The project has progressed positively in both the individual and merged components that compose the final design of the reaction robot. On the individual part, three main aspects of the design of the reaction robot have been developed: the design of the light-harvesting system, the design of the materials for the photocatalytic/separation functions and the development of models to optimize the productivity via machine learning.
In the light harvesting part through the Luminescent Solar Concentrators, a novel reactor design to harvest and down-convert photons within the solar spectrum has been developed. This system consists of a silicon-based photoreactor immersed in high-refractive-index silicon oil, which can be doped with a suitable chromophore. The reactor´s efficiency has been validated through a benchmark reaction, showing stable operation up to 250 °C and 5 bar. The modular design enables a seamless, non-invasive switch between reactor types, allowing for greater flexibility in experimental conditions.
In terms of material design, a temperature-dependent strategy for 3D (sub)nanostructured carbon nitrides (CN) has been developed. By systematically tuning the synthesis temperature we have achieved a precise control over the structural evolution, morphology, and electronic properties of carbon nitride photocatalysts, especially for water-splitting reactions into H2. Beyond such photocatalytic functions, the application of Chemical Vapor Deposition (CVD) enabled the formation of uniform, well-adherent carbon nitride nanostructures on flat C-membranes, while preserving their intrinsic permeability. The intimate functions governing the reactivity of CN materials have been deeply studied through a diversified synthesis approach, either by studying carbon-nitride alone, functionalized with proper metal-based co-catalysts (e.g. cobaloxime), or in combination with different materials, including graphene and metal-organic frameworks. Their active role has been corroborated under flow-conditions in photocatalysis, as well as advanced spectroscopic and modelling methods.
Concerning the introduction of product separation functions via membranes, a new polymeric precursor capable of increasing H2 permeance one order of magnitude was successfully experimented. In order to improve the affinity of the membranes towards H2, Pd was added in the carbon matrix. Results about the specific performance of Pd-containing membranes will be collected right after the completion of the aging process of the membranes. This has allowed a transition from tubular to flat supported carbon membranes (CM) and polystyrene pre-treatment to optimize the CVD step: an optimized method to produce flat supported CM via ultrasonic-spray-coating has been developed. Additionally, a layer of polystyrene on the opposite side of the supported flat CM was applied in order to better localize the effects of the chemical vapor deposition treatment.
Concerning the automatization of the reaction robot, progress has been made to introduce the machine learning models into a photocatalytic systems. The knowledge generated in the individual components of the reaction robot has been implemented in the design of the hardware for the reaction robot and currently the models are being implemented to start the optimization of the reaction performance. In addition, the experimental data will be validated and compared to a techno-economic assessment that compares the developed technology with those reported for similar purposes.
Regarding the dissemination, exploitation and management of results, the CATART consortium has attended a significant number of congresses to disseminate the main breakthroughs. In addition, actions have been taken to patent the design of photocatalysts. The participation of CATART project in the organization of a Solar-to-X conference in Belgium, in synergy with different EU projects, as well as in venture programs allowed reformulating the exploitation plan thanks to the feedback received from experts and stakeholders. Finally, 13 deliverables have been submitted during the RP2.
