation to produce electricity at the cell. Thus, fuel utilization improves with increasing CO₂/CO ratio at the exhaust, and this influences the overall conversion efficiency of the cell.

In this work, an operational model for a tubular CFC was developed that takes into account heat transfer and temperature distributions within the cell. The parameters in the model were determined experimentally. The model was then used to map out the operational space for power density and cell efficiency. Furthermore, geometrical parameters such as fuel bed height and tubular placement can be tuned to minimize the mole fraction of CO in the exhaust. The model was implemented for multiple tubular geometries and spacing between tubes in order to determine how these parameters affect overall fuel cell performance.

This presentation will address the dependence of cell efficiency and power density on the cell and carbon bed geometries, which will then aid in the optimization of tubular cell design for the air-carbon fuel cell. Optimal operation conditions are identified for maintaining high efficiencies while also achieving realistic power densities.

References

1. A. C. Lee, S. Li, R. E. Mitchell, and T. M. Gür. Electrochem. Solid State Lett. 11(2), B20-B23 (2008).

2. B. R. Alexander, R. E. Mitchell, and T. M. Gür. Proceeding of theCombustion Institute 34, 3445 (2013).

3. B. R. Alexander, R. E. Mitchell, and T. M. Gür. J. Electrochem. Soc.159(3), B347-B354 (2012).

Introduction

The molten carbonate fuel cell (MCFC) has emerged as one of the leading devices to convert chemical energy into power. To date 1.9 GWh of electricity has been produced commercially using this new technology. The molten carbonate fuel cell uses a mixture of alkali carbonates as the electrolyte and operates in the temperature range of 550-650°C. The high operating temperature dramatically improves the reaction kinetics and eliminates the need for a noble metal catalyst. During cell operation, hydrogen is oxidized at the anode and oxygen is reduced at the cathode, equations (1) and (2)

H₂ + CO₃^2- → H₂O + CO₂ + 2e^- (1)

1/2O₂ + CO₂ + 2 e^- → CO₃^2- (2)

The current carbonate fuel cell design (illustrated in figure 1) has evolved over many years from extensive research conducted in fifteen companies and many prestigious laboratories. The cell hardware is made from stainless steel, the electrodes are nickel based, and the separator called matrix is a porous ceramic. The anode design is an important consideration from both performance and cost considerations.

The state of the art MCFC anode is a porous Ni electrode stabilized against sintering by Cr and/or Al additives. Typical anodes have porosities of 40-60% and 3-6 µm as average pore diameters. To ensure long-term stability against creep, corrosion and oxidation, major considerations need to be taken into account in material selection. Mechanical and chemical stability as well as the electrochemical activity for anode are key factors for performance and life stability.

Accelerated lab-scale technology stacks (30kW) and long-term field tests have shown that the current anode materials have the desired performance and stability of the useful life of >5-7 years. As shown in figure 2, anode maintains stable pore structure and morphology over long-term operation. Parameters such as conductivity, pore size distribution, wetting and electrolyte fill level affect anode performance and its stability. To achieve an

The concept of Hybrid Direct Carbon Fuel Cell (HDCFC) uses a mixture of solid carbon with (Li-K-Na) carbonates, which melt at operating temperature, fed on top of a solid oxide anode. This setup offers several advantages: the presence of carbonates helps both to increase the contact between the solid carbon and the solid oxide electrode and to extend the electrolyte into the solid carbon. Furthermore the HDCFC allows the use of the current SOFC configuration, like commercial anode supported cells.

In this study the performance of anode supported NiO/YSZ│YSZ│LSM/YSZ cells towards direct carbon oxidation has been measured. Particularly, the effect of both CeO₂ infiltration into the NiO/YSZ anode and CeO₂ infiltration of solid carbon have been investigated between 750°C and 815°C through the use of polarization curves and electrochemical impedance spectroscopy (EIS). The impedance spectra are modeled and discussed with the objective to study the effect of CeO₂ infiltration on the processes dominating the impedance response of the anode.

Figures 1 and 2 show the Nyquist plots of the impedance spectra for the CeO₂ infiltrated anode cell and the blank cell, respectively. The spectra were recorded at open circuit potential at 815°C, 800°C and 760°C in the presence of 96% N₂ and 4% CO₂.

A zirconia oxygen sensor was designed to monitor the gas-solid oxidation reaction characteristics of metal/metal-oxide powder beds for energy storage applications. Energy is stored in the metallic state and released during oxidation. This in-situ monitoring device allows for determination of the oxidation state of the metal oxides during reactions in real time, as well as calculation of the chemical reaction rate constants. The signal output of the sensor was analyzed for oxidation of and diffusion into the powder bed. The oxygen transport mechanisms occurring inside the sensor were described to further understand the signal outputs from the sensor. Two metal/metal-oxide systems were examined using this device to demonstrate its performance. The simple Ni/NiO system was chosen to demonstrate feasibility, and the complex W/WO₃ system was chosen to demonstrate versatility of the sensor for actual energy storage applications.

The Direct Carbon Fuel Cell (DCFC), which uses solid carbon as fuel and molten carbonate as electrolyte, has had resurgence of interest due to very high electrochemical conversion efficiency, nearly 100%, no requirement of fuel reforming, and potential for CO₂capture and sequestration. At the cathode, carbon dioxide is converted into carbonate ions. The main reaction at the anode is generation of carbon dioxide from carbon and carbonate, the net cell reaction being oxidation of carbon to carbon dioxide. Additional reactions that may be occur at the anode are a two-electron reaction resulting in co-generation of carbon dioxide and carbon monoxide and the Boudouard reaction leading to conversion of carbon and carbon dioxide to carbon monoxide, the so-called "carbon corrosion". The performance of DCFC can be appreciably limited by this reaction. The reverse Boudouard reaction has not been studied in solid-melt-gas systems. It is important to study this reaction in conjunction with wetting behavior of carbon in molten carbonate.

With this in mind, the wetting behavior of carbon electrode in molten carbonate, comprised of a Li-K eutectic mixture, was studied in this study at different temperatures. The wetting behavior is complicated due to reactions occurring at the carbon surface before or after the start of wetting. Before wetting, carbon rods were exposed to a gas atmosphere with varying partial pressures of carbon dioxide to examine the extent of Boudouard reaction and the resulting structural changes in carbon rods. This reaction modifies the lateral surface of carbon electrode due to carbon corrosion, which leads to reduction in interfacial tension between solid carbon and molten carbonate after dipping carbon rods in molten carbonate.

Wetting behavior of carbon rods was studied for a period of 24 hours. The effects of temperature and pre-exposure to varying levels of carbon dioxide on speed of formation of liquid (molten carbonate) meniscus, propagation of liquid film thereafter, and the ultimate length of liquid film were studied. During the wetting process, bubble evolution occurred at the surface of carbon electrode not only below the meniscus, but also above the meniscus. This indicates that wetting of carbon electrode in molten carbonate is influenced not only by capillary forces but also by the reverse Boudouard reaction. The reverse Boudouard reaction is promoted at higher temperature, leading to changes to greater extent in the surface structure and therefore solid-liquid and solid-vapor interfacial tensions.

