Periodic Reporting for period 3 - SONAR (Modelling for the search for new active materials for redox flow batteries)
Reporting period: 2022-09-01 to 2023-12-31
SONAR has developed a framework for the simulation-based screening of electroactive materials for aqueous and nonaqueous organic redox flow batteries (RFBs). It adopted a multiscale modelling paradigm, in which simulation methods at different physical scales were further advanced and linked by combining physics- and data-based modelling. For the traversal of the different scales, exclusion criteria like solubility, standard potentials and kinetics were defined. The results for individual candidates are stored in a database for further processing. To increase the throughput of the screening, SONAR exploited advanced data integration, analysis and machine-learning techniques. The models were validated by comparison with measurements of redox potentials for known chemistries and measurement data of RFB half-cells and lab-sized test cells.
As planned, the complex procedure is now available to all interested persons, companies and organizations as a service managed by Fraunhofer SCAI on request. To this end, the project website (sonar-redox.eu) has been redesigned and serves as a contact point.
As planned, the complex procedure is now available to all interested persons, companies and organizations as a service managed by Fraunhofer SCAI on request. To this end, the project website (sonar-redox.eu) has been redesigned and serves as a contact point.
A demonstrator of the high-throughput screening process was developed for external use at redoxfox.scai.fraunhofer.de.
The capability of automated potential energy surfaces (PES) exploration algorithms to find products and reactions paths of degradation reactions in organic flow battery electrolytes has been explored. The framework has been applied to degradation reactions of substituted quinones across neutral, basic and acidic conditions. These proof-of-principle applications showed that the proposed framework is able to find the degradation products and reaction mechanisms without their a priori knowledge.
An automatic workflow to optimize the geometry structure of the carbon felt electrode, with consideration of the electrode compression, has been developed. The objective of the optimization algorithm was to maximize the electrolyte utilization rate, in which a homogeneous electrolyte distribution and a high specific surface area was required. The parameter set of the optimization workflow included the fiber diameter, the porosity of the fiber, the fiber direction, the amount of in-plane fiber and the compression ratio.
A flow battery model (RFB-SCL-3D) for performance predictions of a single lab-sized flow battery cell has been developed. The model builds upon and extends the cell performance model for high-throughput screening. However, in contrast to the RfbScFVM model, the RFB-SCL-3D model performs a fully resolved 3D discretization of the electrolyte flow in the porous electrodes and flow field structures. The electrolyte flow is described by the incompressible Navier-Stokes equation in the free-flow regions and the Brinkman equation in the porous electrode structure.
Topology optimization calculations were conducted to enhance the design of a flow battery cell. A homogenized 2D cell model with simple chemistry is developed in COMSOL Multiphysics® and the existing topology optimization framework of the software is used for the calculations.
A spatially resolved 3D microstructure battery model has been developed. For this purpose, a simulation tool developed in-house using the finite volume method. It is set up to simulate a battery half-cell including the membrane as a boundary condition, and consists of coupled, non-overlapping fluid and solid regions. The simulated domain can either be obtained from image reconstruction of the electrode or constructed using CAD software. Both the all-vanadium and TEMPO systems are implemented and successfully validated. In general, the model is capable of providing performance data like half-cell potential or active material concentration distribution within the microstructure and their dependency on the flow regime.
A combined hydraulic/electrochemical/thermal stack model has been developed which incorporate an input interface that allows input parameters to be varied for different chemistries and system designs. System-level modelling considers sizing and energy requirements for pumps.
An optimized techno-economic model was created to enable consideration of further factors influencing the costs and technical properties of organic RFBs. An optimized version of the technoeconomic model was programmed in Python (FLOTE) as a stand-alone software, using the extensive input variables determined in the laboratory itself.
The capability of automated potential energy surfaces (PES) exploration algorithms to find products and reactions paths of degradation reactions in organic flow battery electrolytes has been explored. The framework has been applied to degradation reactions of substituted quinones across neutral, basic and acidic conditions. These proof-of-principle applications showed that the proposed framework is able to find the degradation products and reaction mechanisms without their a priori knowledge.
An automatic workflow to optimize the geometry structure of the carbon felt electrode, with consideration of the electrode compression, has been developed. The objective of the optimization algorithm was to maximize the electrolyte utilization rate, in which a homogeneous electrolyte distribution and a high specific surface area was required. The parameter set of the optimization workflow included the fiber diameter, the porosity of the fiber, the fiber direction, the amount of in-plane fiber and the compression ratio.
A flow battery model (RFB-SCL-3D) for performance predictions of a single lab-sized flow battery cell has been developed. The model builds upon and extends the cell performance model for high-throughput screening. However, in contrast to the RfbScFVM model, the RFB-SCL-3D model performs a fully resolved 3D discretization of the electrolyte flow in the porous electrodes and flow field structures. The electrolyte flow is described by the incompressible Navier-Stokes equation in the free-flow regions and the Brinkman equation in the porous electrode structure.
Topology optimization calculations were conducted to enhance the design of a flow battery cell. A homogenized 2D cell model with simple chemistry is developed in COMSOL Multiphysics® and the existing topology optimization framework of the software is used for the calculations.
