Graded Membranes for Energy Efficient New Generation Carbon Capture Process

Final Report Summary - GREEN-CC (Graded Membranes for Energy Efficient New Generation Carbon Capture Process)

Executive Summary:
The aim of the project was to develop membranes and a PoC-module for an integrated oxygen generation in cement production and for Oxyfuel- and IGCC fossil power plants. As these industries have a range of different requirements for incorporation of a membrane module into their specific process environment, one of the first steps was to establish a list of boundary conditions (e.g. Temperature, Pressure, Gasses, Dust, Durability, Maintenance etc.), which serves as a guideline for the material development and testing of the membranes. It turned out that the amount of harmful elements and dust is very different in the different applications, so that dependent on the membrane stability process modifications in the flue gas cleaning may be required. In addition, process simulations for an optimal integration of the membranes were performed. The calculations allow a comparison of the newly designed processes with existing processes in terms of net-efficiency and investment and operating costs for selected cases. The IGCC-case was identified to be inefficient due to additional technology required independent of membrane performance and cost.
Four design concepts for the asymmetric planar pilot Proof-of-Concept-module were investigated by numerical modeling (Computational Fluid Dynamics and FEM stress analysis). Advantages and disadvantages of each concept were evaluated, and the most promising concept was selected at month 24 (MS 3). Also our Australian partner, the University of Queensland, contributed to these modeling activities. Two tape-cast asymmetric membranes and an interlayer providing stability elements have to be laminated, sintered, and sealed into the metallic housing.
A wide range of membrane materials were identified and manufactured at lab scale, aiming at high performance and stability. Consequently, permeation experiments and stability tests in CO2 and SOx were carried out. La0,58Sr0,4Co0,2Fe0,8O3-δ (LSCF) was decided to be scaled up for the PoC-module as a reference material because it is stable in CO2 and are successfully sealed to metallic housing materials using Ag/CuO-based reactive air brazing. However, stability towards SOx is a critical issue. In the last period membrane components are developed in 7x10 cm size for the PoC module. The module with 420 cm2 membrane area was built, sealed and assembled in the last project period. Operation of module started due to some problems with sealing delayed in M50, i.e. demonstration of the proof of concept operation for 1000 hours in application relevant conditions. The stability of the membrane was investigated. Unfortunately due to some leakage the flux through the membranes could not be measured so that Milestone 9 could not been fullfilled. This work will be continued after the project duration.
In addition, a lot of dual phase composites were investigated, because of a high stability in CO2 and SOx. In the recent period one favorite composition (Ce0,8Gd0.2O2-δ - FeCo2O4) including catalyst was selected from this research which is proven to be stable in the targeted conditions. To increase the performance of the membranes several catalytic materials were tested and selected for porous surface layers on top of the membranes. Also for these materials stability tests in CO2 and SOx containing atmospheres were carried out and the electrochemical behavior was measured.
Most of deliverables and milestones were fulfilled according to the work plan in Annex I or with small delays. Operation time of 1000 h (measured 650 hours) of the PoC module under realistic atmosphere was demonstrated in the last period of the project. Only the measurement of the 1% aging rate could not be measured due to leakage in the sealing of PoC module. The work will be continued after finalization of project period. The interaction between the work packages and partners is very effective. The project has a high potential to achieve its goals developing a membrane, stable at operation conditions defined by the industrial partners.

Project Context and Objectives:
The objective of the project is the development of a new generation high-efficiency capture process based on oxyfuel combustion utilizing novel oxygen transport membranes.
All scenarios about the future global energy requirements anticipate an increasing demand for electricity, which in 2030 is predicted to be twice the current demand. Increasing global population and fast growing economies especially in developing countries are identified as main drivers for this development. These scenarios also indicate that this increase can only be met by considerable use of the fossil fuels coal, natural gas and oil. It becomes clear that fossil fuels will contribute to more than 60 % of future electricity generation. However, fossil fuel power plants are by far the biggest point sources of CO2 and contribute more than 40% of the worldwide anthropogenic CO2 emissions. Carbon Capture and Utilization (CCU) or Storage (CCS) in fossil power plants is an important strategy to reduce CO2 emissions in order to counteract global warming, offering a key contribution to allow future electricity supply to be environmentally safe and sustainable.
Major sources for human made CO2 emissions comprise the energy and the industrial sector including cement, glass and steel production. One of the most appropriate concepts to capture CO2 from such point sources is the oxyfuel combustion. The main energy demand for this method results from the O2 generation, which is usually done by air liquefaction. This energy demand can substantially be lowered using thermally integrated separation modules based on ceramic oxygen transport membranes (OTM). It is least if the OTM is integrated in a 4-end mode, which entails that the permeating oxygen is swept and directly diluted using recirculated flue gas. Up to 60% reduction in capture energy demand compared to cryogenic air separation and up to 40% reduction compared to post-combustion capture approaches can be achieved. GREEN-CC provides a new generation high-efficiency capture process based on oxyfuel combustion. The focus lies on the development of clear integration approaches for OTM-modules in power plants and cement industry considering minimum energy penalty related to common CO2 capture and integration in existing plants with minimum capital investment. This is attained by using advanced process simulations and cost calculations. GREEN-CC also explores the use of OTM-based oxyfuel combustion in different energy-demanding industrial processes, e.g. oil refining and petrochemical industry.
Oxygen transport membranes consist of gastight Mixed Ionic Electronic Conductors (MIEC), which allow oxygen diffusion through vacancies in the crystal lattice and simultaneous transport of electrons in the opposite direction, thus obviating the need for an external electrical short circuit. Their major advantage is infinite oxygen selectivity, providing that leakage through the membrane layer or the sealing is prevented, resulting in high purity oxygen. However, highly permeable membrane materials show a chemical instability against CO2 and other flue gas components. One major challenge faced by GREEN-CC was therefore to identify and develop membrane materials, components, and a Proof-of-Concept-module for the 4-end mode OTM integration. The desired membrane assembly consists of a thin membrane layer supported on substrates with engineered porosity and oxygen reduction catalysts with high and stable activity in flue gas. As proof of concept, a planar membrane module were developed which opens up challenging hurdles like joining technology. Modelling at all levels supported the development ranging from gas transport modelling through the membrane assembly, finite element modelling of the Proof-of-Concept- module as well as process simulations of the targeted applications, i.e. power plant and cement industry. Specific attention was paid to cost efficiency in terms of capital and operational costs.

Project Results:
The following chapter describes the major results in the project. To structure the topics the results are discussed for the single work packages and the exploitable foreground. The Attached version of the Final GREEN-CC Report is identical to Deliverable 7.21 Final Report.