The lateral surfaces of graphite rods before wetting and after wetting, liquid films formed on the graphite rods, and cross-sections of cylindrical carbon rods above and below the meniscus were analyzed using SEM. The liquid films and circular cross-sections of carbon rods were also analyzed using energy-dispersive X-ray spectroscopy (EDX) and distributions and variations in oxygen and potassium were examined. Oxide ions are generated prior to and during wetting of carbon rods and play an important role in surface modification of these prior to and during wetting.

A reaction mechanism based on the experimental observations is proposed. The roles of Boudouard reaction, including equilibrium characteristics, and oxide ions on surface modification of carbon rods before and after wetting and the implications thereof on wetting behavior of carbon rods are explained satisfactorily by the reaction mechanism.

The alloys of Fe-Cr-Al-RE system are the high temperature oxidation resistant materials and they are normally are intended to apply at the temperatures up to 1200^OC. Usually the composition of the mentioned system is varied by the main elements values as well as RE's and other alloying elements amount and the variety of the latter. The high temperature behavior of the Fe-Cr-Al-RE bulk alloys and the coatings with the chromium content of up to 25wt% and the Al that of up to 5wt% is investigated widely, whereas higher chromium content alloys of the mentioned system are barely studied.

The oxidation peculiarities of high (>40wt%) chromium content Fe-Cr-Al-RE model alloy with some minor additives were compared to that of the two commercial (Aluchrom YHf and PM2000) alloys of the same system. According to the experimental results the new material investigated exhibited the slowest oxidation rate (as it is shown in the Fig.1a) and the thinnest oxide scale formation comparatively to the other two alloys which makes it a quite promising high temperature oxidation resistant material [1-3].

The processes taking place on the surface of model alloy exposed to high temperatures are discussed and compared to those of the two commercial alloys. The attempt to disclose the ways differentiating the oxide scale formation on the investigated low and high chromium containing alloys of Fe-Cr-Al-RE system was made and the factors influencing the slower oxide development process for the model alloy were suggested.

The obtained kinetic data after the high temperature cyclic oxidation (120/15 minutes of heating up/cooling down cycling) in laboratory air at 1100^OC during 2000 hours is compared for all the investigated tree alloys having either polished or rough surface conditions before the exposures at the elevated temperatures. Thus, the initial surface condition on the protective oxide scale growth kinetics was studied. Besides, the oxide scale growth rate dependence on the specimen thickness was studied as well for the alloys exposed at the same elevated temperatures during 1000 hours, where new model alloy has showed the outstanding priority comparing to the other two commercial alloys investigated (Fig. 1b). SEM analysis was applied for the evaluation of the oxidized specimens' surfaces as well as their cross-sections. References

• Tsurtsumia O., Kutelia E., Bulia B., Mikadze O., Materials Science Forum Vols. 595-598, 2008, (Trans Tech Publications) pp. 833-840

• Tsurtsumia O., Kutelia E., Bakhtiyarov S., Proceed. Of 17th Int. Corrosion Congr. USA, Las Vegas, Nevada, October, 2008

• Kutelia E., Tsurtsumia O., Eristavi B., Adanir H., Bakhtiyarov S., Journal on Engineering and Technology, 2006, Vol. 1, No. 4, pp. 61-71

Hydrogen is widely used by petroleum refineries in upgrading processes, necessary to facilitate the hydrotreating/catalytic hydrocracking of heavy hydrocarbon molecules, and the reduction of sulfur content. At an industrial scale, three major technologies are presently used to produce hydrogen, depending on the primary energy sources or available feedstock: coal gasification, natural gas steam reforming, and water electrolysis.

Petroleum coke (or petcoke), similar to coal, is a byproduct of refinery processes. The U.S. refineries produce more than 125,000 short tons petcoke per day (st/d), most of which is produced in Calfornia, Texas and Louisiana. With a low heating value (LHV) of 6.024 MMBtu/barrel (equivalent to 8.826 MWhr/st), petcoke produced in the U.S. potentially is worth 17,689 MW electricity assuming 40% efficiency. However, due to larger activation energies, higher sulfur content and less volatile materials than coal, petcoke is not a desirable feedstock for a conventional coal-burnt power plant and more than 62% of petcoke was exported annually.

Instead of purchasing hydrogen over the fence, a refinery potentially can build on-site hydrogen production infrastructure directly utilizing its available resource (petcoke). Materials & Systems Research Inc. currently is developing an advanced hydrogen production process built-upon the solid-oxide electrochemical technology with reduced energy consumption and low costs for CO₂ separation & sequestration. In this talk, the hydrogen production process and the development of the electrochemical device will be discussed in detail. Proof-of-concept demonstration of hydrogen production at a 200W stack level will also be presented.

Mass transport property in nano-size structure controlled materials are attracting much interest because such effects have a possibility to improve the performance of the electrochemical solid state devices such as Solid Oxide Fuel Cells (SOFCs). There are many approaches and ideas for the improvement of oxide ion conductivity in electrolyte by nano size effects. In this study, we demonstrated the 3 dimension (3D) tensile strain in Pr₂NiO₄, mixed conductor, was successfully introduced by dispersing Au particles in grain. The maximum lattice distortion estimated from X-ray diffraction measurement was 0.23 % at 2 mol% Au dispersion. Au dispersed Pr₂NiO₄ shows a large positive effect on oxide ion diffusivity and surface exchange coefficient.

The doubled oxygen permeation flux was achieved with dispersing 2 mol% Au in PNCG at 873 K, indicating the tensile strain in PNCG lattice positively works for the oxygen permeation property. Since oxygen permeation is controlled by interstitial oxygen concentration and its mobility, further details study on effects of tensile strain on oxygen diffusivity were studied by using ¹⁸O tracer diffusion. The depth profiles of x mol% Au/PNCG (x=0, 1, 2, 3) are studied with SIMS analysis, and the estimated oxygen tracer diffusivity (D*) and the surface exchange coefficient (k*) are much increased with dispersing Au, and the enhancement in D value was well agreed with that in oxygen permeation rate. Redox titration measurement suggests that 3D tensile strain increased amount of excess oxygen. Interstitial oxygen in Pr₂NiO₄ is introduced in Rock salt layer of K₂NiF₄ structure, and this is origin for the fast oxygen diffusivity and high surface activity for oxygen dissociation.

Recently, there is much attention of surface composition in oxide ionic conductors using the low-energy ion scattering (LEIS) technique to understand the surface reaction and degradation. LEIS analysis was also performed on Au dispersed Pr₂NiO₄ and it was found that surface of Pr₂NiO₄ was enriched with Pr. However, Pr segregation was prevented by introduction of Au and so tensile strain also positively works for preventing Pr segregation. On the other hand, surface oxygen content is also decreased by Au dispersion and so surface oxygen vacancy seems to be much enriched. Therefore, increased surface exchange coefficient by Au dispersion could be assigned to the high concentration of oxygen vacancy on surface of Pr₂NiO₄ dispersed with Au.

In summary, although Au dispersion was not uniform, effect of tensile strain on oxide ion mobility could be observed in Pr₂NiO₄ by Au dispersion. Furthermore, tensile strain affected the surface composition, which may enhance the k* values. If the better dispersion state with fine particles could be achieved, there is a possibility to develop the higher oxide ionic conductor with uniform strained material.

