A causal artificial intelligence (AI) model was developed to support the discovery of new superconducting materials by analysing causal relationships between interval-valued elemental descriptors and the superconducting critical temperature, Tc. The aim is to explore a broad elemental composition space without requiring prior knowledge of material structure. Using a University of California, Irvine dataset comprising 21,263 materials with chemical formulae, 81 features, and Tc values were aggregated into temperature-based intervals and analysed within a reproducing kernel Hilbert space framework to infer a causal directed acyclic graph. Three interval features emerged as direct causal drivers of T: the standard deviation of mass density, the weighted geometric mean c of electron affinity, and the weighted geometric mean of valence. A random forest model using all predictors achieved an R2 of approximately 92.9 %, while the causal model, using only these three features, achieved anR2 of approximately 89.7 %. Under out-of-distribution splits, the causal model demonstrated superior robustness. Estimated interventional ("do") effects revealed nonlinear behaviour, and counterfactual analyses of hypothetical interventions further demonstrated the potential of causal AI for guiding exploration of new materials.
The hydrogenation of naphtha is critical to producing stable, clean gasoline, yet current catalysts often lack selectivity and stability. In this work, a novel palladium–magnesium oxide (Pd/MgO) nanocatalyst was developed to address these challenges. The catalyst was prepared by reduction of Na2PdCl4 on MgO support using sodium borohydride, resulting in well-dispersed Pd nanoparticles with an average size of 1.7 nm and a Pd loading of 0.9 wt.% (nominally 1 wt.%). The size, composition, and dispersion of the nanoparticles were confirmed using transmission electron microscopy, X-ray diffraction, X-ray photoelectron spectroscopy, hydrogen pulse chemisorption, inductively coupled plasma atomic emission spectroscopy, and hydrogen temperature programmed reduction measurements. Catalytic tests showed great activity of quinoline hydrogenation at 150 °C and 40 atm H₂, with a corrected turnover frequency (TOFcorr) of 6400 h⁻¹, which was almost fourfold higher than Pd/SiO₂ and Pd/Al₂O₃ commercial catalysts (TOFcorr = 1600-1800 h⁻¹). Linear alkenes were hydrogenated with the catalyst at mild conditions (25 °C and 10 atm H₂). Moreover, during biodiesel upgrading, the conversion of polyunsaturated fatty acid methyl esters to stable monounsaturated products was achieved at (100 °C, 1 atm of H₂ with >80 % conversion). Recyclability tests proved that the alkene hydrogenation activity was stable in three cycles, and only slight deactivation in quinoline hydrogenation occurred. This work shows that the Pd/MgO nanocatalysts are promising for enhancing the quality of gasoline, minimizing the formation of gums, and increasing the stability of biodiesel because of their nanoscale dispersion, high selectivity, and recyclability.
A causal artificial intelligence (AI) model was developed to support the discovery of new superconducting materials by analysing causal relationships between interval-valued elemental descriptors and the superconducting critical temperature, Tc. The aim is to explore a broad elemental composition space without requiring prior knowledge of material structure. Using a University of California, Irvine dataset comprising 21,263 materials with chemical formulae, 81 features, and Tc values were aggregated into temperature-based intervals and analysed within a reproducing kernel Hilbert space framework to infer a causal directed acyclic graph. Three interval features emerged as direct causal drivers of Tc: the standard deviation of mass density, the weighted geometric mean of electron affinity, and the weighted geometric mean of valence. A random forest model using all predictors achieved an R² of approximately 92.9 %, while the causal model, using only these three features, achieved an R² of approximately 89.7 %. Under out-of-distribution splits, the causal model demonstrated superior robustness. Estimated interventional (“do”) effects revealed nonlinear behaviour, and counterfactual analyses of hypothetical interventions further demonstrated the potential of causal AI for guiding exploration of new materials.
Magnesium alloys have attracted considerable attention, particularly in the biomedical field, due to their excellent biocompatibility and bone-like properties, which make them suitable materials for biomedical applications. However, their low corrosion resistance and rapid degradation lead to poor functional performance. Therefore, this review focuses on the biomedical applications of magnesium alloys and examines the corrosion mechanisms and corrosion types affecting these alloys when used in the medical field. Among the many techniques, this review focuses on surface coating preparation. Surface coating techniques have helped increase corrosion resistance and biocompatibility while reducing the susceptibility of magnesium alloys to premature degradation. These coatings represent a modern approach for improving the clinical performance of these alloys. However, several challenges remain, including balancing degradation rates with structural integrity and enhancing their compatibility with biological tissues. Accordingly, this review summarizes recent advances in surface coating technologies and provides an overview of the corrosion mechanisms associated with coated magnesium alloys, thereby highlighting future directions and opportunities for their broader application in the biomedical field.
In the present paper, the potential of a locally sourced plant-based adsorbent, Ziziphus lotus stone, for the removal of Congo Red (CR) dye from aqueous media was evaluated. The adsorbent was characterized using X-ray diffraction (XRD), Fourier-transform infrared spectroscopy (FTIR), scanning electron microscopy (SEM)/energy-dispersive X-ray spectroscopy (EDS), thermogravimetric analysis (TGA), differential scanning calorimetry (DSC), and Brunauer-Emmet-Teller (BET) surface area analysis. These combined characterization methods provided a comprehensive profile of the material, confirming the presence of active surface functional groups, a porous structure and high thermal stability. The BET surface area was estimated at 4.7758 m2 g-1, with a pore size of 4.3578 nm. About 10 % enhancement in the adsorption efficiency of Congo Red dye was achieved using response surface methodology (RSM) based on the Box-Behnken Design (BBD), considering three key factors: adsorbent dose, pH, and initial CR dye concentration. As a result, the overall adsorption efficiency reached up to 95 %. The isotherm study, based on non-linear isotherm models, indicated that the adsorption data were best described by the Freundlich model, with maximum adsorption capacity Qmax of 11mg g-1. Thermodynamic analysis revealed negative values of triangle G0, triangle H0 (-11.0959 kJ mol-1), and triangle S0 (-29.587 J K-1 mol-1), indicating that the adsorption process was exothermic, spontaneous, and accompanied by decreased randomness at the solid-liquid interface. Furthermore, a techno-economic assessment of the based-plant adsorbent and the overall adsorption process was conducted. The estimated production cost of the Ziziphus lotus stone adsorbent was estimated at around 0.11304 USD/kg, while the overall adsorption cost was approximately 0.112 USD/m3. These findings confirm the viability and high efficiency of Ziziphus lotus stone as a low-cost adsorbent for largescale and low-cost environmental applications.
