This research study describes a machine learning (ML)-driven model for producing smart structural materials via additive manufacturing (AM) by extrusion. A 3D concrete printing system was used to make cementitious composites that were reinforced with carbon nanotubes (CNTs) and graphene nanoplatelets (GNPs). Random Forest (RF), Support Vector Machine (SVM), and Artificial Neural Network (ANN) models were used to undergo supervised learning on an experimental dataset consisting of 320 specimens to predict compressive strength, electrical conductivity, and print quality as dependent on process parameters and material composition. The highest R2 of compressive strength prediction of SVM was 0.946, whereas RF had the highest R2 of 0.987, which was used to predict electrical conductivity. Optimization of parameters guided by ML had a 61.8% enhancement of compressive strength and 30.5 times increase in electrical conductivity in comparison to non-optimized baselines. Nanomaterial networks were also found to be conductive, allowing individual networks to detect their strain levels through changes in current at a strain of 0.1%, which facilitates real-time structural health sensing. The artificial system showed a 31% decrease in CO2 emissions and a 58.8% decrease in material wastage compared with the usual way of building, proving to be a valid route towards intelligent and sustainable infrastructure.
Poly(butylene adipate-co-terephthalate) (PBAT) is a promising biodegradable polyester, but its low strength limits broader application. In principle, blending PBAT with polylactide (PLA) can combine toughness and stiffness, yet severe immiscibility usually leads to poor interfacial adhesion and unsatisfactory overall performance. Here, a bio-based lignin-epoxy composite compatibilizer (E-FL) was developed by premixing ethanol-fractionated lignin (FL) with a protocatechuic-acid-derived epoxy compound and introducing it into PBAT/PLA blends through reactive melt processing. Fractionation enriched lignin fractions with lower molecular weight and higher hydroxyl content, thereby improving reactivity and dispersibility. During melt blending, E-FL promoted interfacial reactions with PBAT and PLA end groups, increased melt torque and molecular weight, refined the dispersed PLA domains, and reduced the Tg gap between the two phases. At an E-FL loading of 3 wt%, the blend exhibited the best balance of performance, with a tensile strength of 36.1 MPa, an elongation at break of 1035%, and a fracture toughness of 238.3 MJ/m3. This work provides a sustainable strategy for converting lignin into a high-efficiency reactive compatibilizer and offers a practical route to high-performance PBAT-based biodegradable blends.
Gastrodin is a phenolic glycoside and the principal bioactive compound of Gastrodia elata. Owing to its potent neuroprotective, antioxidant, and therapeutic properties, gastrodin has attracted increasing attention and is now widely applied in the pharmaceutical, healthcare, and food industries. Traditional extraction of gastrodin is constrained by limited raw material availability and low yield, making it insufficient to meet the growing market demand. In recent years, microbial biosynthesis has become a preferred route for gastrodin production due to its sustainability, economic feasibility, and high safety. Therefore, developing metabolically engineered strains with enhanced genetic stability, high productivity, and efficient substrate utilization has become an urgent priority for achieving gastrodin biosynthesis. This review introduces the discovery and biosynthetic routes of gastrodin, summarizes its production methods, and discusses recent advances across various microbial chassis systems. It further highlights recent advances in pathway reconstruction and metabolic optimization, with an emphasis on strategies to enhance precursor flux, optimize UDP-glucose biosynthesis and regeneration, and improve glycosyltransferase catalytic activity through protein engineering. Overall, this review provides insights and future directions for developing efficient, genetically stable, and industrially scalable microbial cell factories for sustainable gastrodin production.
Extensive investigations have revealed the precipitation of nanometer-scale silicides, identified as G-phase, within the ferritic matrix of duplex stainless steels during prolonged thermal aging. These silicides typically exhibit a well-defined coherent orientation relationship with the ferrite matrix, specifically (100G//100F, 110G//110F, 111G//111F). Consequently, the authors and their research team proposed a novel concept in 2015: utilizing the G-phase as a primary strengthening phase. It was proposed that through strategic alloy design, these silicides—ordinarily considered deleterious in duplex stainless steels—could be used to develop a new generation of dispersion-strengthened ferritic stainless steels. This approach aims to significantly enhance the yield strength of the alloy while maintaining excellent tensile ductility. Over the past decade, the authors and their research team have focused on nanoscale G-phase dispersion-strengthened ferritic stainless steels. By combining first-principles calculations with thermodynamic database-driven alloy design, a series of new ferritic stainless steel systems based on G-phase strengthening has been developed. These efforts have yielded extensive fundamental results regarding the compositional control, microstructural design, and mechanical properties of silicide-strengthened 20Cr ferritic stainless steels. Based on a comprehensive review of the existing literature, this paper further summarizes the compositional design criteria and microstructural control strategies for G-phase strengthened steels. It is hoped that this work will encourage further fundamental research and industrial applications in this field.
Persistent SARS-CoV-2 antigen has been proposed as a driver of post-COVID condition (PCC), with targeted mass spectrometry multiple reaction monitorin/selected reaction monitoring (MRM/SRM) increasingly invoked as quantitative evidence. We appraise the targeted-MS literature on SARS-CoV-2 antigen in genuine human clinical specimens and re-analyse a focal study, which reported spike and nucleocapsid “protein” concentrations in ng/µL from two proteotypic peptides per target with 13C/15N internal standards. These values are either physically impossible as intact protein or, more likely, raw peptide concentrations reported without the required ≈122-fold molecular-weight correction. Only 15 of 65 patients (26%) had cellular pellet spike above the authors’ own limit of quantification; nucleocapsid was essentially undetectable; and in those 15, the nucleocapsid: spike molar ratio was strongly inverted relative to intact virions, incompatible with a viral source. Critically, no targeted-MS method has ever quantified spike in human blood—the prior literature is nucleocapsid detection in respiratory specimens and spike quantification in vaccine or recombinant material—so the reported blood-spike values lack any validated precedent and exceed the most sensitive validated platform (single-molecule arrays) by several orders of magnitude, with no enrichment step. Finally, 77% of the cohort was vaccinated, and a measurable spike was concentrated among vaccinated individuals. The source’s own supplement inconsistently reports vaccination status. Their 2024 predecessor publication withheld it entirely. The MRM/SRM data, therefore, do not support persistent viral antigen as a general driver of PCC. Minimum standards are proposed: molar reporting, strict limit-of-quantification (LOQ) compliance, qualifier-ion confirmation, vaccine-discrimination peptides, stoichiometric cross-validation, and vaccination-status disclosure. We suggest that the cellular blood component, routinely discarded, warrants direct investigation in the context of spike persistence and PCC symptoms.
