Search

International audience; Mixed Ionic and Electronic Conductors (MIECs) are promising membrane materials for oxygen separation fromair at high temperature. The transient stage is critical due to the stress induced by the chemical strain, combinedwith the oxygen activity field through the membrane. In the literature, permeation models are able to predict theoxygen flux in steady state or to identify the slowest mechanism. In this paper, the proposed model aims topredict the oxygen activity fields through the membrane both in transient stage and steady state. A macroscopicapproach is proposed, the balance of a temporary species is introduced to take into account separately thedissociative adsorption, associative desorption and interface reaction between surface and bulk. New densitiesof probability are proposed to model dissociative adsorption and associative desorption in order to take intoaccount the difference between “classical” catalysis and the MIECs case. This model is implementing in the finiteelement software COMSOL Multiphysics© thanks to the combination of the modules “coefficient form PDE”and “coefficient form boundary PDE”. The results of the model are in good agreement with steady state andtransient stage experimental data. The different kinetics of oxygen release and oxygen intake are wellreproduced.

This work aims at studying the volume change of bauxite corroded by a molten slag. Cylindrical samples were prepared by mixing ground bauxite with slag. Optical measurement at high temperature (1450 °C) of deformation with a high-resolution camera has been developed. Image processing allowed for determining the change in diameter of the sample. We showed that the deformation was induced by the precipitation of new expansive crystallised phases observed by SEM-EDS analyses. Adding pellets of the same slag upon the samples allowed to emphasize the effect of the slag amount on the size change. The change in diameter significantly increased in the impregnated area.

Mixed ionic and electronic conductors (MIEC) are promising materials for ion transport membrane (ITM) based technologies. A reliable design of these membranes requires a better knowledge of their macroscopic mechanical properties. These properties are linked to the composition of membrane materials, but also in a large part to the manufacturing process. Then, the mechanical characterization set-up must allow using specimen very similar in term of manufacturing process as the final parts in. The use of the diametric compression test on thin plates is investigated here, combined with full-field measurements. Guideline for sample size, solution for classical artifact and a dedicated I-DIC algorithm are proposed. Then, Young modulus and tensile strength of seven MIEC compositions of the lanthanum-ferrite perovskite series (La (1− x ) Sr x Fe (1− y ) Ga y O 3−δ , La 0.5 Ba 0.5 Fe 0.7 Co 0.3 O 3−δ and La 0.5 Ca 0.5 Fe 0.7 Co 0.3 O 3−δ ) were determined at room temperature and at 900 °C.

International audience; The transient stage is critical due to the stress induced by the chemical and thermal strain. In order to predict this strain, the oxygen activity field through the membrane needs to be known. Usually, the membrane is divided into three zones: the bulk where diffusion takes place and the two surfaces where exchanges between atmosphere and membrane take place. Oxygen bulk diffusion is well described by the Wagner theory. A consensus has not yet emerged regarding the surface exchange models proposed in the literature. Moreover, these models describe the permanent state, and cannot be extended to the transient stage. A new macroscopic surface exchange model which allows computing transient stage is proposed. This model assumed that the oxygen flux is governed by the association/dissociation of adsorbed oxygen and by the high energetic cost of oxygen reduction/oxidation. Then, the balance of transient specie only present on the surface is introduced to account for these two phenomena. The oxygen activity fields predicted by the proposed model are in agreement with the measures of chemical potential drop between the membrane and the atmosphere in permanent state. Transient stage measured during isothermal expansion test is partially reproduced.