Work Package 1
WP1.1 Material development/selection
The main objective of the work conducted in WP1.1 was focused towards development and further optimization of membrane materials with appropriate oxygen flux and stability under flue gas conditions as defined by WP5. Since membranes with a mixed ionic electronic conductivity (MIEC) are required a two-route approach was chosen from the beginning, i.e. single phase as well as dual phase MIEC materials.
Route A: Single-phase perovskites are known to have the highest permeability, but limited stability. Therefore, two candidates with good performance and expected stability against CO2 (as minimum stability requirement) were chosen to be analyzed in detail with regard to the specific needs in the project (in particular stability against SOx). La0.6Sr0.4Co0.2Fe0.8O3-δ (LSCF) is a well-established cathode material in SOFC and already investigated as membrane material showing good performance and stability in CO2. In addition La1-xCaxCo1-yFeyO3-δ (LCCF) was considered because the substitution of Sr with Ca should lead to improved stability against acid gases in the flue gas such as CO2 and SOx. However, it turned out that the calcium did not sufficiently dissolve in the perovskite lattice so that the stability was even decreased. Therefore, this material was discontinued.
Reasonable stability of LSCF in CO2 could be confirmed, but relatively low levels of SO2 lead to the formation of SrSO4, which blocks oxygen permeation. Nevertheless, LSCF was agreed to be continued as reference material because it shows reasonably high fluxes and based on the experience with this material scale up to real components is expected to be possible. Moreover, a lot of knowledge exists about engineering properties of LSCF such as thermal expansion, mechanical strength etc. which was required by WP4 in order to engineer the PoC-module at a relatively early stage of the project.
Route B: Major attention was focused towards development of dual-phase membrane materials, in which mixed conductivity is facilitated through a network of two-co-existing and percolating phases in a composite. Both zirconia-based and ceria-based dual phase membrane materials were investigated.
Experiments were conducted to test the thermochemical stability of dual-phase membrane materials under conditions of CO2 and SO2 exposure. Most dual phase composites showed very good stability. Additional criteria for material selection include permeance, sinterability, thermal expansion, technical issues with regard to the development of asymmetric membranes in WP1.2 as well as availability and costs of raw materials. In summary, a total of 17 dual-phase compositions were investigated. Research on 2 of these compositions was continued. A third reasonable candidate was chosen as backbone material for surface exchange activation layers in WP2.
The final decision on the route B-material in close cooperation with WP1.2 was the composite Ce0.8Gd0.2O2-δ – FeCo2O4 (CGO-FCO) with optimized composition, i.e. 85 wt% CGO and 15 wt% FCO. The zirconia based composite (Y2O3)0.01(Sc2O3)0.10(ZrO2)0.89– MnCo2O4 (ScYSZ – MCO) is also a good candidate and, hence, was chosen as fall back option. The major reason for scale up of CGO-FCO was that the partner FZJ responsible for scale up had already gained experience with this material. As shown in D1.4 the composite Ce0.8Gd0.2O2-δ – FeCo2O4 (CGO-FCO) shows transport limitations by the ionic conductivity of CGO. Therefore, a study was started increasing the ceria content resulting in a maximum permeance of the composite using 85 wt% CGO and 15 wt% FCO (vol%-ratio: 81.5:18.5).
At sintering temperature of 1200 °C phase transformations occur, i.e. irreversible formation of a donor doped GdFeO3 perovskite and reversible formation of cobaltowurzite. This facilitates densification of the composite. Moreover, it could be proven that they contribute to oxygen permeation. Therefore, a novel synthesis route was developed, i.e. mixing of commercially available CGO with Fe-oxide and Co-oxide powders. The formation of the intended phases (FCO spinel and GCFCO perovskite) during the sintering process was confirmed by XRD. The permeation rate is slightly below that of the initially synthesized powder most likely due to differences in microstructure showing that there is potential of improving this further. Nevertheless, since the manufacturing is much more scalable and cost effective this novel route was decided to be followed.
Stability tests with varying high amounts of CO2 and SO2 were performed with the 60CGO-FCO and 85CGO-FCO in WP3. No carbonate and/or sulphate formation could be detected using different analysis techniques. Only the monoxide phase with rock salt structure was detected as observed during sintering. However, this phase occurs due to the low pO2 in the annealing gas and does not result from the presence of CO2 or SO2. Since in the application always a minimum content of approx. 3-5% O2 is present in the flue gas, this is not critical. The details of stability testing are reported in WP3.
WP1.2 Membrane development
La0.6Sr0.4Co0.2Fe0.8O3-δ (LSCF) was chosen as reference material. In order to establish a robust and reproducible manufacturing process of asymmetric membranes a fundamental study on tape casting slurries was performed. All commercially available LSCF powders are synthesized aiming in a high specific surface area (for the use as SOFC cathode). However, it turned out that such powders are not very suitable for tape casting. Ultimately, the company Solvay Flour GmbH, Hannover, Germany, was willing to synthesize customized powder within an internal R&D project. An adapted slurry preparation route and slurry recipe for tape casting was developed, which made reproducible processing and scale up of membrane components possible (Deliverable 1.3). In the course of the project, in particular for the scale up in WP1.5 three batches (in total 50 kg) LSCF powder were purchased.
Oxygen flux of the thin supported membrane is about 10 times higher compared to the disc in an air/Ar-gradient. At the targeted operation temperature of 850 °C the thin membrane shows a flux of approx. 1 ml min-1 cm-2 and, thus, is in the targeted range. Applying a porous catalytic layer the permeation rate is expected to increase significantly.
The manufacturing process was further developed with regard to reproducibility and for delivery of samples to the partners. A batch of 150 small disc-shaped asymmetric samples was prepared and all of the delivered membranes were tested in advance to be sufficiently gas-tight. The microstructure of the newly prepared membrane batch was analysed via SEM and compared to the reference reported in D1.3. The porosity of the support layer as well as the thickness of the membranes are comparable validating the reproducibility of the preparation method. Also the oxygen flux is very comparable to the first developments made in D1.3. The further developed membrane shows a slightly improved performance. When a porous surface activation layer made of LSCF is applied the flux further increases as expected. At 850 °C the flux is now approx. 1.4 ml cm.2 min-1 in the given conditions (air-Feed vs Ar-Sweep).
For route B CGO-FCO was selected for scale up. In case of optimized 85wt%CGO – 15wt%FCO (85CGO-FCO) the asymmetric membrane was successful using commercially available CGO as well as Co- and Fe-oxide powders and the reactive sintering process described in WP1.1. On top of the membrane a porous LSCF layer was coated to accelerate surface exchange. Permeation shows good performance reaching 1 ml cm-2 min-1 @1000 °C and reveals that surface exchange limitations at the membrane-support interface limit the transport concluded from the drastically increased apparent activation energy. The activation energy of the both-sided activated bulk sample is equal to that of the ionic conductivity of CGO clearly indicating that this is the rate limiting step in the overall permeation process. In general, surface exchange kinetics and concentration polarization (gas phase diffusion) show higher and lower activation energies, respectively. Therefore, an asymmetric CGO-FCO membrane was surface activated by WP2 leading to increased performance. There is still some room for improvement indicated by the activation energy, which is now much lower, but still slightly above the bulk membrane. But, for the current status of development this is considered as very good result.
WP1.3 Mechanical characterisation
The work task concentrated on the assessment of the mechanical properties at room and application relevant temperatures for materials selected within the framework of the project. Tests were based on properties that are considered to be important for the application, i.e. elastic modulus, fracture strength and creep behaviour, where materials availability and the dynamic development of the material was one of the boundary conditions that limited property determination for example for the dual phase material. Overall, compared to other membrane materials discussed in literature the studied material appear to be suitable for further development towards application, where low Weibull moduli and enhanced creep behaviour of porous substrate materials are aspects that require further materials and microstructural optimization. The work task was finalized and comprehensively reported in D1.9.
WP1.4 Transport modelling
The objective of the task 1.4 was to evaluate the mass transfer resistance of oxygen in asymmetric membranes by means of transport models that describe the oxygen transport in the gas phase and the solid membrane phase. Two different approaches have been chosen to achieve this goal: The first approach considers the porous support of the membrane as a homogeneous (effective) medium. Furthermore, the oxygen concentrations in the feed and sweep gas are assumed to be spatially homogeneous. This enables a facile evaluation of the overall mass transfer resistance of the support layer and a straightforward comparison to experimental data. The second approach includes a detailed model description of the support microstructure and, by employing fluid dynamics model, takes into account the oxygen distribution in the feed and sweep gases, as well as in the pores of the membrane support. While more elaborate, this approach allows the calculation of the spatial distribution of oxygen in the membrane as well as in the feed and sweep. Regions in the membrane support structure which actively contribute to the oxygen transport can be identified. Furthermore, the fluid dynamics modeling reveals any concentration polarization in the feed and sweep gas, if present.
Both model approaches are supplemented by experimental studies to investigate the oxygen transport predominantly in the membrane support. They provide complementary information that will assist in optimizing the microstructure and geometry of the membrane support.
WP1.5 Scale up of asymmetric membranes
WP 4 defined requirements for a planar membrane component comprising of two membrane layers and supporting elements in order to withstand the foreseen outer pressure of 5 to 10 bar. The resulting voids between the supporting elements are acting as flow channels for a better gas transfer in the membrane component.
Although scale up was planned in M24-M36 according to the DoW, investigation was started already in the first reporting period based on the very encouraging results of WP1.2. In a training activity between JÜLICH and RSE, Francesca Drago from RSE visited Forschungszentrum Jülich for 4 month (Feb-May 2015) closely cooperating with Dr. Falk Schulze-Küppers at JÜLICH/IEK-1. A first scale up step using LSCF was successful according to the design made in WP4. It was manufactured using two asymmetric membranes and an intermediate layer in which the channels are machined. These three layers are laminated and co-sintered, so that the thin membrane layers are on the outside and the porous supporting layers are inside providing gas flow channels. Numerous technical problems could be solved in this development particularly related to sintering. Due to the high amount of organic additives and pore formers inside a relatively dense structure, the debinding became critical. Therefore, an optimized combined debinding/sintering cycle could successfully be developed.
However, in the further scale up process additional serious problems came up. It turned out that the weight of the full size membrane element was too high. Since the whole component (approx. 100 g in full size) linearly shrinks about 20-25%, scratches occurred due to the intensified contact with the sintering plate damaging the thin membrane layer. Several attempts using different sintering strategies failed. Therefore, eventually the whole process chain was changed to a half-cell concept. This means that first one asymmetric membrane is laminated with an intermediate layer of half thickness. The laminate (half-cell) is subsequently sintered lying on the channeled interlayer. This results in fewer scratches due to approx. half weight. Moreover, the smaller scratches are no longer in the membrane layer but the channeled part, which is porous anyways. Afterwards two of this sintered half cells are garnished together, i.e. second sintering step with a roll coated paste of LSCF as “ceramic glue”. This resulted in very good mechanical stability after sintering, i.e. the components are not breaking easily during any handling. Finally, porous surface activation layers are printed and calcined on top of both membrane layers. The entire process chain is comprehensively described in D1.10.
A positive side effect is that the machining of the channels in the (half-thickness) interlayers can be done by punching. This is very cost effective in mass production compared to other methods such as milling, which was used for the thick interlayers. Therefore, the novel concept using half-cells solved the problems so that the required three full-size components could be delivered for the module test to WP4 although this was highly delayed.
In case of the route B material, i.e. Ce0.8Gd0.2O2-δ – FeCo2O4 (CGO-FCO), the debinding problem again occurred. It turned out that the FCO apparently catalyses ignition of the organic compounds at least as effective as LSCF or even worse. Therefore, development with a CGO-FCO membrane layer on a porous structural ceramic, i.e. partially stabilized zirconia, was started at DTU. The manufacturing of an approx. 10x10 cm2 dense membrane layer on a 3YSZ support was successful. However, the thickness of the support was only 200 µm, which is too low for the sandwich-structured components required by WP4. Moreover, the thermal expansion behaviour is quite different compared to LSCF. Therefore, a complete new module would have to be built up using different housing and sealing materials. Although this is favourable in terms of cost efficiency, it could not be realised in the given frame of time and resources within GREEN-CC.
Nevertheless, this concept is very promising because it serves several advantages. Structural ceramics, in particular partially stabilized zirconia, provide highest strength and fracture toughness among ceramic materials. Moreover, it is cheaper compared to functional ceramics such as LSCF. Also the choice of housing materials (metals) widens up because the coefficient of thermal expansion now fits to ferritic steels and it is no longer necessary to rely on (expensive) nickel based alloys. The desired multi-layer assembly will consist of a structural support, a fine porous interlayer for surface activation, the dense membrane layer, and another activation layer on top. The surface activation is researched in WP2 and a lot of progress was done there in terms of materials development as well as concepts for technical realisation. However, in particular the choice of the catalysts is very dependent on the targeted application and, hence, the surface activation will remain a technical issue.