Solid oxide fuel cells (SOFCs) are ideally suited for environmentally benign conversion of chemical energy in hydrocarbons to electricity. While medium scale SOFC power systems in the hundreds of kWe have been demonstrated, larger scale application of such systems have been hobbled by still too high cost. In order to bring down the cost of these systems, manufacturing costs have to be decreased while the power densities have to be improved. Achieving these twin goals will require the reduction of operating temperatures of the cells. On a single cell level, the largest source of performance loss at lower operating temperatures occurs in the cathode. Thus many different groups are presently investigating a slew of new cathode materials. While the performance of the new cathode materials are impressive under controlled operating conditions, it is presently not clear whether they will stand up to the rigorous and punishing operating environments of actual fuel cell stacks. In particular the interfacial and bulk stability of the new generation of cathode materials such as strontium doped lanthanum cobaltites, strontium doped lanthanum cobaltites , and lanthanum nickelates when exposed to anodic gases such as carbon dioxide and water vapor is presently unknown. Exposure to at least trace amounts of anodic gases during the lifetime of the stack is a likely occurrence and could arise from back diffusion, pinholes in the electrolyte and/or interconnection, or compromised seals. Thus, it is important to have a clear understanding of the thermodynamic and kinetic stability of both the interfaces and the bulk of the cathode material under exposure to the aforementioned anodic gases. In this paper, we present results from a study that combines electrochemical and x-ray measurements of the impact of carbon dioxide exposure on strontium doped lanthanum cobalt iron oxide (LSCF), as a function of strontium dopant concentration, gas composition, and temperature. These are combined with computational studies of the stability of the LSCF cathode to carbon dioxide exposure. The longer term goal of this study is to identify cathode compositions which have long term stability to exposure to trace anode side gases.

Degradation of SOFC cathodes is the result of microstructural and/or chemical alterations taking place during operation [1,2]. The latter originate either from impurities or secondary phases that segregate or form at the electrochemical active solid/gas interface or from compositional changes of the cathode material itself. On the one hand, the cathode material can react with species from adjacent or nearby solid materials within the multi-layer cell setup. On the other hand, surrounding gas species or impurities can initiate the formation of secondary phases on the surface of the cathode material or the decomposition of the cathode material itself.

This contribution will focus on the solid/gas interaction between nanoscaled La_0.6Sr_0.4CoO_3-δ (LSC) thin films and the surrounding gas atmosphere in terms of gas composition and partial pressures and discusses the resulting influence on the electrochemical properties. The thin-film cathodes were derived by a sol-gel process based on metal organic precursors and were directly deposited onto Ce_0.9Gd_0.1O_1.95electrolyte pellets [3].

Differently designed experiments and thermodynamic calculations were performed in order to investigate the influence of water-vapor or carbon dioxide in the surrounding atmosphere and of the oxygen partial pressure on the cathode performance. The latter plays an important role during the cathode fabrication process. Low oxygen partial pressures occur during the thermal decomposition of the metal organic precursors of the cathode material, which lead to the formation of a beneficial hetero-interface of (La,Sr)₂CoO_4±δ and La_0.6Sr_0.4CoO_3-δ, as could be shown by thermodynamic calculations [4]. The influence of H₂O(g) and CO₂ on the electrochemical performance was investigated experimentally in a temperature range of 400 ... 650 °C [5]. Both gases negatively affect the solid/gas electrochemistry of the cathode (cf. Figure 1). CO₂ leads to poisoning of the cathode, which is reversible at temperatures above 500 °C. At temperatures below 500 °C a regeneration of the cathodes could only be achieved by a treatment in pure oxygen. In contrast to this behavior, H₂O leads to an irreversible and continuously increasing degradation. A detailed analysis of electrochemical impedance spectroscopy data disclosed, which polarization processes are majorly affected. CO₂ mainly affects the surface-exchange reaction, whereas H₂O – in the time frame of the test – impedes the oxygen ion diffusion within the cathode.

The results demonstrate the changeability of the solid-gas electrochemistry at mixed conducting cathodes. Based on these results operation strategies in terms of optimal operation temperature and required gas preprocessing (e.g. drying of oxidant gas) can be deduced, potentially decreasing the degradation during operation.

References

[1] S. B. Adler, Chem. Rev., 104(10), p. 4791 (2004).

[2] W. Lee, J. W. Han, Y. Chen, Z. Cai and B. Yildiz, J. Am. Chem. Soc., 135(21), p. 7909 (2013).

[3] J. Hayd, L. Dieterle, U. Guntow, D. Gerthsen and E. Ivers-Tiffée, J. Power Sources, 196(17), p. 7263 (2011).

[4] J. Hayd, H. Yokokawa and E. Ivers-Tiffée, J. Electrochem. Soc., 160(4), p. F351 (2013).

[5] J. Hayd and E. Ivers-Tiffée, J. Electrochem. Soc., 160 (11), p. F1197 (2013).

La₂NiO₄ and related materials with the K₂NiF₄ structure are known to support high electronic conductivity (about 100 S•cm^-1) and oxygen ionic conductivity (comparable to or exceeding that of yttria-stabilized zirconia).^{1, 2} Unlike other well known oxygen ion conductors, these materials conduct via an interstitial rather than a vacancy mechanism. While the K₂NiF₄ crystal structure is robust across a fairly wide solid solution compositional space, nearly all known compositions exhibit mixed ionic electronic conduction due to one or more transition-metals on the B-site. One exception, LaSrAlO₄,³ was previously found to be unstable to the introduction of ionic defects. For these reasons, despite high oxygen ion conductivity, oxides with the K₂NiF₄ structure have been limited in application as electrode and not electrolyte materials in solid oxide fuel cells (SOFC). In this work, we aimed to develop a new type of SOFC electrolyte material from transition-metal-free oxides with the K₂NiF₄ structure. The first successful composition, La_1.6Sr_0.4Al_0.4Mg_0.6O₄,⁴ was created by solid-state reaction. Oxygen ion defects were able to be created in the lattice by adjusting the La/Sr ratio. We hypothesize that defects are stabilized in this composition but not LaSrAlO₄ due to higher charge separation between the rocksalt and perovskite layers in the crystal lattice. Similar compositions substituting Al, Mg or Sr with Ga, Zn or Ca, respectively, were also synthesized and were confirmed to maintain this crystal structure. X-ray diffraction revealed that the crystal structure was consistently more stable to higher concentrations of vacancies relative to interstitial defects. Four-electrode conductivity measurements indicated that the B-site composition had a larger effect on total electrical conductivity than the A-site composition. Specifically, replacing Al or Mg with Ga or Zn, respectively, improved the conductivity by 1 to 2.5 orders of magnitude. Unfortunately, further electrochemical studies suggested significant hole conduction in these samples. Substituting Sr with Ca in the baseline composition had no effect on the total conductivity value, and La_1.6+xCa_0.4-xAl_0.4Mg_0.6O₄ series samples were all found to be pure ionic conductors. Nevertheless, the oxygen ion conductivity values achieved in the compositions to date remain less than technologically desirable. Results from experiments on newer compositions will show that Li can be placed on the B-site. Such compositions may have utility in both Li-ion and Li-air batteries.

