This study investigates machine-learning-based prediction of bio-oil yield and quality indicators during the hydrothermal liquefaction (HTL) of lignocellulosic biomass. A curated literature-derived database was used to develop multilayer perceptron, support vector regression, decision tree, random forest (RF), and extreme gradient boosting (XGB) models for predicting bio-oil yield, elemental ratios, carbon retention, oxygen removal efficiency, and higher heating value. Predictor selection combined process knowledge with correlation analysis, while model hyperparameters were optimized using particle swarm optimization. XGB and RF generally provided the strongest performance; however, the models developed for most secondary outputs exhibited limited generalization. Consequently, detailed post-prediction analysis was restricted to bio-oil yield, for which the XGB model showed the most reliable predictive performance. Inclusion of the effective water fill ratio (Vw/Vr) substantially improved yield prediction by representing variations in solvent availability, reactor filling, and associated physicochemical interactions. SHapley Additive exPlanations ranked Vw/Vr as the most influential predictor, whereas Spearman analysis identified reaction time as the strongest monotonic factor (ρ ≈ −0.43), followed by Vw/Vr (ρ ≈ +0.34). Segmented regression and localized partial-dependence analyses reconciled these rankings by showing that the influence of reaction time was concentrated mainly below approximately 80 min, while Vw/Vr produced stronger nonlinear and regime-dependent changes, particularly below a model-derived breakpoint near 13.9. Joint-bootstrap Monte Carlo and residual-conformal analyses were further used to quantify predictive uncertainty and establish a model-credibility envelope based on interval width and applicability-domain support. Overall, the proposed framework provides an interpretable and uncertainty-aware approach for modeling lignocellulosic HTL and prioritizing conditions for subsequent process optimization.
This study investigates the influence of ash content on heat transfer during biomass pyrolysis. Lumped capacitance and thermally thin particle models were used to predict the thermal behavior of biomass particles in a furnace, and results were compared with measurements from a thermogravimetric system (TGA). In the lumped capacitance model, the biomass particle is assumed to behave as a thermally uniform body subject to radiative heating in the furnace. The predictive model establishes three heating stages, including the initial heating stage (dehydration), pyrolysis, and the post-pyrolysis stage. Empirical correlations were developed to account for the effects of biomass biochemical composition and the heating rate on the decomposition process. This study addresses a key knowledge gap by incorporating the thermal effects of ash content into predictive models using its effective thermophysical properties. The results indicate that particle size, ash content, and heating rate significantly affect the pyrolysis characteristics, including pyrolysis temperature, particle conversion, and reaction duration. The effect of K2CO3 was observed through a shift in the pyrolysis temperature and a reduction in the pyrolysis initiation time at higher heating rates. Overall, these findings highlight the importance of accounting for ash in biomass pyrolysis and provide new insights into the ash effect on temperature profiles, pyrolysis time, and conversion behavior.
Probiotic Bacillus species are being investigated as sustainable interventions to enhance health and disease resilience in aquaculture. However, the functional basis, biosafety profile, and genomic determinants of probiotic suitability in shrimp gut-associated Bacillus strains remain insufficiently characterized. In this study, a Bacillus strain (KNSH39) isolated from the intestine of Pacific white shrimp (Litopenaeus vannamei) was evaluated using integrated phenotypic, functional, and genome-resolved approaches. Classical assays assessed morphology, sporulation, antibiotic susceptibility, gastrointestinal tolerance, storage stability, and antibacterial activity of cell-free supernatant under thermal and pH stress. Hybrid whole-genome sequencing using Oxford Nanopore Technologies and Illumina platforms enabled high-quality assembly, followed by comprehensive functional annotation, mobilome analysis, biosynthetic gene cluster prediction, and comparative genomics. KNSH39 exhibited strong sporulation capacity (98.04
In response to the increasing global population and economy, water security has become a serious concern, and thus, the enhancement in wastewater treatment efficiency is of utmost importance. One solution is integrating multiple wastewater treatment approaches, such as coupling advanced oxidation processes (e.g., Fenton oxidation) with adsorption. Therefore, this study Cu-Fe bimetallic biocomposites prepared from activated hydrochar/biochar to remove methylene blue (MB). The results showed that Fe-Cu@activated hydrochar obtained at 800 degrees C with a Fe/Cu molar ratio of 1:0.5 (AHC-800-1) showed the highest MB removal performance compared to other biochar-based biocomposites, achieving > 90 % removal within 20 min. After the screening study, the efficiency of MB removal over AHC-800-1 was optimized by investigating AHC-800-1 concentration, H2O2 concentration, and pH value. The maximum MB removal efficiency was attained at 100 ppm biocomposite, 5 % H2O2, and a pH of 7. The recyclability study showed a significant drop in removal efficiency from 99.21 % to 47.86 % after three cycles, mainly due to the reduced surface area, lower total pore volume, and metal leaching. The mechanistic study revealed that adsorption over the porous structure was the major contributor to MB removal, while Fenton oxidation contributed much less. In short, this study provides a potential new solution to remove contaminants using bimetallic biocomposite materials prepared from activated hydrochar/biochar via adsorption coupled with Fenton oxidation.
We study dynamic treatment regimes under contemporaneous confounding: at each stage, an unmeasured factor may affect both treatment and the next observed state, but has no further direct effect on later stages. From the observational distribution, the causal graph, and specified structural restrictions on possible next states, we construct at each state–treatment pair the set of transition probabilities compatible with this information. These local sets require no sensitivity parameter and can be combined by backward induction to obtain lower and upper bounds on the expected outcome under any given treatment regime. Our main result shows that these bounds are sharp: their endpoints are exactly the smallest and largest expected outcomes generated by causal models compatible with the same observational distribution, causal graph, and structural restrictions. The same backward-induction method gives a maximin rule for choosing treatments by maximizing the worst-case expected outcome. Thus, for this class of models with contemporaneous confounding, the framework generalizes the classical g-formula of for evaluating treatment regimes and the Q-learning framework of for selecting them. When the local transition probabilities are identified, the two recursions reduce to these classical methods.