Work Package 2
Within GREEN-CC project, WP2 is focused on the development of catalysts for the activation of Oxygen Transport Membranes (OTMs) operating under oxyfuel conditions and the further optimization for achieving a high performance level with a reliable and stable behaviour. The consecution of this global objective was achieved by conducting actions on:
(1) Development and optimization of highly effective and highly stable catalytic materials for the oxygen reduction/oxidation reaction in OTMs exposed to CO2-SO2 rich gas environments during long periods.
(2) Development of methods for the efficient activation of both flat membrane surfaces by diverse coating techniques and porous substrates by infiltration/impregnation procedures by using the optimized catalyst materials.
The proposed methodology consisted of two main approaches consisting of theoretical (DFT analysis) and experimental (EIS, PIE, ECR and oxygen permeation) techniques.
Large scale first-principles calculations of the basic properties of (La,Sr)(Co,Fe)O3 (LSCF), serving as a reference membrane material, were performed for its different compositions and crystalline structures, both bulk and surfaces. The very accurate calculations of defect properties were achieved using the state-of-the-art hybrid exchange-correlation functionals within the Density Functional Theory (DFT). The effective atomic charges, magnetic moments on Fe, Co, the structure relaxation, total and one-electron energies of defective supercells were carefully analyzed. Sr dopant greatly improves permeation properties of LSCF due to decrease of the oxygen vacancy formation energy and thus increase of their concentration. These vacancies are shown also to stabilize cubic phases of both LSCF and BSCF with respect to transformation into hexagonal or rhombohedral phases, respectively. Modeling of the interaction of the LSF (100) surface with water and CO2 in a flue gas have shown visible proton uptake and chemisorption of CO2 molecules. A study of Ag doping as a model catalyst on LaMnO3 surface has shown an increase of the O2 adsorption energies which could improve oxygen reduction efficiency. Moreover, The basic properties of pure ceria were calculated using the state of the art first principles method. It was shown that the hybrid PBE0 (HSE06) exchange correlation functionals reproduce experimental data for a pure ceria very well. A novel approach for the defective crystal calculations in the supercell model was suggested and applied. The novel approach was also used for the calculations of Tb impurities in CeO2 with oxygen vacancies. It leads to 2 orbits for the position of Tb in CeO2 and two orbits for the position of oxygen vacancy in the 96 atoms supercell. The calculations of Tb impurities were done within the DFT+U approach. All the positions both the Tb impurity and oxygen vacancy were tried in combination with different distances between Tb ion and oxygen vacancy. The oxygen vacancies reduce not only the Tb impurity but also the Ce ions into 3+ state, reducing6 the formation energy down to 2.5- 1.2 eV. Thus, Tb doping strongly fuels oxygen transport in ceria.
For confirming the DFT findings on O2-CO2 adsorption and advancing in the establishment of a model describing the oxygen permeation mixed electronic ionic conducting materials, it was performed a broad study on LSF55 membranes. This study consisted of establishing a theoretical model allowing to predict the permeation behaviour of LSF membranes under different conditions of temperature, pO2 and CO2 content. Therefore, 96 different conditions were tested, thus obtaining a big amount of data for model fitting adjustment. Different models describing the oxygen permeation and the interactions of O2 and CO2 molecules on membrane surfaces were considered. Observing from the experimental results that CO2 presence produces a decay in oxygen permeation, it was considered that the reaction of CO2 with the O2 vacancies causes a decrease of the O2 vacancy concentration in the sweep surface and, therefore a decrease of the oxygen flux across the membrane because the driving force decreases. Therefore, a model was built aiming to predict this decrease of the oxygen flux as the CO2 partial pressure of the sweep gas increases. The adjustment of experimental data obtained from LSF55 permeation studies allowed the fitting to some considered models. With this model it is possible to predict J(O2) values in dependence of temperature, pO2 gradients and CO2 concentration on sweep side. Moreover, with this study the adsorption constants and enthalpies of O2 and CO2 were also determined, thus complementing DFT findings.
An optimization study on porous catalytic layers formulation for the activation of OTMs was conducted. A preliminary evaluation on several materials was conducted in order to select those showing a stable behaviour under oxyfuel conditions. Firstly, materials potentially stable under environments with presence of CO2 and SO2 were analyzed. Nearly 100 compounds materials (and composites) were identified as potential candidates for operation in those conditions. The selection was done on the basis of activity towards O2 oxidation/reduction and stability against carbonation and sulfation. The materials were crystalline based on the following structures: fluorites, perovskites, swedenborgites, cromites, titanates as well as dual-phase composites thereof. The screening revealed that 36 out of 43 finally tested materials were stable under the considered test conditions. Knowing from other partners findings (see Deliverable D1.1) that Sr-containing materials lead to SrSO4 formation when exposed to higher SO2 concentrations (400 ppm or more) then the materials with Sr in their compositions were discarded. From the remaining stable materials, the interest was focused on the composites, due to their interesting mixed ionic-electronic properties. The material with composition 60NFO-40CTO was finally selected as reference material for conducting the optimization catalytic study, mainly for its well-known better MIEC properties than the rest of the candidates. Therefore, O2&SO2 stables material based on Fe2NiO4-Ce0.8Tb0.2O2-d (NFO-CTO) composite were selected as reference material for acting as porous backbone. Then, a complete electrochemical characterization consisting of electrochemical impedance spectroscopy (EIS), electrochemical conductivity relaxation (ECR) and O18/O16 pulse isotopic exchange (PIE) studies was carried out in order to (a) identify the best catalysts for the promotion of oxygen oxidation reduction reactions on infiltrated NFO-CTO porous catalytic layers; and (b) achieve further understanding of the most important parameter of the exchange process and about the nature of the elemental mechanisms. Catalytic activation of membranes was done by means of screen printing technique, thus obtaining homogeneously porous and well-attached NFO-CTO composite layers. A high porosity degree was achieved, resulting in a proper infiltration of the catalysts. The activation of the catalytic layers was conducted by capillary infiltration of solutions containing the catalysts precursors. Several elements such as lanthanides (Ce, Pr, Tb, Sm), metals (Zr, Nb, Mo, Co, Al) were selected for performing the screening study. Single, binary and ternary combinations of catalysts were considered for the conduction of the experiments, thus determining the best formulations for the activation of OTMs subjected to oxyfuel-like environments. EIS tests performed on single catalysts revealed lower polarization resistance values associated to surface processes for Ce and Pr under all the tested conditions, especially when inducing harsh oxyfuel environments. After a first screening, binary combinations of the catalysts were tested resulting in an outstanding performance for a catalyst based on Ce-Pr formulation. These finding were confirmed by PIE tests, measuring high surface exchange rates for Ce-Pr and Pr-Al catalysts. And finally, optimal polarization resistance values were obtained with ternary catalysts consisting of Ce:Pr:Zr combination.
In addition to the reference material 60NFO-40CTO, other dual-phase materials were also considered for oxyfuel applications, thus studying their electrochemical performance under CO2 and SO2-containing environments. In a joint investigation with DTU and WP3, several formulations of dual phase materials with 8YSZ or ScSZ as ionic phase and Mn2CoO4 as electronic phase were characterized. Other dual-phase compositions considered consisted of Y0.8Ca0.2Cr0.8Co0.2O3-x-Ce0.9Gd0.1O2-x (50:50) material. These formulations are some of the reference compositions used by DTU in the activities conducted throughout WP3. These materials were characterized for being used as catalytic layer when activated with Ce-Pr catalyst. The obtained results showed a very good performance for some cases, especially for the ScSZ-MCO and the YCCC-CGO compositions, being the first also stable under oxyfuel environments.
Finally, and according to the results obtained in the EIS and PIE measurements, several catalysts were selected for activating 60NFO-40CTO membranes and thus study the oxygen permeation performance under representative oxyfuel conditions. Selected elements were: Ce, Pr, Al, Pr-Al and Ce-Pr. The results showed a significant improvement when adding a 60NFO-40CTO catalytic layer, obtaining even better results when including active catalysts as resulting from EIS and PIE findings. 6-fold improvement in oxygen permeation was obtained under clean conditions (Air/Ar gradient) for a Ce-Pr activated membrane, whereas oxygen flux was improved up to 2.6 times with the same catalyst under harsh CO2&SO2-containing atmospheres for periods of 24 hours at 850 ºC.
Other studies conducted on oxygen permeation optimization consisted in catalytic layer thickness variation and on the consideration of other dual-phase compositions. . For that, two catalytic layers were considered: a 15 micron-thick and a 30 microns thick 60NFO-40CTO backbone layers, both infiltrated with Ce-Pr catalyst and both deposited on bulk dense 60NFO-40CTO membranes. Higher oxygen flux values were obtained with the thicker layer for all the temperature range under Air/Ar gradient. This can be ascribed to the fact that thicker layer would present more active sites and TPBs for oxygen reactions, resulting in a higher surface active area. The other tested dual-phase options were 40NFO-60YSZ and 40NFO-60ScSZ membranes, which were activated with layers of the same composition and infiltrated with Pr-Al catalyst. From the obtained results, it was found that Sc doping favours significantly the oxygen permeation with respect to YSZ ionic phase. 40NFO-60ScSZ activated membrane showed a performance similar to 60NFO-40CTO membranes.
Finally, optimized catalytic layers consisting of 60NFO-40CTO porous backbone activated with Ce-Pr catalysts were deposited on asymmetric membranes for the conduction of permeation studies, reaching an oxygen permeation of 3.1 ml/min·cm2 at 875 ºC and 50% CO2 with a freeze-casted LSCF membrane. Further tests were conducted on an 60NFO-40CTO+Ce-Pr activated asymmetric FCO-CGO membranes, obtaining lower oxygen permeation rates, in the range of 0.5-0.3 ml/min·cm2 at 850 ºC under CO2 containing atmospheres.
Therefore, it can be concluded that WP2 objectives were fully accomplished since CO2 and SO2 stable catalytic layers were performed resulting in significant improvements in the oxygen permeation of OTMs subjected to harsh oxyfuel environments. After GREEN-CC project and as a result of WP2 activities, it can be stated that new and functional materials leading to a satisfactory implementation of ceramic membranes in oxyfuel processes have been obtained, thus improving the existing state-of-the-art materials before GREEN-CC project.