References

1. H. S. Kim and H. I. Yoo, Physical Chemistry Chemical Physics, 2010, 12, 4704-4713.

2. V. V. Kharton, A. P. Viskup, N. E. N. and F. M. B. Marques, Journal of Materials Chemistry, 1999, 9, 2623-2629.

3. E. S. Raj, S. J. Skinner and J. A. Kilner, Solid State Sciences, 2004, 6, 825-829.

4. N. Ye and J. L. Hertz, Acta Materialia, 2013, http://dx.doi.org/10.1016/j.actamat.2013.10.013.

Figure Caption

Figure 1 (a) X-ray diffraction patterns of La_1.6Sr_0.4Al_0.4Ni_1-xMg_x-0.4O₄, (b) Arrhenius plots of the total electrical conductivities of La_1.6Sr_0.4Al_0.4Ni_1-xMg_x-0.4O₄ samples in air

8 mol % Y₂O₃-ZrO₂ [YSZ] has been widely used as electrolyte material in SOFCs. (Sc₂O₃)_0.1(CeO₂)_0.01(ZrO₂)_0.89 [SCSZ] has higher ionic conductivity in the intermediate temperature range (600^oC - 800^oC) compared to YSZ, but YSZ has better chemical and phase stability. In this work YSZ and SCSZ were used in developing layered electrolytes with a unique design to incorporate both materials, resulting in an electrolyte with enhanced ionic conductivity and improved robustness. The design involved placing SCSZ layers between two outer YSZ layers, so as to produce four- and six-layered electrolytes. Tape casting, lamination and pressureless sintering techniques were used in the development of the electrolytes. Due to the mismatch of the coefficient of thermal expansion between the two materials, thermal residual stresses arise between the layers. These stresses contribute to an enhancement of the ionic conductivity of the layered electrolytes. In addition, the compressive residual stress significantly affects the mechanical properties of layered electrolytes, and improves the electrochemical performance. Biaxial flexure strength was measured using ring-on-ring strength testing at room temperature and 800^oC. A finite element method was employed to calculate the maximum principal stress at fracture. The results showed that the layered YSZ/SCSZ/YSZ electrolytes have improved flexure strength at both room temperature and 800 ^oC because of the appearance of compressive residual stresses in the outer YSZ layers of the electrolyte. The calculated compressive stress values were also verified using Weibull statistics of strength data measured at room temperature.

Doped ceria is a good electrolyte candidate for intermediate temperature solid oxide fuel cells (IT-SOFC) because of its enhanced ionic conductivity. Segregation of dopants and impurities at grain boundaries may degrade the performance of these materials. This grain boundary segregation can be mitigated through manipulation of processing conditions. Recent studies have shown that, at 550°C, Sm_0.75Nd_0.75CeO_2-δ made by co-precipitation from nitrates followed by microwave sintering exhibits enhanced ionic conductivity (2.8 × 10^-3 S·cm^-1) as compared to the ball milled and conventionally sintered (1.9 × 10^-3 S·cm^-1) samples (1). Here we focus on the effects that the composition and processing have on the bulk and grain boundary contributions to ionic conductivity and gain a systematic understanding of the underlying processes. Scanning electron microscopy (SEM), electron energy dispersive spectroscopy (EDS) and electron backscattered diffraction (EBSD) studies indicate that the co-precipitated and microwave sintered samples exhibit smaller grain size, lower impurity content and less segregation at grain boundaries and triple point junctions as compared to conventional solid state processing route. These results are correlated with ionic conductivity data obtained from electrochemical impedance spectroscopy (EIS).

REFERENCES:

• S. Omar, "Enhanced Ionic Conductivity of Ceria-Based Compounds for the Electrolyte Application in SOFCs", Ph.D. Dissertation, University of Florida (2008).

Typical electrolyte materials in solid oxide fuel cells (SOFC) and solid oxide electrolyzer cells (SOEC) are yttria-stabilized zirconia (YSZ), Sr and Mg-doped LaGaO₃ (LSGM) or doped cerium oxide. These solid oxides exhibit good oxygen ion conductivity at elevated temperatures, ranging between 600^oC and 800^oC. It is preferable to lower the operating temperature to 600^oC, and possibly even lower. During the last two decades, considerable effort has been devoted to lower the ohmic resistance, primarily by using thin electrolyte film, anode-supported cell design. Over the past decade, considerable effort has also been devoted to lowering the cathode polarization resistance. This has led to the development of a number of excellent cathodes, so that at least in many cell designs, the cathode is no longer the component which limits cell performance. Our recent work [1] has shown that the ohmic loss may actually be the dominant loss even in anode-supported thin electrolyte cells, necessitating additional efforts to further reduce the ohmic contribution to the net cell resistance. It is known that at low temperatures, the electrolyte resistance is dominated by grain boundary resistance, which in turn is often dictated by space charge effects.

The purpose of the present work was to examine the role of space charge on grain boundary contribution to the electrolyte resistance. Figure 1 shows that grain boundary resistance dominates the overall ohmic resistance of a typical electrolyte (in this case GDC) at medium-to-low temperature range. Thus, it is deemed necessary to determine parameters that control ion transport across grain boundaries. It is known that grain boundary resistance usually is attributed to either a silicate phase and/or oxygen vacancy depletion. As the purity of raw materials has been improved greatly over the years, oxygen vacancy depletion is considered to be the primary reason for the grain boundary resistance [2]. The objective of this work is to explore the effect of space charge on conduction properties when one of the defects is far more mobile than the other defects. In such a case, at low temperatures, only the mobile defects can move, but the less mobile defects are essentially frozen from the high temperature annealing step. The present work first examines an example of a simple, undoped NaCl system. Later, the work will be extended to YSZ and other oxide systems of interest. This work is based on the classic work of Kliewer and Koehler [3]. Figure 2 shows that in order to maintain charge neutrality in the grain interior of a grain of sufficiently large thickness, the electrostatic potential inside the grain should be negative for NaCl, which is realized by anion vacancy depletion near the surface. Supposing both cations and anions are mobile and have reached their equilibrium spatial positions at each temperature, Figure 2 shows that, as the temperature increases, the grain boundary thickness (the region of space charge) decreases and both cation and anion vacancy concentrations increase. If, however, the anion vacancies are less mobile and thus are frozen when the sample is cooled to a lower temperature, equilibrium will be attained by the re-distribution of cation vacancies to minimize the free energy.

Figure 3 shows the corresponding cation vacancy profile as a function of distance from the crystal surface (grain boundary) at various temperatures. Assuming much of the ionic transport occurs by cation vacancies (this for NaCl), the grain boundary resistance is expected to be different in the two cases. Similar calculations are conducted on oxygen ion conductors such as doped zirconia and ceria. The results of the calculations and the corresponding experimental results will be presented.

Funded by DOE EFRC Grant Number DE-SC0001061 as a flow-through from the University of South Carolina.

References:

1. Liangzhu Zhu, Lei Zhang, Feng Zhao, Virkar A.V., ECS transcription (2014)

2. X. Guo, J. Maier, J. Electrochem. Soc., 145, E121, 2001

3. K. L. Kliewer, J. S. Koehler Physical Review, 140, 4 (1965)

Fig. 1A and 1C: Electrochemical impedance spectra on a GDC electrolyte; 1B and 1D: Arrhenius plots of grain and grain boundary resistance.

Fig. 2: Simulated anion vacancy distribution. Both cations and anions are assumed mobile and to have reached equilibrium distributions at each of the temperatures.

Fig. 3: Simulated cation vacancy distribution assuming the anion vacancies to be frozen at the highest annealing temperature.