Mixed Ionic and Electronic Conductors are promising materials for oxygen transport membranes at high temperature. The mixed conduction leads to oxygen semi-permeation properties. This oxygen transport properties is linked to the ability of the membrane material to reversibly adsorb and desorb oxygen and to diffuse oxygen through the lattice [1]. However, the defect equilibrium with the surrounding atmosphere induces a “chemical” expansion in the same order than the thermal expansion. Mechanically, the transient stage is critical due to the stress induced by the chemical and thermal strain [2]. To predict this strain, the oxygen activity field through the membrane needs to be known. Usually, to model, the membrane is divided into three zones: the bulk, where oxygen bulk diffusion takes place and the two surfaces where oxygen exchanges between atmosphere and membrane take place. Oxygen bulk diffusion is well described by the Wagner theory [3]. The surface exchanges include the adsorption/desorption, dissociation/association and incorporation/exclusion of oxygen. A consensus has not yet emerged regarding the oxygen surface exchange models proposed in the literature. Two categories of models are reported: one corresponds to an extension of the Wagner theory to the surface [4] and the second corresponds to a chemical approach by kinetic laws [1, 5]. Different experiments have been performed to characterize the surface exchange and bulk diffusion parameters. The measures reported by Geffroy et al. on membrane surfaces during steady oxygen flow exhibits a large chemical potential gap between the membrane surfaces and the surrounding gases [6]. From our knowledge, this experimental data cannot be reproduced by a single model from literature. Moreover, although the combination of Wagner extension approach on the high oxygen side and kinetic approach on the low oxygen side makes it possible to reproduce results in steady state, it does not allowed to investigate transient stage. The proposed talk deals with a new macroscopic surface exchange model describing both oxidizing and reducing surface for transient stage. This model assumes that the oxygen flow is governed by the association/dissociation of adsorbed oxygen and by the high energetic cost of oxygen reduction/oxidation. Then, the balance of a transient species at the membrane surface is introduced to account for these two phenomena. The first results obtained are in accordance with oxygen semi-permeation measurements reported by Geffroy et al. [6]. The computation of isothermal expansion tests realized by Adler [7] in transient stage, using the proposed exchange model and the chemical expansion model proposed by Valentin et al. [8], present also a good correlation with experimental results. [1] S.B. Adler et al, Mechanisms and rate laws for oxygen exchange on mixed-conducting oxide surfaces, J Catal, (245), 91-109, 2007. [2] O. Valentin et al, Chemical expansion of La0.8Sr0.2Fe0.7Ga0.3O3-δ, Solid State Ionics, (193), 23-31, 2011. [3] C. Wagner, Equations for transport in solid oxides and sulfides of transition metals, Prog Solid State Ch, (10), 3-16, 1975. [4] H.J.M. Bouwmeester et al, Importance of the surface exchange kinetics as rate limiting step in oxygen permeation through mixed-conducting oxides, Solid States Ionics, (72), 185-194, 1994. [5] S. J. Xu et al, Oxygen permeation rates through ion-conducting perovskite membranes, Chem Eng Sci, (54), 3839-3850, 999. [6] P.M. Geffroy et al, Influence of oxygen surface exchanges on oxygen semi-permeation through La(1-x)SrxFe(1-y)GayO3-δ dense membrane, J Electrochem Soc, (158), 1-9, 2011. [7] S.B. Adler, Chemical expansivity of electrochemical Ceramics, J. Am. Ceram. Soc., (84), 2117-2119, 2004. [8] O. Valentin et al., Thermo-Chemo-Mechanical Modelling of Mixed Conductors in Transient Stages, Adv. Sci. Rech., (65), 232-237, 2010.

International audience; This paper suggests a macroscopic model describing the thermo-chemo-mechanical behaviour of ceramic dense membrane for oxygen separation application. This work takes in account to oxygen permeation and strain induced by stoichiometry variation with working conditions. This model, developed within the traditional framework of phenomenological approach, is based on the assumption of strain partitions and requires only three state variables: oxygen activity, temperature and total strain. Oxygen bulk diffusion and surface exchanges are described thanks to the thermodynamic approach developed by Onsager. While many works focused on semi-permeation induced strain, the proposed model also includes the temperature effect on chemical expansion. Strains predicted by the proposed model are validated thanks to experimental test on La0.8Sr0.2Fe0.7Ga0.3O3-δ. Implemented in F.E.A code Abaqus, this model permits studying the design and the process management effects such as chemical shocks on the membrane reliability.