A spatially resolved 3D microstructure battery model has been developed. For this purpose, a simulation tool developed in-house using the finite volume method. It is set up to simulate a battery half-cell including the membrane as a boundary condition, and consists of coupled, non-overlapping fluid and solid regions. The simulated domain can either be obtained from image reconstruction of the electrode or constructed using CAD software. Both the all-vanadium and TEMPO systems are implemented and successfully validated. In general, the model is capable of providing performance data like half-cell potential or active material concentration distribution within the microstructure and their dependency on the flow regime.
A combined hydraulic/electrochemical/thermal stack model has been developed which incorporate an input interface that allows input parameters to be varied for different chemistries and system designs. System-level modelling considers sizing and energy requirements for pumps.
An optimized techno-economic model was created to enable consideration of further factors influencing the costs and technical properties of organic RFBs. An optimized version of the technoeconomic model was programmed in Python (FLOTE) as a stand-alone software, using the extensive input variables determined in the laboratory itself.
The developments made in SONAR beyond the state of the art were in the areas
1) the development of a high-throughput screening process.
2) development and optimisation of individual scale models and
In SONAR, a high-throughput screening process was developed that enables the rapid search for novel organic active materials for redox-flow batteries. In the field of atomistic modeling and simulations, redox potentials and especially solubilities can be calculated much more accurately by considering solvation energies on the basis of hybrid cluster continuums. In addition, it was shown using quinones that side reactions can be found on the basis of an automatic pathfinder method for creating energy level schemes and can therefore potentially be taken into account automatically in the future. Kinetic data of redox pairs were obtained by kinetic Monte Carlo simulations, taking into account side reactions such as dimerization.
Three-dimensional electrode structures were optimized using Lattice-Boltzmann simulation and enable improved flow battery cell structures that can be automatically adapted specifically for the respective redox pair and the respective electrolyte. A 0D cell model allows rapid calculation of battery performance values, which was integrated into the high-throughput process and is publicly available as open-source code. A more complex 3D cell model was also developed and will be commercially available. Another cell model was based on OpenFoam, also taking into account the complex electrode structure. Automated parameter optimizations were performed to develop optimized 3D electrode structures that have better properties than conventionally used ones.
Coupled electrochemical-hydraulic-thermal models for mapping the cell, stack and system properties of batteries with organic active materials have been developed and will be made available as standalone software. A techno-economic model was developed and simulations carried out to calculate total costs and the distribution of costs among components. Functions such as the optimization potential were introduced, which provides a quantifiable statement about the degree of cost reduction when optimizing components. The techno-economic model was programmed in Python as standalone software and was published as open source.
SONAR’s planned impact is to enable a high-throughput screening to find new active materials for future novel organic redox flow batteries. By offering a service to interested representatives from industry and academia, it should become possible to simulatively identify new organic active materials for future redox flow batteries as stationary storage devices for renewable energies. This service has been implemented on the sonar-redox.eu website and is available. REDOXFOX offers a demonstration of the possibilities of screening at redoxfox.scai.fraunhofer.de.
1) the development of a high-throughput screening process.
2) development and optimisation of individual scale models and
In SONAR, a high-throughput screening process was developed that enables the rapid search for novel organic active materials for redox-flow batteries. In the field of atomistic modeling and simulations, redox potentials and especially solubilities can be calculated much more accurately by considering solvation energies on the basis of hybrid cluster continuums. In addition, it was shown using quinones that side reactions can be found on the basis of an automatic pathfinder method for creating energy level schemes and can therefore potentially be taken into account automatically in the future. Kinetic data of redox pairs were obtained by kinetic Monte Carlo simulations, taking into account side reactions such as dimerization.
Three-dimensional electrode structures were optimized using Lattice-Boltzmann simulation and enable improved flow battery cell structures that can be automatically adapted specifically for the respective redox pair and the respective electrolyte. A 0D cell model allows rapid calculation of battery performance values, which was integrated into the high-throughput process and is publicly available as open-source code. A more complex 3D cell model was also developed and will be commercially available. Another cell model was based on OpenFoam, also taking into account the complex electrode structure. Automated parameter optimizations were performed to develop optimized 3D electrode structures that have better properties than conventionally used ones.
Coupled electrochemical-hydraulic-thermal models for mapping the cell, stack and system properties of batteries with organic active materials have been developed and will be made available as standalone software. A techno-economic model was developed and simulations carried out to calculate total costs and the distribution of costs among components. Functions such as the optimization potential were introduced, which provides a quantifiable statement about the degree of cost reduction when optimizing components. The techno-economic model was programmed in Python as standalone software and was published as open source.
SONAR’s planned impact is to enable a high-throughput screening to find new active materials for future novel organic redox flow batteries. By offering a service to interested representatives from industry and academia, it should become possible to simulatively identify new organic active materials for future redox flow batteries as stationary storage devices for renewable energies. This service has been implemented on the sonar-redox.eu website and is available. REDOXFOX offers a demonstration of the possibilities of screening at redoxfox.scai.fraunhofer.de.