1. Introduction

Finding out new candidates that alternative typical YSZ electrolyte was researched for long periods to overcome the low conductivity at low temperature. dopants doped ceria-based oxide electrolytes such as Sm (SDC) or Gd (GDC) considered as attractive material due to their high ionic conductivity at low temperature but electron conduction was observed at low oxygen partial pressure because of reduction reaction of Ce⁴⁺ to Ce³⁺ [1]. In previous research, H. Iwahara el al. reported various perovskite materials can shows oxide ionic conduction behavior by partial substitution of lower valence such as La_0.8Sr_0.2AlO_3-δ, CaTi_0.8Al_0.2O_3-δ and so on[2]. T. Ishihara et al. discovered Lanthanum gallium-based oxide (La_0.8Sr_0.2Ga_0.8Mg_0.2O_3-δ) having perovskite structure was regarded as promising electrolyte due to their high stability at low temperature as well as high ionic conductivity although La and Ga are considered as expensive elements [3]. Meanwhile, S. Hashimoto el al. reported CaTi_0.9Sc_0.1O_3-δ has comparable ionic conductivity with YSZ electrolyte at 800^oC although the fair ionic transportation number at other temperature is unclear [4].

In this research, conductivity behavior of CaTi_1-xSc_xO_3-δ (x=0.05, 0.1) was analysed as function of P(O₂) dependence at various temperature from 500^oC to 1000^oC. Based on the conductivity behaviour, the ionic transference number of CaTi_1-xSc_xO_3-δwas calculated from defect chemistry point of view for mixed-conductor.

2. Experimental Procedures

Commercial CaCO₃, TiO₂, Sc₂O₃were used as starting materials and synthesized by a conventional solid state reaction method depending on required ratio. The mixed powders were calcined in air at 1200ºC for 10h and grounded by satellite-type ball milling process for 2h using ethanol. The calcined powder (x=0.05, 0.1) was compacted into a bar with hydrostatic pressure and sintered at 1550ºC for 10h in air.

Lattice parameter and phase characterization were investigated using X-ray diffraction (XRD) analyser with CuKα. To confirm the electrical properties of CaTi_1-xSc_xO_3-δ (x=0.05, 0.1), impedance analysis was carried out as function of oxygen partial pressure, P(O₂)(10^-30to 1bar) and temperature (1273-773K) assisted by 4-terminal method using chemical impedance meter (Hioki 8334-30, Japan).

3. Results and Discussion

Total conductivity behavior was shown depending on P(O₂) and temperature and fitted total conductivity (solid line) based on experimental analysis were shown as Fig. 1. Fitting process was carried out to separate total conductivity by conduction parameter (σ_elec, σ_ion σ_hole) as function of P(O₂). Conductivity was increased with increasing temperature and also decreased at high P(O₂) region and increased at low P(O₂) region showing typical mixed ionic-electronic conductor (MIEC). On the contrary to this their ionic domain range was decreased with increasing temperature. These kinds of phenomenon were considered as dominant hole and electron conduction due to generation of holes and electrons at high temperature. In addition, effect of electron conduction seemed like not so strong compare with hole conduction with decreasing temperature. Moreover, conductivity was increased with increasing of Sc content. Sc seemed not to improve conductivity in case of CaTi_0.95Sc_0.05O_3-δ. But if Sc content was increased, conductivity was increased and showed high conductivity compare with YSZ, contrastively. This behavior observed more clearly low temperature than high temperature.

References

[1] B. Cales and J. Baumard, J. Phys. Chem. Sol, 45(8/9), (1984), 929-935

[2] T. Ishihara, H. Minami, H. Matsuda, H. Nishiguchi and Y. Takita, Chem. Commun, (1996), 929-930

[3] T. Takahashi and H. Iwahara, Energy Conv. 11(1971), 105-111

[4] S. Hashimoto, H. Kishimoto and H. Iwahara, S. S. Ionics, 179, (2001), 179-187

The search for suitable electrolytes for intermediate temperature fuel cells remains a formidable challenge. Intermediate-temperature (100-400 °C) devices would enable the development of simple, low cost fuel cells, free from the water management and fuel flexibility issues of polymer membrane fuel cells and without the system and materials challenges of high temperature fuel cell systems ¹.

Proton transport in inorganic phosphate crystalline compounds and glasses has been reported for many systems such as LaPO₄, CsHPO₄, metal pyrophosphates (MP₂O₇) and a variety of composite glass ceramic materials.

Nagao et. al ² reported in 2006 that the anhydrous proton conductor In doped SnP₂O₇ exhibited ionic conductivity on the order of 0.1 S/cm at temperatures of 150-350°C in water-free atmospheres. Attempts to reproduce the high conductivity material have had varying degrees of success ^3-6. Since the initial reports, many new publications on pyrophosphates have appeared. The authors of these reports attribute the proton conductivity of these materials to the generation of oxygen vacancies via aliovalent doping with concurrent hydrolysis of water filling the vacant sites and thus creating mobile protons. The mechanism cannot reconcile the high conductivities of the parent un-doped MP₂O₇materials and the high conductivity of materials with phosphate to metal ratios greater than 2.

Other investigators have hypothesized that amorphous, phosphate rich phases derived from the material synthesis precursors may remain in grain boundaries and serve to enhance conductivity, as suggested for proton conducting LaPO₄ materials⁷, Sn_0.9In_0.1P₂O₇⁸ and also demonstrated for un-doped SnP₂O₇⁶.

We have extensively studied the temperature and atmosphere dependent conductivity, thermal behavior, crystallography^{9, 10} and hydrogen vibration spectra of In³⁺-doped SnP₂O₇systems, varying the In concentration, and the metal to phosphate ratios.

Inelastic neutron scattering on dehydrated, hydrated and deuterated samples was particularly useful for probing the dynamics of hydrogen within the material and for identification of the hydrogen bonding within the materials. A complex picture of low temperature proton transport dominated by an amorphous polyphosphate intergranular material and a high temperature conductivity that is influenced by indium doping has arisen.

Our study provides insights towards the proper investigation of phosphate materials and perhaps, new strategies to produce composite ceramic-glass proton conductors. The effects of crystalline/amorphous interfacial interactions also cannot be discounted.

References

1. O. Paschos, J. Kunze, U. Stimming and F. Maglia, J Phys Condens Matter 23(23), 234110 (2011).

2. M. Nagao, A. Takeuchi, P. Heo, T. Hibino, M. Sano and A. Tomita, Electrochemical and Solid-State Letters 9(3), A105 (2006).

3. S. R. Phadke, C. R. Bowers, E. D. Wachsman and J. C. Nino, Solid State Ionics 183(1), 26-31 (2011).

4. S. Tao, Solid State Ionics 180(2-3), 148-153 (2009).

5. Y. Sato, Y. Shen, M. Nishida, W. Kanematsu and T. Hibino, Journal of Materials Chemistry 22(9), 3973 (2012).

6. X. Xu, S. Tao, P. Wormald and J. T. S. Irvine, Journal of Materials Chemistry 20 (36), 7827 (2010).7. G. Harley, R. Yu and L. Dejonghe, Solid State Ionics 178(11-12), 769-773 (2007).

8. M. L. Einsla, R. Mukundan, E. L. Brosha and F. H. Garzon, ECS Transactions 11(1), 347-355 (2007).

9. C. R. Kreller, M. S. Wilson, R. Mukundan, E. L. Brosha and F. H. Garzon, ECS Electrochemistry Letters 2(9), F61-F63 (2013).