Work Package 3
The main objectives of WP3 was to investigate the stability of powders and materials suggested in WP1 and WP2 under conditions expected to be encountered in the flue gas of a power plant. Information on the stability was delivered back to WP1 and WP2 for refinement/improvement of the compositions. The stability studies of powders are carried out mainly in WP3.1 “Thermochemical Stability”. The second main objective of WP3 is to carry out permeation tests to establish the performance of the developed components under conditions resembling those expected in a power plant. The permeation tests are carried out mainly in WP3.2 “Permeation Test”. Additionally, methods on how to re-establish performance of these membranes after exposure to simulated flue gas were established. Several material combinations showed sufficient stability in the expected flue gas environment, and it was possible to re-establish the membrane performance after exposure to CO2 and SO2. All milestones and targets related to WP3 were met.
WT3.1 Thermochemical Stability
The activities in WT 3.1 focused on the thermochemical stability testing of several materials, which potentially could be used in the manufacturing of oxygen transport membranes (OTM) 4-end modules developed by other partners of the project. These materials have been subjected to a series of tests at high temperature with an atmosphere containing CO2, SO2 and H2O, simulating the operating conditions that the membrane will need to withstand in real processes. The tests have been performed at temperatures between 850 and 900 °C with gas compositions containing up to 2000 ppm of SO2 (on a dry basis) in CO2 and moisture contents up to 30% vol. The materials have been characterized before and after the treatments with the aim of determining possible changes that could have undergone during the process, revealing low resistance in the operation conditions of OTM modules. With this objective, a complete set of techniques has been used, including X-ray diffraction (XRD), Fourier transformed infrared spectroscopy with attenuated total reflectance (ATR-FTIR), Raman spectroscopy, X-ray fluorescence spectroscopy (XRF), elemental analysis and scanning electron microscopy with energy X-ray dispersive or wavelength-dispersive spectroscopy (SEM-EDS or SEM-WDS).
A total of 19 materials (10 membrane and/or support materials, 8 catalysts and 1 asymmetric membrane) supplied by DTU, FZJ and ITQ-CSIC have been received in different shapes (powders, bars or discs). From the materials studied, single-phase LSCF perovskite (La0.6Sr0.4Co0.2Fe0.8O3) has been chosen as the reference material since it is a good MIEC in terms of oxygen transport properties but it presents low resistance against CO2 and SO2, as it has been found in this project and in previous studies. Two other single phase materials were tested: LCCF (La0.6Ca0.4 Co0.8Fe0.2O3) and 3YSZ (Y-doped ZrO2) as structural support material. LCCF has been found to be unstable against both CO2 and SO2 and has been discarded as a potential candidate for its application in OTM. 3YSZ has been found to be stable in all the conditions tested, thus proving to be a good material for this application. The other 7 membrane materials studied were dual phase composites: CGO-FCO (Ce0.8Gd0.2O2 - FeCo2O4) with two different CGO/FCO ratios, YSZ-MCO (Sc/Y-doped ZrO2 - MnCo2O4), YSZ-AZO (Sc/Y-doped ZrO2 - Zn0.98Al0.02O1.01) YSZ-LCM (Sc/Y-doped ZrO2 - La1.0Cr0.9Mg0.1O3) YSZ-LCCN (Sc/Y-doped ZrO2 - LaCr0.85Cu0.01Ni0.05O3) and NFO-CTO (Ni2FeO4 - Ce0.8Tb0.2O2 (60%/40%). From these materials, CGO-FCO and YSZ-MCO have been found to be unstable in low pO2 atmospheres (e.g. in pure CO2) since the reducing conditions provoke the segregation of Co from the electronic conductors. Nevertheless, they are stable in highly CO2 containing atmospheres. For this reason, these materials are suitable candidates for this application from thermochemical stability viewpoint in flue gas environments where minimally 3-5 % of O2 is expected to be present. The other materials (YSZ-AZO, YSZ-LCM, YSZ-LCCN and NFO-CTO) have been found to be completely stable in all the conditions studied, conserving their structures after the treatment. These four materials are also suitable candidates for their use in OTM modules.
After assessing the stability of the substrates, the stability of a series of potential catalysts for the O2 dissociation and recombination has been evaluated. The catalysts have been provided by ITQ-CSIC and the NFO-CTO (also prepared by ITQ-CSIC) has been selected as catalyst support since it has been found to be stable in all the conditions studied. The catalysts tested have been composed of mono-, bi- and trimetallic combinations of CeO2, PrO2, ZrO2 and Al2O3. It has been found that all the combinations of CeO2, PrO2 and ZrO2 are stable in all the conditions defined in this project. On the contrary, the presence of Al2O3 in the composition of the catalyst has resulted in lower resistance against the operating conditions due to the formation of sulphates when both SO2 and H2O have been present in the gas feed.
Finally, a disc replicating the architecture of an asymmetric membrane was tested under the harshest conditions defined for the project. The disc was composed of a dense membrane of CGO-FCO supported on a porous layer of CGO-FCO and with a screen-printed catalytic porous layer of NFO-CTO infiltrated with CeO2-PrO2 on top of the dense membrane. It was found that the treatment did not affect the structure of the membrane and the layers stayed attached with no signs of incompatibility.
The final conclusion from WT 3.1 is that, from the thermochemical viewpoint, the best candidate materials for their application in 4-end OTM modules are the 3YSZ (Y2O3/ZrO2), the YSZ-AZO (Sc/Y-doped ZrO2 - Zn0.98Al0.02O1.01) the YSZ-LCM (Sc/Y-doped ZrO2 - La1.0Cr0.9Mg0.1O3) the YSZ-LCCN (Sc/Y-doped ZrO2 - LaCr0.85Cu0.01Ni0.05O3) and the NFO-CTO (Ni2FeO4 - Ce0.8Tb0.2O2 (60%/40%) as main membrane materials and the combinations of CeO2, PrO2 and ZrO2 as catalysts. Also CGO-FCO (Ce0.8Gd0.2O2 - FeCo2O4) with a 85:15t CGO/FCO ratio and YSZ-MCO (Sc/Y-doped ZrO2 - MnCo2O4) are suitable candidates in sufficiently oxidizing atmospheres.
WT3.2 Permeation Test
The results on application oriented testing are grouped into two sections. The first section “Permeation testing of first generation materials” describes results obtained for materials developed and tested during the initial phase of the project, but which development was discontinued or not chosen for scaling-up due to various reasons. The part “Permeation testing of second generation materials” describes the development and testing of materials, which successfully passed benchmark-testing in the initial phase and which were scaled-up and tested in more detail in the remaining part of the Green-CC project.
Permeation testing of first generation materials: LSCF (La0.6Sr0.4Co0.2Fe0.8O3):
Asymmetric LSCF membranes, supplied by FZJ as the reference membranes chosen in the Green-CC project, have been tested at high temperature and atmospheric pressure. Permeation tests have been performed in the temperature range of 800 – 950°C by feeding synthetic air or pure oxygen on the membrane side (with a flow rate of 250 and 200 Ncc/min, respectively). The effects of CO2 and SO2 in the permeation process and the membrane stability in CO2-rich atmosphere have been studied by feeding different carbon dioxide concentration in the sweep gas stream. The performed permeation experiments indicate that LSCF Membranes are not suitable for long-term operation in SO2 containing flue gases. The formation of different sulfates on the surface of the membrane inhibits the oxygen diffusion, and already after a short exposure time a drastic decrease in the oxygen flux is observed.
Dual phase membrane:
Different composite membranes were prepared in WP1 by conventional powder mixing and sintering and further tested for CO2 stability in WP3. The selected ionic conductor was 10Sc1YSZ (Sc0.2Y0.02Zr0.78O1.89) and as electronic conductor 5 materials (MnCo2O4, NiFe2O4, NiCr2O4, NiFe1.8Cr0.2 LaCr0.85Cu0.10Ni0.05O3 and Al-doped ZnO (Al0.02Zn0.98O1.01) were tested. Oxygen permeation fluxes of the different composite membranes (1 mm thick) under air/N2 and air/CO2 gradients showed that the best performance was achieved by 10Sc1YSZ-MnCo2O4 (70-30 vol.%) membranes (850°C, 0.21 Nml min-1 cm-2 in N2 and 0.22 Nml min-1 cm-2 in). The membrane composed of the 10Sc1YSZ-LaCr0.85Cu0.10Ni0.05O3 (70-30 vol%) + 3 vol% NiO shows the second highest oxygen permeation. At 950°C, it reaches 0.265 mlN min-1 cm-2 in N2 and 0.246 mlN min-1 cm-2 in CO2. Based on the results above it was decided to continue with these two composites.
Permeation testing of second generation materials:
In the second project phase, two composite materials were selected from the potential candidate materials described in WP1 and the section above. These dual phase membranes are based on i) Ce0.8Gd0.2O2 - FeCo2O4 (CGO-FCO) and (Y2O3)0.01(Sc2O3)0.10(ZrO2)0.89 - MnCo2O4 (ScYSZ-MCO). Their stability towards CO2 and SO2 was evaluated and additionally results on how to re-establish performance of these membranes after exposure to simulated flue gas were reported. Both material combinations showed sufficient stability in the expected flue gas environment, and it was possible to re-establish the membrane performance after exposure to CO2 and SO2.
Conclusion for CGO-FCO based membranes:
Concluding the permeation experiments with CO2 and SO2, there are two processes that can be observed as of now. The first effect is caused by the oxygen partial pressure gradient. Both cobalt and iron diffuse in the FCO phase to the oxygen rich side. Cobalt seems to be faster, resulting in the formation of cobalt enriched spinel crystals on the feed side and an iron enriched porous sweep side surface. This process can be observed with and without SO2 present on the sweep side. The described processes change and specifically increase the exchange surface area on both the feed and sweep side due to a coarsening of the surfaces. This explains the increase of the oxygen flux during the first few days of the permeation experiment, conducted on non-activated, i.e. surface sensitive, specimens. The process is significantly decreased in the second-generation material with a higher CGO content (85 % compared to 60 % in the first-generation material). Based on the long-term testing performed so far, there is no indication that this process could be problematic, but future experiments are required to investigate this diffusion behavior over a longer period investigating the applicability of the selected materials in long term operations. The second process is caused by the presence of SO2. When SO2 is added to the sweep gas the oxygen flux decreases abruptly. This is probably due to the SO2 molecules blocking active oxygen surface exchange sites (“competitive adsorption”), which will occur in any oxide material. A chemical reaction with SO2 would decrease the oxygen flux irreversibly. Moreover, no sulfur containing compounds could be identified in both the EDS and XRD analysis. Therefore, it can be concluded that the CGO as well as FCO are stable in SO2 and CO2.
Conclusion for(Y2O3)0.01(Sc2O3)0.10(ZrO2)0.89 - MnCo2O4 (ScYSZ-MCO) membranes:
ScYSZ-MCO has been selected as the second material for further investigation is the GREEN-CC project and was considered as back-up material in case the CGO-FCO line fails. The membranes were obtained in a two stage sintering process followed by infiltrations of CGO and LNC (LaNi0.6Co0.4O3). Thin dual-phase membranes (7 µm) made of 70 vol.% ScYSZ and 30 vol.% MnCo2O4 were successfully prepared and characterized by thermochemical stability tests and oxygen permeation tests.
The stability of the 7 µm thick asymmetric membranes was studied in conditions relevant for oxy-fuel combustion (CO2, SO2, H2O containing atmospheres). The test was performed at 850 °C for 7 days using a gaseous mixture containing 250 ppm of SO2, 3 vol.% of H2O, and 5 vol.% of O2 balanced with CO2. The analyses of the exposed membrane by XRD, Raman spectroscopy and SEM revealed excellent stability. In addition to this test, a longer stability test (1700 h) was performed at 850 °C using pure CO2 as sweep gas. The test showed initial degradation of the oxygen permeability during 1100 h, after which stable performance was observed during the remaining 600 h of test. The observed degradation was attributed to coarsening of the infiltrated catalyst phase. Oxygen permeation tests were carried out in clean atmosphere (N2), in pure CO2 and in conditions relevant for oxy-fuel combustion. In clean atmosphere, oxygen fluxes-up to 1.41 mLN cm-2 min-1 and 2.23 mLN cm-2 min-1 were obtained at 940 °C in air/N2 and O2/N2 atmospheres, respectively. In pure CO2, the highest oxygen permeation fluxes achieved were 0.81 mLN cm-2 min-1 and 1.91 mLN cm-2 min-1 (940 °C, air/CO2 and O2/CO2, respectively). The study showed that in both N2 and CO2 atmospheres the main rate-limiting factor for the oxygen permeation was the surface exchange kinetics; therefore, further studies were performed to improve the catalytic surface layers on the membrane. Oxygen permeation tests were realized on 0.5 mm thick 10Sc1YSZ-MCO membranes (self-standing membranes) coated with different porous backbones ((i) 8YSZ, (ii) 10Sc1YSZ-MCO (70-30 vol.%) and (iii) CTO-NFO (40-60 vol.%)) and infiltrated with a Ce-Pr containing catalytic solution. The good catalytic performance of the CTO-NFO backbone resulted in an increase of about 50 % of the oxygen permeation flux (0.28 mLN cm-2 min-1, 850 °C, air/Ar) when compared to the membrane coated with the 8YSZ backbone (0.19 mLN cm-2 min-1, 850 °C, air/Ar). Significant influence of SO2 on the oxygen permeability by competitive adsorption was observed (decrease of the oxygen flux of 37 – 57 % depending on the coated backbone material), nevertheless no microstructural degradation was found after SO2 exposure and initial oxygen permeation fluxes could be recovered in most of the cases (depending of the coated backbone material recovery of 65 - 100 % of the initial flux).
Overall, all the tests confirmed, that in terms of stability, ScYSZ-MCO composite is a good candidate for use in OTMs directly integrated in oxy-fuel combustion processes.