10. C. R. Kreller, M. S. Wilson, R. Mukundan, E. L. Brosha and F. H. Garzon, ECS Transactions 57(1), 1009-1018 (2013).

Acknowledgements

The authors would like to acknowledge the financial support of the US Department of Energy, Los Alamos Internal Directed Research Program (LANL-LDRD) for financial support

Neodymium doped ceria (NDC) is a promising candidate for use as an electrolyte in a solid oxide fuel cell (SOFC) due to its high ionic conductivity, low electronic conductivity, and good sinterability and stability. It is well known that grain size and grain boundary effects can significantly impact the observed ionic conductivity of ceria in the bulk and particularly in thin films. This work focuses on the influence grain size (film thickness) and grain boundary configuration have on the ionic conductivity of NDC. To this end, textured NDC thin films of varying thicknesses and grain sizes were grown on sapphire substrates using pulsed laser deposition (PLD). We will present broadband impedance data analysis on across plane and in-plane electrode configurations collected as a function of temperature. The separate effects of film thickness, grain size, and grain boundary on the observed ionic conductivity will be discussed based on the contrasting data from the two measurement configurations.

Recently, solid electrolytes with reasonable room temperature lithium ionic conductivity are being examined for high energy density batteries with safe operation. Several solid state electrolytes based on sulfides and oxides have been investigated [1]. Among them, Garnet phase materials have shown significant improvement in conductivity [2]. Garnet has cubic and tetragonal phases of which the cubic is high temperature with a high ionic conductivity value of 10^-4 S/cm [3]. Several doped compositions have also been reported in order to improve the cubic phase ionic conductivity [4]. LLTO Perovskite system has also been reported for high ionic conductivity value of 10^-4 S/cm [5,6]. Currently, we have synthesized a Ti, Cr, Ta doped Li₇La₃Zr₂O₁₂and LLTO by a simple Pechini process to investigate their ionic conductivities.

A Pechini method was used to synthesize these compoundsusing nitrate salts, ethylene glycol and citric acid in a 38:36:26 ratio respectively in de-ionized water. The samples were dried in an oven overnight at 120°C and then heat-treated at different temperatures to achieve the desired phases. The optimized product was heat treated at 1000°C for six hours, pelletized and then sintered again at higher temperatures (1200°C) to perform the ionic conductivity measurements using impedance spectroscopy.

Samples were characterized by XRD for phase analysis and electrochemical impedance performance has been investigated under varying temperatures and voltages. The effect of doping on the phase transition in LLZO and LLTO will be presented. Figure 1 shows the impedance plot for the Ta doped garnet (Li₇La₃Zr_1.5Ta_.0.5O₁₂) at room temperature. Ionic conductivity value of 1.27(x10^-4) S/cm has been obtained.

Detailed impedance measurements on systematic doping of Ti, Cr, Ta etc. in LLZO and LLTO phases using various sample preparation conditions, various ohmic contacts for the pellets will be discussed.

Figure. 1 Impedance plot for Li₇La₃Zr_1.5Ta_0.5O₁₂ at room temperature.

Acknowledgements

This work has been supported by the US Dept. of Energy/NETL, EERE program. FRB acknowledges Oak Ridge Institute for Science and Education (ORISE) fellowship.

References

[1] J.W. Fergus, Journal of Power Sources, 195 (2010) 4554-4569.

[2] E. Rangasamy, J. Wolfenstine, J. Sakamoto, Solid State Ionics 206 (2012) 28-32.

[3] I. Kokal, M. Somer, P.H.L. Notten, H.T. Hintzen, Solid State Ionics 185 (2011) 42-46.

[4] Y. Jin, P.J. McGinn, Journal of Power Sources 196 (2011) 8683-8687.

[5] B. Antoniassi, A. H. M Gonzalez, S. L. Fernades, C. F. O. Graeff, Materials Chemistry and Physics 2011, 127, 51.

[6] O. Bohnke, Q.N. Pham, A. Boulant, J. Emery, T. Salkus, M. Barre, Solid State Ionics 2011, 188, 144.

The unique photoelectric properties of TiO₂, which make it a better choice of dye-sensitized solar cells with high efficiency, have attracted a lot of interest in recent years. Especially, single-crystal rutile TiO₂ nanorod arrays have been extensively investigated. In this study, in order to improve the photoelectric properties of TiO₂ and analyze the effects of different electrolytes on the performance of TiO₂ based dye-sensitized solar cells, the solar cells were fabricated with iodine base electrolyte, sulfur based electrolyte and inorganic solid state electrolyte, CsSnI_2.95F_0.05, as electrolyte respectively, and Pt deposited on the conductive face of fluorine-doped tin oxide (FTO) glass as the counter-electrode. TiO₂ nanorod arrays were treated with ethanol solution of cis-diisothiocyanato-bis-(2,2^,-bipyridyl-4,4^,-dicaboxylato)-ruthenium()bis(tetrabutyl-ammonium) (N719, DYESOL) for 18h before the cell assembly. Open-circuit voltage (V_OC), short-circuit current density (J_SC) and impedance of the solar cells with these three kinds of electrolytes were measured respectively, and then their conversion efficiencies were calculated. The TiO₂ based solar cell fabricated with CsSnI_2.95F_0.05 showed the highest conversion efficiency of about 8.9%, which was much higher than those of the solar cells fabricated with the other two kinds of electrolytes.

Some perovskite-type oxides have shown high proton conductivity in wet hydrogen atmosphere. In these perovskite oxides the partial substitution of an acceptor dopant for B-site results in formation of oxygen vacancies, which lead to the formation of protonic defects by the absorption of water in wet atmosphere. Among various proton conducting perovskite-type oxides, Y-doped Barium cerate and barium zirconate are nearly isomorphic. Therefore, the basic phases of a binary solid solution where B-sites are randomly occupied by either Zr⁴⁺ or Ce⁴⁺. This is because the chemical instability of Y-doped BaCeO_3-δ in the presence of CO₂-, H₂O- or SO₂-containing atmosphere below 973K was substantially improved by replacing Ce with Zr to form a solid solution, and the mechanical strength was also enhanced when the proton conductivity and thermal stability remained adequate over a wide range of operation conditions for the fuel cell.

To elucidate the transport properties and defect structure of a Ce/Zr-coexisting preovskite structure, BZCY(BaZr_xCe_1-xY_0.15O_3-δ, x=0, 0.2, 0.4, 0.6) was chosen as the model system. First, we study the electrical properties of BZCY in various oxygen and water vapor atmospheres by four-probe DC measurements as a function of temperature. And then, these properties are analyzed analytically, based on the defect structure of the material. Furthermore, we use the conductivity relaxation measurement technique to monitor the monotonic/non-monotonic(two-fold) relaxation for the BZCY under various themodynamic conditions and extract the chemical diffusivities by solving Fick's second law for each case on the basis of hydrogen- and oxygen- decoupled diffusion as proposed by Professor Yoo^1-3.

Fig. 1 shows the typical relaxation profiles of the mean total conductivity upon oxidation and reduction across an identical oxygen activity window. During the redox reaction driven by the oxygen chemical potential gradient at a fixed water vapor activity, oxygen may be incorporated into the BZCY by ambipolar diffusion of V^••_o and 2h^• in the p-type oxidizing regime. In Fig. 2, conductivity relaxation profiles upon hydration and dehydration at a fixed pO₂(≈0.21 atm) are shown as a function of water vapor activities at 1023K. The figures clearly reveal the violation of monotonic relaxation governed by the chemical diffusion of water, according to the ambipolar diffusion of oxygen vacancies and proton. As shown in the fig. 2, non-monotonic twofold relaxation behavior was clearly observed for the BCZY system.