Work Package 4
According to the DoW, the objectives of the WP are:
• Design and build a membrane module with a planar stack configuration and an effective area of at least 300 cm2
• Proof the engineered performance of a membrane module capable of operating between 750 and 900oC with an air leak rate lower than 2%.
• Long term (1000 hours) proof-of-concept based on testing under laboratory synthetic flue gas stream.
WP4.1 Module development
Five different basic design concepts able to house asymmetric planar membranes operated in a 4-end mode have been considered. Membrane characteristics and process conditions have been defined according to the input of WP1 and WP5, respectively.
Advantages and disadvantages of each design concept have been evaluated, as discussed in D4.1 and the two most promising concepts (Concept 3 and Concept 5) have been selected for further development. Both design concepts are based on planar stacks of three membrane elements 10x10 cm in size. Each membrane element is made by two La0.6Sr0.4Co0.2Fe0.8O3-d (LSCF6428) membranes and up to 20 porous support elements of the same materials which have a structural function. Membranes have an asymmetric structure e.g. they are made of a 30 mm thick dense layer on a 0.7 mm thick porous support.
The final choice between the two alternatives has been taken by considering both minimization of mechanical stresses of the module components and optimization of the fluid-dynamic regime. A parametric study concerning the fracture probability in the membrane element in dependence of number of porous stability elements and feed pressure has been performed by using a simplified 3-D FEM-model. Optimization of fluid-dynamic regime of the module was carried out by CFD simulations performed using a 3-D FEM method based on Comsol Multiphysics coupled with SolidWorks.
Results indicate that Concept 5 design exhibits the best mechanical behaviour in terms of fracture probability, and a uniform flow distribution, as evidenced by CFD simulations. Based on these results, Concept 5 has been chosen as a final design of membrane module (D 4.2).
In Concept 5 each asymmetric membrane element (LSCF6428) is supported by two metallic frames (Nicrofer 6025HT). Feed and sweep streams are fed perpendicular to the membrane stack and flow parallel to the membrane surface in opposite directions. Supporting elements are placed to increase the mechanical stability of the membrane element. Ceramic-metallic sealing between membrane and metallic frame and metallic-metallic sealing between the metallic frames are required.
WP4.2 Membrane assembly and joining
The objective of this WP is to obtain leak tight membranes sealing under expected operating conditions. The activity has been first addressed to define sealing material requirements for metallic-ceramic and metallic-metallic joints and, thereafter, to develop the brazing process for the full size module.
Sealing material requirements for ceramic (LSCF) – metal (Nicrofer) joints are defined in D4.3 (Joining method specification). As a first step, a literature research allowed to select the following candidates for experimental testing: crystallizing glass-solder, glass-composite solders, silver and reactive air brazing alloys Ag-xCuO (1 < x < 20 mol%) and Ag-0.5Al2O3.
Based on wetting angle tests, reactive air brazing has been selected as a joining method, while Ag-3CuO foil and Ag-xCuO (1< x < 3) paste have been selected as sealant materials.
Gas tightness of ceramic-metal joints was further examined using a bubble test bench in which asymmetric LSCF membrane discs are brazed on top of Nicrofer tubes. Moreover wetting tests were made in order to Investigate the influence of reactive element (Al2O3, CuO) concentration in silver on the microstructure. Finally annealing of reactive air brazed joints at 850 °C in air up to 3500 h followed by bubble tests and microstructure examinations were performed.
As far as concerns Nicrofer-Nicrofer joints, since no infiltration can occur, wetting angle tests are not requested, and braze composition is selected only by considering their joint strength (tensile tests). Three different compositions where considered: Ag1CuO, Ag3CuO and Al2O3. Both as brazed samples and samples after thermal cycling were tested. Based on tests results, Ag-3CuO was selected for brazing Nicrofer-Nicrofer joint.
All the experimental activities leading to the above results have been performed with small size samples (brazing area of 0.85 cm2). Since the brazing total area of the membrane module is much higher, about 270 cm2, a suitable process to obtain a leak tight braze with the actual geometry and size of module should be developed.
The brazing process was developed starting from metallic frames. 20 cm x 14.5 cm in size. Since preliminary tests indicate difficulties for the application of Ag3CuO foils, it was decided to focus on braze pastes. To enable a reproducible thickness and width of applied braze paste two different methods were tested:
• Braze paste application by using stencils made of plastic or stainless foil
• Braze paste application by using a commercial 3D printer that was modified to a syringe extruder
Tests indicated that paste application using stencils is not suitable due to the formation of countersinks, while syringe extrusion showed better results and, therefore, was selected for braze paste application.
For the syringe extrusion of braze paste, a 3D printer kit Vellemann K8200 was assembled and used to print the specially designed syringe holder. A parametric study allowed defining process conditions. The braze paste Ag-3CuO is applied through a nozzle (d = 1.2 mm) on the Nicrofer plate. The braze paste is applied in four layers of 0.2 mm to obtain a braze thickness of 0.8 mm. The binder content is 40 wt. %. Stainless steel foil (t=50 µm) is used as spacer during the brazing process. The brazing width is fixed at 150 µm. A level is used to avoid an accumulation of the braze in a sink.
Further brazing tests with assembly of an half size membrane element in a small scale module were planned; however the late delivery of leak tight membrane prevented to accomplish them. Therefore it was decided that the Ag3CuO paste will be applied by syringe extruded both for the Nicrofer-Nicrofer joints and Nicrofer-LSCF joints.
WP4.3 Module and test facilities design and construction
Difficulties experienced in the membrane element scale-up in WP1 have determined the selection of a different manufacturing route and a change of membrane element dimensions. The width of the membrane element was reduced from 10 cm to 7 cm (see WP1 results in this report), leading to a total membrane area of 420 cm2 , still higher than the value (300 cm2) indicated in the DoW. Consequently, the executive design of the module was revised, as described in D4.7.
The process of assembling the proof-of concept module is described in detail in D4.5. All welding steps were made before brazing, in order to avoid thermal stress for the joined module and to ensure leak-tight welding seams. The three membrane elements were joined to Nicrofer plates and, thereafter, assembled in the module. All components were joined in parallel to avoid re-melting of the braze.
Braze paste Ag-3CuO was applied by syringe extrusion, as described above. Nabertherm furnace with a capacity of 1 m³ was used for the brazing process. Parameters of the brazing program, heating rate and holding time were tested in a trial run. The module was shielded to provide a homogeneous temperature distribution.
The integrity test of the membrane module was made with pressurized air and leak detection spray. Leakage was found between all Nicrofer plates after the first brazing cycle. Brazing was repeated for two more times by increasing pressure; however a leakage at the top plate was still observed. Nevertheless it was decided to install the module in the pilot loop for testing.
The pilot loop has been designed by considering that proof of-concept testing of the membrane module is performed at 850°C by feeding air at 5 bar on the feed side and pure carbon dioxide at 1 bar on the sweep side. The P&ID of the pilot loop is shown in Fig. 3. Sweep gas and feed trains are shown in the upper side and in the lower side, respectively. Process conditions are adjusted by mass flow controllers at the beginning of the gas pipe and pressure controllers after gas cooling down. Electrical preheaters are used to rise both feed and sweep gas temperature up to the operating value before entering the membrane module. The membrane module is housed in an insulated oven in order to keep the temperature constant. An oxygen analyser using a lambda sensor and a mass spectrometer for the leakage detection by N2 quantification are put before sweep gas leaves the pilot loop. Process parameters such as gas concentration, temperature, pressure and flow rate, are measured and recorded continuously, thus allowing to adequately test the proof-of–concept module. Specifications of the components are given in D 4.4. Pre-existing racks are used for placing the components. After construction, testing of the pilot loop was made. The feed and sweep gases were transported without any leakage from the gas supply system to the full-size module and to the measuring components.
WP 4.4 Testing of module performance
The membrane module has been tested in isothermal condition (850°C) by feeding air on the feed side and pure carbon dioxide on the sweep side, for 700 hours. In a first phase, low pressure was applied to the membrane module in order to preserve the membrane elements and the brazed seams from damaging. Pressurized air and CO2 were supplied to the membrane module at 0.5 bar. However a severe leakage of air into the sweep side has been detected since the beginning of tests (almost the 60% of the feed air passed in the permeate). Despite the very high leakage, the feed pressure was increased up to the operating value (5 bar) for a few hours. The air leakage still increased up to the 77% and, thereafter, remained stable, although feed pressure was decreased at the initial value (0.5 bar). The test was then stopped.
The presence of such a severe leakage prevented to obtain a meaningful measurement of the O2 permeation. As a consequence, objectives of tests (leakage less than 2%, continuous operation for 1000 hours and production of 50 NL/h of O2) could not be met.
Leakage could occur both through the membrane elements and through the brazed seams. Brazed seams along gas inlets and gas outlets have been evaluated using an endoscope. Optical evaluation with the endoscope showed that the quality of the brazing is satisfactory. However, about a half of the brazed area and the condition of the membrane elements cannot be evaluated non-destructively. Moreover during examination with the endoscope, a piece of a membrane was found at the sweep inlet. The piece origins from the edge of the membrane, traces of braze are slightly visible, The membrane is not wetted sufficiently by the braze. Possible reasons are too wide brazing gap or lower amount of available braze due to drop formation.