1. H.I. Yoo et al., Phys Chem Chem Phys, 10, 974 (2008)

2. H.I. Yoo et al., Solid State Ionics, 180, 1443 (2009)

3. J.I. Yeon et al., Solid State Ionics, 181, 1323 (2010)

4. S. Hamakawa et al., Solid State Ionics, 148, 71 (2002)

5. S.-J. Song et al., Solid State Ionics, 167, 99 (2004)

6. S.M. Haile et al., Nature, 125, 271 (1999)

7. D. Pergolesi et al., Nature Materials, 9, 846 (2010)

8. H. Iwahara et al., Solid State Ionics, 125, 271 (1999)

9. Y.C. Yang et al., Sensors and Actuators B, 140, 273 (2009)

10. C. Zuo et al., Advanced Materials, 18, 3318 (2006).

Lithium-stuffed garnets have become one of the most studied lithium ionic conductors in the oxide system owing to the high ionic conductivity and good stability [1, 2]. The compound Li₇La₃Zr₂O₁₂ (LLZ) was reported to crystallize in either the cubic or tetragonal symmetries in different studies [3, 4]. The tetragonal phase transformation was accompanied by a complete ordering of Li. It has been proved that cation doping could stabilize the cubic phase.[5] Alternatively, H₂O and CO₂ were shown to be responsible for phase transition in aged LLZ samples.[6] In this study, we examine the effect of Ta doping and H₂O/CO₂ exposure on the phase transition of LLZ materials using diffraction techniques and in-situ impedance measurement. It's important to note that the strict control over synthesis condition is a necessity for carrying out such study and we took cautions to ensure purity.

We have shown computationally that the Li-ordering may be driven by the increasing first neighbor Li-Li repulsion as Li content increases.[7] Therefore, any cation doping strategy that results in a lower Li content has the potential of disrupting Li-ordering thus transforming the phase. As shown from the X-ray diffraction results on the series of carefully prepared Ta-doped LLZ (Li_7-xLa₃Zr_2-xTa_xO₁₂, x=0-0.6, Al-free, minimal air exposure), the intermediate compositions (x=0.1-0.5) consist of coexisting cubic phase and tetragonal phase possibly with different chemical makeup, contradicting results in previous studies.[8] It is likely that the high conductivity phase Li_6.75La₃Zr_1.75Ta_0.25O₁₂ (LLZT0.25) is not thermodynamically stable. Synchrotron experiments (APS proposal accepted) are expected to help us determine the phase fraction and exact composition. Two possibilities could explain the cubic phase of high Li content compositions in previous study: 1) unintentional doping of Al from firing medium which further lowered Li content; 2) phase transition induced by extended exposure to the air. To evaluate the effects of moisture and CO₂ on phase transition and conductivity of LLZ, we measured the impedance change of impurity-free LLZ pellet in-situ under H₂O or CO₂ gas flow at different temperatures and characterize the phase ex-situ. In the control group, conductivities of LLZ sample up to 750 °C under argon were measured and the phase transition at around 630 °C was clearly evident from the Arrhenius plot. On the other hand, tetragonal LLZ transformed to cubic phase upon exposure to moisture at 250 °C or CO₂ at 120 °C. Impedance measurement indicated that despite having cubic symmetry, the transformed LLZ had a lower conductivity than the initial phase at room temperature in both cases. It should be noted that the phase transition of LLZ under argon and under H₂O or CO₂ are different in nature: the former is first order phase transition due to increasing entropic contribution; the latter may be due to compositional changes, for instance, removal of Li ions and/or incorporation of protons in garnets.

Figure 1. (a) Arrhenius plot of LLZ under argon flow; (b) X-ray diffraction patterns showing the tetragonal-to-cubic phase transition induced by H₂O and CO₂; (c) X-ray diffraction patterns of the series Li_7-xLa₃Zr_2-xTa_xO₁₂, x=0-0.6.

References

[1] R. Murugan, V. Thangadurai and W. Weppner, Angew Chem Int Edit, 46, (2007), 7778-7781.

[2] V. Thangadurai, H. Kaack and W. J. F. Weppner, J Am Ceram Soc, 86, (2003), 437-440.

[3] J. Awaka, N. Kijima, H. Hayakawa and J. Akimoto, J Solid State Chem, 182, (2009), 2046-2052.

[4] J. B. Goodenough, H. Xie, J. A. Alonso, Y. T. Li and M. T. Fernandez-Diaz, Chem Mater, 23, (2011), 3587-3589.

[5] A. Logeat, T. Koohler, U. Eisele, B. Stiaszny, A. Harzer, M. Tovar, A. Senyshyn, H. Ehrenberg and B. Kozinsky, Solid State Ionics, 206, (2012), 33-38.

[6] G. Larraz, A. Orera and M. L. Sanjuan, J Mater Chem A, 1, (2013), 11419-11428.

[7] Y. H. Wang, A; Lai, W, Solid State Ionics, (2013).

[8] Y. X. Wang and W. Lai, Electrochem Solid St, 15, (2012), A68-A71.

Ba₃Ca_1.18Nb_1.82O_9-δ (BCN₁₈) is a complex perovskite, which exhibits excellent proton conduction when hydrated. It is also known to be more stable compared other proton conductors such as doped BaCeO₃ and doped SrCeO₃. BCN₁₈is thus a potential electrolyte for fuel cells and for hydrogen separation. It is generally recognized that efficacy of proton transport at least partly depends on the local atomic environment, which implies that B-site cation ordering and oxygen vacancy occupation site may play important roles, making the study of the B-site cation ordering and oxygen vacancy position of interest.

Earlier studies^1,2 have shown the important role played by electrostatic effects between different ionic species in driving the ordering process. Some recent works^3,4, however, have suggested the ordering is related to the large ionic radius difference (~50%) between Ca²⁺ (1.14 Å) and Nb⁵⁺ (0.78 Å).

Using density functional theory calculations, we investigate the physical mechanism underlying the formation of the B-site cation ordering and the oxygen vacancy site selection in Ba₃CaNb₂O₉. We found that either cation site exchange or oxygen vacancy formation induces negligible lattice strain. This implies that the ionic radius plays a minor role in governing these two processes.

We also found that the electrostatic interactions are dominant in the ordering of mixed valence species on one or more sites; the ionic bond strength is identified as the dominant factor in governing both the 1:2 B-site cation ordering along the <111> direction and the oxygen vacancy site preference in Ba₃CaNb₂O₉. Starting from the 1:2 fully ordered atomic structure, we exchange one of the Ca atoms with one of its surrounding Nb atoms to study the B-site cation ordering in BCN. As shown in Fig. 1, there are several possible exchange types, denoted as 1-1x, 1-1y, 1-1z, 2-2x, 2-2y, 2-2z, and 2-2z', respectively. Specifically, the cation ordering can be rationalized by the preference of mixed Ca-O-Nb bonds over the combination of Ca-O-Ca and Nb-O-Nb bonds; while oxygen vacancy prefers a site to minimize the electrostatic energy and to break the weaker B-O-B bond, as shown in Fig. 2.