Work Package 5
WP 5, process engineering, was focused on the development of integration approaches for OTM-modules in power plants (conventional and IGCC) and cement production plants considering minimum energy penalty related to common CO2 capture and integration with minimum financial investment. In this context, the WP5 partners achieved results which will be individually described in the following.
WP 5.1 Process simulations
Simulation of an oxyfuel IGCC: The feasibility and the potential of oxygen transport membrane (OTM) application in an oxyfuel integrated gasification combined cycle (IGCC) power plant for CO2 capture has been assessed. Therefore, an IGCC plant of ELCOGAS with an electricity output of 287 MW and a plant efficiency of 39.5 % (HHV based) was simulated. The results were comparable to plant data from ELCOGAS.
The main alteration from the common IGCC to the oxyfuel process is that the fuel is combusted with pure oxygen and all moderation and carrier gases are substituted with CO2 or steam to produce a concentrated CO2 stream for underground storage (CCS) or enhanced oil recovery (EOR). The most optimal design was: use of steam as sweep gas with subsequent condensation, 10 bar feed pressure, 1 bar sweep pressure and direct combustion of natural gas in the air stream. Including CO2 compression, a plant efficiency of 26.9 % (HHV based) was achieved. A membrane model was proposed for a SrCo0.8Fe0.2O3-δ oxygen transport membrane. For the production of 135,000 m3/h pure oxygen this lead to an estimated membrane area of 400,000 m2.
Simulation of a pre-combustion IGCC: Since the oxyfuel IGCC plant did not offer advantages, the following work was concentrated on the application of OTM in an IGCC plant with pre-combustion carbon capture. The main modification is the installation of a water-gas-shift (WGS) step which allows a removal of the CO2 for the carbon capture before the combustion in the gas turbine.
As before, the IGCC plant of ELCOGAS located in Puertollano was used as reference plant for the OTM integration (net efficiency of 39.5% HHV). Three concepts of pre-combustion IGCC were investigated: two with full OTM use (concept A and C) and one with partial OTM and partial cryogenic distillation (CD) use (concept B). In the energetic evaluation, concept C showed the highest efficiency of 31.4% (HHV) with a net output of 228 MW. The main reasons for the efficiency loss were: the CO2 compression cycle, the steam used for the WGS and the loss of hydrogen in the CO2/H2 separation step. Concept C were optimal at 10 bar feed pressure, ambient air as sweep gas at 10 kg/s and 1 bar sweep pressure. For this configuration, a total membrane area of 119,000 m2 is necessary using the SrCo0.8Fe0.2O3-δ calculation model.
Simulation of an oxyfuel power plant: The process layout was adapted from the study “reference power plant North Rhine-Westphalia” (1210 MW thermal power, steam parameters of 600°C/285 bar and 620°C/60 bar reheated, net efficiency of 45.9 %) and transferred to oxyfuel conditions. Hot flue gas is recirculated at the aimed membrane temperature of 850°C. The combustion conditions for a stable oxyfuel combustion were an oxygen concentration of 24% and an air excess ratio of 1.15. An oxygen separation ratio of 90% is optimal for efficient heat integration. Highest efficiency is reached for a feed air pressure of 5 bar. However, this still resulted in a total efficiency drop of 5.6 %-points, from which 4.2 %.-points are due to CO2 compression and purification. Flue gas compositions of the proposed concept were calculated for various coals and conditions. As a standard, the composition H2O 25.4%; N2 5.9%; O2 4%; CO2 64.1%; NO 142 ppm; SO2 2150 ppm; CO 10 ppm was set and given to the other WPs for application oriented testing.
Simulations of an oxyfuel cement process: A reference model of a conventional cement plant with a production capacity of 3000 tons clinker per day has been set up according to literature data. Results have been validated against operating data from thyssenkrupp. A full oxyfuel cement design was chosen for the OTM integration where the flue gas which leaves the upper cyclone preheater undergoes a cleaning/dedusting step to be partially recycled through the cooler. An optimal integration point for the oxygen transfer membrane was identified to be between cooler and rotary kiln. As benchmark, an oxyfuel cement plant with cryogenic air separation was modelled to compare with the OMT version.
Influences of the oxygen separation ratio and the feed pressure for 4-end OTM integration were studied. For an energy efficient process, a high oxygen separation ratio of 0.9 was chosen. Two different pressure scenarios were evaluated. One, where the energy output of the turbine was maximised to decrease the consumed electric energy at 10 bar feed pressure (additionally 3.2 MW electric energy and 29.9 MW fuel energy). And one, where the total energy demand was optimized leading to a 5 bar design (additionally 5.8 MW electric energy and 6.7 MW fuel energy).
CO2 emissions were for the standard cement plant 86.9 t/h, for the first OTM case 8.3 t/h, for the second OTM case 5.2 t/h and for the CD 5.6 t/h. Furthermore, the flue gas composition for varying oxygen purity, oxygen excess, fuel composition, raw material composition and false air intrusion were discussed. The standard air composition at 100% oxygen purity and an air excess of 1.3 with reduced false air intrusion is: CO2 78.6%; N2 3.2%; H2O 15.3%; O2 2.9%; Ar 384 ppm.
WP 5.2 Scale up rules
The proposed oxyfuel concepts involve a number of technical hurdles that are not state of the art yet. For this purpose, investigations were conducted.
Rotating Equipment: Blowers and compressors are available standard equipment. Expanders are only state-of-the-art until a temperature of 500°C. Possibly, FCC expanders used in fluid catalytic cracking processes can be used after qualification.
Flue gas cleaning: High temperature flue gas desulphurization can be realized with limestone particle spray which is used in the power industry, especially in fluidized bed combustion boilers. With this method, a sulphur concentration of 250 ppm can be achieved in the sweep gas, which was set as a target for membrane development. Since the particle size of the adsorbent is in the same range as fly ash, both components can be separated in one step by ceramic filters.
Heat transfer equipment: For the air preheating up to membrane temperature, two different concepts have been analysed. The first concept is based on shell-and-tube heat exchangers (STHE) where heat is transferred counter-currently from hot permeate or hot retentate to cold pressurized air. Using STHE, heat can be transferred very efficiently due to the relatively high attainable gas velocities and counter-current heat exchange. Drawback of STHEs, however, is the complicated manufacturing process and the requirement of a separate shell for each heat exchanger, which, depending on operation temperature, consist of expensive material. The second concept is the use of a waste heat recovery unit in form of a flue gas stack. This approach avoids the use of separate shells for each heat exchanger by using a single channel which is equipped with refractory lining for the low-pressure sweep gas. The pressurized air flows through tubes which are distributed as bundles inside the channel. Compared to the STHE concept not only the shells of the heat exchangers can be avoided, but also the steam generation can be integrated, saving additional vessels for boilers. The heat transfer mode is essentially cross flow, hence not as efficient as the STHE concept.
Directly fired heaters: The directly fired heater for the final heat-up of the feed air to membrane temperature (from 800 to 850°C) is not standard equipment due to the elevated pressure, but feasible. In principle, an analogy to gas turbine combustors can be used for the equipment.
Membrane unit: Due to maintenance and safety reasons as well as limited scalability of single modules, the membrane area has to be separated into manageable segments. The necessity to use multiple heat exchangers in parallel provides a natural train concept which defines the largest membrane area of a single train. To avoid a plant shutdown in case of membrane fracture or maintenance it has to be possible to block-in the considered section. In analogy to reforming processes, the method of tube nipping can be applied. The mechanical limit for nipping gives an upper bound on the tube diameter which is 100 mm. Assuming a maximum gas velocity of 25 m/s of the biggest volume flow, the number of blocks per train is defined. Besides the nipping method, also, reversible sealing could be used but suitable high temperature block valves are very expensive.
The segmentation of the total membrane area is a method to compensate for membrane failure probability, which is a statistical value. The specific fracture probability of a single module and the total membrane area determine the required degree of redundancy and thereby the feasibility of the OTM concept within large-scale applications. Due to the high amount of single membranes in a full-scale application, even low fracture probabilities mean that a certain amount of membranes will fail during operation. This has particular relevance for application in all investigated industrial process. E.g. in the cement industry, the clinker kiln lines at cement plants are usually continuously operated for the period of one year. At the current state, required membrane area as well as specific fracture probability are too high.
WP 5.3 Cost estimations
Cost calculations for oxyfuel IGCC: The final step of the process evaluation was the dimensioning of the equipment and the price calculation. Capital expenditures and operational expenditures were estimated to be respectively 1,045 MM$ and 380 MM$. To compare the economic competitiveness of the IGCC power plants the net present value was estimated. Calculations lead to a minimal electricity price of 13.7 c$/kWh for the OTM oxyfuel IGCC compared to 5.8 c$/kWh for conventional IGCC and 7.7 c$/kWh for pre-combustion IGCC from literature. Main cost factors for the oxyfuel OTM IGCC are high membrane costs and high natural gas prices. Apparently at higher CO2 emission costs (of 91 € per ton CO2) oxyfuel IGCC becomes financial attractive compared to the no capture case, but then pre-combustion would be the most favourable. Therefore, research on the oxyfuel application in an IGCC plant was stopped and pre-combustion was put into focus.
Cost calculations for pre-combustion IGCC: Parallel to the oxyfuel case, individual equipment pieces for pre-combustion were dimensioned and prices calculated. The capital expenditures and operational expenditures were estimated to be respectively 944 MM$ and 265 MM$. For the OTM pre-combustion (250 MW), an electricity price of 19.8 c$/kWh is calculated which is 2.6 c$/kWh cheaper than the oxyfuel process, but still 2 c$/kWh more expensive than pre-combustion with cryogenic air separation. After the scale-up of the plant from 250 MW to 750 MW output, the resulted electricity price was 9.7 c$/kWh. An increase of the gas turbine operation temperature to 1200 °C would lead to an electricity price decline of 1 c$/kWh. The reduction of the OTM operation temperature to 200 °C also would cause an electricity price decline of 1 c$/kWh. However, with all these improvements OTM is still not competitive to the cryogenic air separation when considering the CapEx costs.
Cost calculations for an oxyfuel power plant: For the membrane periphery, partly, Linde in-house cost estimates scaled from executed and proposal projects were used. Partly, vendors were asked for budgetary offers. This resulted in a CapEx of 375 M€ with an accuracy of +/- 30%. For the power plant section, costs were estimated by using literature data of 7 power plant cases for functional groups of equipment and suitable scaling methods. Relevant technology used in the GreenCC model and supplementary flue gas cleaning was considered. Further direct and indirect plant costs as well as working capital were added. Finally, the following CapEx values were achieved: OTM 1,744 M$, CD 1,827 M$ and reference Case 1,153 M$.
The levelized costs of electricity (LCOE) were calculated. The LCOE as a basic assessment method of the cost competitiveness which incorporates all costs over its lifetime. For the reference case the LCOE is 5.64 c€/kWh. Neglecting the costs for the membrane module itself, the OTM power plant produces electricity at 9.87 c€/kWh which is 2 c€/kWhel cheaper than the CD case. This leaves a CapEx margin of 520 M€ for the membrane modules before both electricity prices become even. As a result, a price of 1040 € per m2 would be tolerable.
It is evident that at CO2 emission costs of 0 €/t, the conventional power plant is favourable in terms of electricity price. For a value of 96 €/t a tie between OTM and the reference case is reached (12.87 c€/kWhel). In this scenario, the CD case would have the highest electricity price (12.92 c€/kWhel). In conclusion, OTM will become a viable solution when the CO2 emission cost exceeds 96 €/t or when governmental regulation will drastically limit the CO2 emissions.
Cost calculations for an oxyfuel cement plant: Finally, the product costs of the different cement process options were determined. Firstly, the capital costs for the oxyfuel cement process (73.8 M€) and the conventional process (60.3 M€) were calculated. For the OTM and CD option, this process part hardly changes, thus, the same CapEx values can be used for both. For the membrane periphery, the total CapEx was 25.47 M€.
Subsequently, the additional operational costs for OTM and CD were determined to obtain a target membrane price up to which the OTM is favourable over the CD. Compared to the conventional cement plant, for the OTM scenario costs for additional electricity of 7.00 €/t and additional costs for natural gas of 0.40 €/t have to be supplied. In case of the cryogenic distillation, additional costs for electricity of 10.87 €/t and for fuel 0.42 €/t accumulate. With a total price of 103 €/t for the cement the additional costs represent a rise in product costs of 7.2% (OTM) and 10.9% (CD). Since the cement industry is price sensitive, these supplementary costs would have to be forced by governmental regulation.
For the membrane technique to compete with the CD, membrane module costs of 17.0 M€ are acceptable assuming an exchange of the whole modules every 4 years (44 000 m2). With currently achievable oxygen transfer rates, this means 387 €/m2 or more general 1.93 M€/(kg/s).
Membrane Module Investment costs: For the cost estimate of the membrane module, the costs for the ceramic membrane component, the housing and the joining were taken into account. Solvay as powder supplier and Kerafol as manufacturers of ceramic SOFC cells provided information on the raw material and manufacturing costs of the ceramic membrane component in a technologically relevant amount and number. However, the volumes under consideration are far below the quantities required, as individual companies do not maintain such production capacities. Based on the information provided, it is not possible to distinguish explicitly between capital expenditures and operational expenditures. The cost estimate for the housing and assembling was made on the basis of the costs and working hours spent on the test module. In total, membrane module costs of 16,400 €/m2 were accumulated (LSCF ceramic material costs 700 €/m2; membrane manufacturing costs 4200 €/m2; housing and sealing 11500 €/m2). Several improvement possibilities were proposed which could drastically lower the membrane prices in future to meet the required target prices for industrial application.
WP 5.4 Clinker burning process in oxyfuel atmosphere
The oxyfuel technology affects directly the clinker burning process inside the kiln. The changed combustion atmosphere has an influence on the chemical reactions as well as on the temperature profile within the kiln which affects the clinker phase formation and consequently the product quality. For an industrial application of oxyfuel operations, it must be ensured that the product quality is not decreased.
With this background, laboratory tests have been conducted by the European Cement Research Academy, published in their technical reports on CCS, to show the impact of a high CO2 partial pressure on the decarbonation reaction. The conducted experiments have shown an increase of the reaction enthalpy since the decarbonation reaction shifted towards 80 K higher temperatures compared to conventional operation. Variations of the alite / belite ratio were not identified with changing production atmosphere. In addition, the ratio of C3A/ C4AF is almost not affected. Furthermore, microstructural analysis under the microscope did not reveal any transformations. Only a slight increase in the heat of hydration was found indicating a higher reactivity of the samples produced under oxyfuel conditions.

Potential Impact:
The reduction or elimination of CO2 emissions from power plants fuelled by fossil fuels (coal, gas) and cement production plants is a major target in the current socio-economic, environmental and political situation and discussion to reduce greenhouse gas emissions. This mission can be achieved by introducing gas separation techniques using membrane technology, which is associated with significantly lower efficiency losses in the power plant process compared to conventional separation technologies. To reduce CO2-emissions by 50 to 60% before 2050, a significant contribution of carbon capture and storage or use is required. Concerning the emissions from European fossil fuel power plants, the “European Industrial Initiative on CCS” wants to reach a near zero level until 2020 by means of economic feasible CCS technologies. After a further optimization the technology will be transferred to other carbon-intensive industries. A significant cost reduction for oxygen production is expected to be achieved by Oxygen Transport Membranes (OTM), including lower efficiency losses as compared to today’s best cryogenic technology. This is a key requirement to make oxyfuel concepts a feasible future option for emission-free energy generation. The GREEN-CC process concept offers a very promising solution that has great potential in reducing the cost of CO2 capture. Companies like Linde AG, Thyssen Krupp Industrial Solutions, Shell worldwide leaders in their fields, will strongly participate in the GREEN-CC project in order to reinforce technology transfer. In addition to storage CO2 use is a relevant topic for future activities. In a world of increasing amount of renewable energies highly stable oxygen transport membranes like dual phase composites can be also used for “Power/Heat to X (Fuels or Chemicals)” technologies in a membrane reactor.
In the chemical industry, process intensification can be achieved through the integration of catalytic converters and membrane separators. High product selectivity is obtained through the control of oxygen dosing into the catalyst and reducing the risk derived from explosive mixtures air/methane in fixed bed reactor with concomitant reactant feeding. Process intensification strategy permits dramatic reductions in the size of operation units in chemical plants, in order to combine given production objectives with waste and energy consumption reduction. Air separation using OTMs operating at process temperature of the chemical reactions considered in this project – Oxidative coupling of methane – makes the use of methane as feedstock economically and ecologically attractive by energy saving, resulting in mitigation of CO2 emission.
Results from research work related to material development (WP1 and 2) and modelling (WP1.5 2.1 and WP5) within GREEN-CC provide the basis for further technological development and industrial application. Development of membranes with high flux, high selectivity and stability, and a commercial level of throughput require (i) deep understanding of the physico-chemical and thermo-mechanical properties of complex multicomponent perovskites and thin films under application conditions, which is achieved here by computer modelling and (ii) reliable processing of thin film membranes with sufficient long-time stability in application conditions. The close to application or actual application condition material testing will be also be studied thoroughly (WP3). Another focus will be on the development and construction of a proof-of-concept module (WP4 and WP5). To achieve the overall goals of the project and to open up the perspective towards industrial implementation, it is very important to build a bridge between research and a potential industrial application.
The project partners are working towards commercialization of membrane technology, and acknowledges that open dissemination channels are the best means of gathering forces for this cause. GREEN-CC targets in the demonstration of an asymmetric thin-film oxygen transport membrane with superior performance with regard to high oxygen permeation, infinite oxygen selectivity, and high stability.