Funded by DOE EFRC Grant Number DE-SC0001061 as a flow through from the University of South Carolina.

References:

1. L. Bellaiche, D. Vanderbilt, Phys. Rev. Lett.81, 1318 (1998).

2. S. Bhide, A. Virkar, J. Electrochem. Soc.146, 4386 (1999).

3. J. Deng, J. Chen, R. Yu, G. Liu, X. Xing, J. Alloy Compd.472, 502 (2009).

4. Q. Zhou, B.J. Kennedy, J.A. Kimpton, J. Solid State Chem.184, 729 (2011).

Fig. 1 (a) Demonstration of the Ca/Nb exchange types, where 1 and 2 are Ca atoms, 1x, 1y, 1z, 2x, 2y, 2z, and 2z' are the exchange processes involving Nb atoms. (b) Relative system energy (exchange energy) upon cation exchange. "Initial" denotes the 1:2 ordered structure, while the others denote the different exchanged structures.

Fig. 2 (a) Relative energies for oxygen vacancy formation sites vs. the number of Ca atoms in the NNN layer. Black squared denotes the oxygen vacancy formation in AO layers, and red circled denotes the oxygen vacancy formation in BO₂ layers. (b) O ion effective charge vs. the number of Ca atoms in the NNN layer. The number of Ca in the NN layer is fixed at 1.

Conventional analysis of data recorded using the Electrochemical Impedance Spectroscopy (EIS) is a relatively simple procedure from the mathematical point of view. It involves an elucidation of electrode processes to derive their characteristic parameters. Another advantage is that it represents a quick visualization tool being still a very sensitive technique [1,2]. However, it has a number of disadvantages too. Often, the interpretation of EIS data using equivalent circuits is quite difficult due to a tedious search for a good fitting correlation. Moreover, results do not give information about number of moving ions (cations or anions), hopping time etc. Knowledge and consideration of these parameters would be very helpful in many electrochemistry-related fields, such as for instance in the development of solid state batteries and in the semiconductor field.

Being unhappy with these drawbacks and limitations of the conventional data analysis method, we recently developed a new approach [4] in which the Z₁ - Z₂ complex impedance plane (where Z₁ and Z₂are the real and imaginary parts of the impedance of materials) has been analyzed, based on Dyre's random-walk theory. Through this approach we have obtained from EIS data (that are measured anyway), a new set of physical parameters, yet unseen and unmined:

i) the diffusion coefficient D and

ii) the number of moving ions N_ions,

both parameters further distinguishable in the bulk region of the sample as well as at the interface.

The presentation will explain in detail our recently developed approach that has the potential to find a widespread use in various electrochemical fields, such as in batteries, solar cells, fuel cells and semiconductor industry in general. The presentation will show, how helpful this approach can be to understand the electrode polarization as well as the ionic transport mechanism in various conductors. In particular, we will discuss in detail recent results achieved on various materials, including selected ionic conductors [3-5], silicons and titanium dioxides [6].

Literature:

• D. D. Macdonald, Transient Techniques in Electrochemistry, Springer US, 1977.

• E. Barsoukov, J. R. Macdonald, Impedance Spectroscopy Theory, Experiment, and Application, Second Ed., John Wiley and Sons, New Jersey, 2005

• S. Stehlik, J. Orava, T. Kohoutek, T. Wagner, M. Frumar, V. Zima, T. Hara, Y. Matsui, K. Ueda, M. Pumera, J. of Solid State Chem. 183 (2010) 144.

• S. Stehlik, K. Shimakawa, T. Wagner and M. Frumar, J. Phys. D: Appl. Phys. 45(2012) 205304.

• D.S. Patik, K. Shimakawa, V. Zima, J. Macak, and T. Wagner, J. Appl. Phys. 113 (2013) 143705.

• T.Wagner et al., Ms in preparation.

Electrochemical impedance spectroscopy (EIS) is frequently used in conjunction with simple equivalent circuit models to quantify grain, grain boundary, and electrode properties in electrochemical cells involving ionic or mixed conducting ceramics. Constant phase elements (CPEs) are often used in the equivalent circuit to obtain a better fit of experimental data. However, the CPE is a mathematical abstraction with no clearly defined physical meaning, and its use remains poorly justified. Depression of the grain boundary impedance arc was previously shown to occur in samples with heterogeneous grain boundary conductivity [1]. Here, we use numerical modeling to investigate the utility of CPE-based equivalent circuit fitting in quantifying heterogeneous grain boundary conductivity and/or permittivity.

A 2D electrical model of a polycrystalline ceramic conductor with individually defined grain boundary properties was developed, allowing impedance spectra to be generated for arbitrary distributions of grain boundary conductivity and/or permittivity. For the present study, the relevant properties of each grain boundary were assigned randomly according to a specified distribution. Heterogeneous conductivity values followed a log-normal distribution, while heterogeneous permittivity values followed an exponential distribution with a maximum equal to the bulk grain permittivity. Simulated impedance spectra (Fig. 1) were produced by solving Poisson's equation and fitted to an equivalent circuit consisting of two parallel R-CPE elements in series. Grain boundary conductivity and permittivity values were calculated from the circuit parameters R_gb, Q_gb, and n_gb.

As expected, increasingly heterogeneous grain boundary properties correlated to increasingly depressed impedance arcs and reduced CPE exponent. When only the conductivity or the permittivity was heterogeneous, the spread of parameter values could be estimated from the CPE exponent. However, this was no longer possible when both parameters were heterogeneous. As one might hope, in all cases, the grain boundary conductivity determined from the CPE-based equivalent circuit was within ±30% of the actual mean grain boundary conductivity. Conversely, simple use of the CPE's Qparameter as a capacitance results in severe error in calculating the mean grain boundary permittivity. A commonly used equivalent capacitance expression given by Brug [2] still caused the calculated permittivity to deviate by up to 60% from the actual mean value. However, a new empirical equation [3] allowed a more accurate estimate that differed from the mean by no more than 35% (Fig. 2). Recent extensions of this work to quantification of the impedance of the electrode–solid electrolyte interface will also be presented.

[1] J. Fleig, Solid State Ionics 131 (2000) 117–127.

[2] G.J. Brug et al., J. Electroanal. Chem. Interfacial Electrochem. 176 (1984) 275–295.

[3] B.E. McNealy and J.L. Hertz (under review).

The dependence of oxygen non-stoichiometry on oxygen partial pressure and temperature is critical to many of the applications of non-stoichiometric oxides in electrochemical energy conversion processes. It has been increasingly recognized that electrochemical methods, in particular, AC impedance spectroscopy, can be used to evaluate this non-stoichiometry. The method is particularly useful for determining the properties of thin films, for which conventional thermogravimetric methods are inapplicable. To date, analysis of the ACIS data to extract non-stoichiometry values has relied on application of various simplifying assumptions for the thermodynamic behavior, limiting the approach to oxides with ideal solution behavior. In this work, we determine the oxygen non-stoichiometry of cerium oxide, which displays non-ideal behavior under certain thermodynamic conditions. The chemical capacitance of single crystal, epitaxial thin films of cerium oxide deposited on yttria stabilized zirconia substrates is measured and related to the oxygen non-stoichiometry using thermodynamic principles. This work will demonstrate the validity of the chemical capacitance measurements to reliably extract non-stoichiometry of a generalized oxide system.