In the duration of the GREEN-CC Project from 01.09.2013 until 31.12.2017 the scientific results was already published in a large number of peer-reviewed journals and in international conferences and workshops. This is reported already in Deliverable 6.7. This publishing of the results will be continued in the following years from all partners. That means that Deliverable 6.7. fulfills already the targeted values of the project but it currently is not final document. Also some PhD thesis’s at the universities are not completed yet and will be finalized and published in near future. At the ICIM 2018 (International Conference of Inorganic Membranes), which is one of the most relevant conferences in the Inorganic Membrane community a lot of partners, will present their results of the project. Presented results will be published after ICIM conference in a special issue of Separation and Separation Technology edited by Ingolf Voigt form Fraunhofer IKTS and Wilhelm A. Meulenberg, the coordinator of the GREEN-CC project. This is only one example and will be continued with other relevant conferences in the following years. In addition to this, a press release will be prepared about the outcome of the GREEN-CC project in the first half year of 2018, which is planned to be published in different media and in worldwide web.
The collaboration of the coordinator Wilhelm A. Meulenberg resulted to a full part time Professorship in the field “Ionic Conducting membranes” at Partner Organization University of Twente and a strategic partnership with JÜLICH. Some results of the GREEN-CC Project will be used for teaching aspects within this field at University of Twente.
In parallel to these actions the project results built a basis for a lot of new projects related to these issues. Examples for this are project proposals between different partners of the consortium like a bilateral application of JÜLICH with RSE to the Helmholtz-Association in the European Partnering Program submitted in February 2018. In this project the GREEN-CC results are planned to be used and continued, to solve e.g. sealing problems and to optimize the GREEN-CC stack design. Also the manufacturing of larger cells of second generation GREEN-CC dual phase membranes is planned. Other knowledge of the GREEN-CC project regarding materials for membranes, catalysts and manufacturing issues are addressed in the MARCAS “Membrane Assisted Reactors for CAtalytic C1-C4 conversionS” proposal in HORIZON 2020 program which was submitted in February 2018 by the coordinator Dr. Fausto Gallucci, Eindhoven University of Technology. GREEN-CC Partners CSIC and JÜLICH will participate here. In addition to these examples a lot of other bilateral ideas for proposals are generated and will be submitted to different funding organizations. The knowledge build in GREEN-CC could be applied in advanced chemical reactors for hydrocarbon upgrading and conversion of renewables energy (mostly solar and geothermal heat) and waste heat from industrial processes.
Some of these projects are related to the field “Catalytic Membrane Reactors” for the production of energy carries or commodity chemicals which was a strong recommendation of the reviewers to transfer the knowledge also to this field. Based on these recommendations JÜLICH wrote a review article published in Journal of Membrane Science in 2017 and a confidential strategy paper of the potential use of the GREEN-CC activities and hydrogen conductors within the field of Membrane Reactors inside JÜLICH.
It is planned to build a large program in JÜLICH in the next budget period 2020ff which includes only in JÜLICH activities in seven different Institutes (IEK-1, IEK-2, IEK-3, IEK-10, ER-C, ZEA-1 and ZEA-3) in collaboration with several international GREEN-CC Partners. This planned program covers all activities from material development to system and integration aspects. It is focused to bring membrane applications closer to the market. Discussions with relevant companies like PRAXAIR (USA), Linde (GER), Shell (NL), Thyssen-Krupp (GER), BASF (GER), HELPE-RES (Greece) and other companies are started and an exploitation workshop was carried out in the frame of the GREEN-CC project. PRAXAIR organized in November 2017 a workshop, held in Buffalo (NY, USA), and related to the application of ionic conductors within this field. In this workshop a lot of large, medium and small size companies participated accompanied by a minor amount of scientists. The coordinator of the GREEN-CC project also participated, but not directly in the function of GREEN-CC coordinator. One outcome of the workshop was that the industry sees a large amount of potential applications in pure gas supply with 3-end modules and production of commodity chemicals or synthetic fuels in membrane reactors. Typical applications are e.g. production of pure oxygen, pure hydrogen, and syngas or CO2 utilization to e.g. methane. Carbon Capture in large scale (Oxyfuel in Power Plants or Cement Industry) is still very interesting but too large for first stage applications, because of missing trust in industry in new technologies. Nearly all applications are targeted for the reduction of carbon dioxide emissions to the atmosphere. Another clear outcome was that stable long term operation for first stage industrial application and market integration is much more relevant than performance. Industry wants first small or medium sized stable standalone devices, which are easy to operate (one button to start up or shut down). After showing the long term stable operation at industry the second stage will be the integration in processes on a medium or large scale. Reason for this is to build up trust in industry in the technology. Used materials in the first stage are mostly highly stable but low or intermediate flux materials, often used as bulk membranes. Costs are relevant but not the most important in first stage. More important for industry is the potential of in cost reduction in future by using of cheap manufacturing steps and materials. A cost study by industry was not shown but cost targets, e.g. for oxygen production were shown. The presentations are available for the coordinator but it was agreed between the parties to keep them confidential. For this reason they are not directly available for GREEN-CC consortium. A second workshop end of 2018 is discussed but currently not decided.

List of Websites:
http://www.green-cc.eu/green-cc/DE/Home/home_node.html

Final Report Summary - GREEN-CC (Graded Membranes for Energy Efficient New Generation Carbon Capture Process)

Diese Seite teilen

Herunterladen