Taufiq Choirul Amri, Riyanarto Sarno, Dwi Sunaryono, Rizqy Ahsana Putri, Alya Kamilah, Irfan Mirda, Kelly Rossa Sungkono, Ratih Nur Esti Anggraini, Sholiq Sholiq, Agus Tri Haryono
{"title":"CALPIR: CALibrated Prediction Intervals and a Rejection rule for machine-learning Cd2+ voltammetry","authors":"Taufiq Choirul Amri, Riyanarto Sarno, Dwi Sunaryono, Rizqy Ahsana Putri, Alya Kamilah, Irfan Mirda, Kelly Rossa Sungkono, Ratih Nur Esti Anggraini, Sholiq Sholiq, Agus Tri Haryono","doi":"10.1016/j.ceja.2026.101522","DOIUrl":"10.1016/j.ceja.2026.101522","url":null,"abstract":"<div><div>Machine learning reads cadmium concentration from cyclic voltammograms and reports a single number. However, an operator judging a discharge against a limit must know how far that number may be from the truth. Default uncertainty estimates are not calibrated, assign one width to every measurement, and place no restraint on a model answering outside the calibration standards. This study proposes CALPIR, CALibrated Prediction Intervals and a Rejection rule, which scales interval width by the disagreement among tree-ensemble members, then rejects a voltammogram that falls outside the calibrated space. Evidence comes from 1497 voltammograms acquired on 2 potentiostats sharing one cell. Empirical coverage matched nominal levels of 0.95, 0.90, and 0.80 to within 1 percentage point on both instruments. Scaling by ensemble disagreement narrowed the interval by factors of 3.61 and 4.48 over the unscaled construction. CALPIR was narrower by factors of 6.8 and 5.8 than a deep ensemble tuned across 40 architectures. At the 0.90 level, reported concentrations fell within a factor of 1.09 to 1.15 of the truth. The rejection rule rejected 96% and 88% of measurements from outside the calibration set, while rejecting 15% of in-range measurements. Coverage on accepted measurements remained at 0.935 and 0.919. Signed-rank tests over 20 seeds rejected equal width against the unscaled construction as well as against the deep ensemble at <span><math><mrow><mi>p</mi><mo><</mo><mn>1</mn><msup><mrow><mn>0</mn></mrow><mrow><mo>−</mo><mn>5</mn></mrow></msup></mrow></math></span>, while coverage showed no departure from nominal. A nested selection returned coverage of 0.905 and 0.910. Evidence covers prepared standards in a defined electrolyte rather than environmental samples. An operator therefore receives a concentration with a stated range, or no number.</div></div>","PeriodicalId":9749,"journal":{"name":"Chemical Engineering Journal Advances","volume":"28 ","pages":"Article 101522"},"PeriodicalIF":6.9,"publicationDate":"2026-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149346884","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Sina Habibi, Mahsa Mehravaran, Karim Asadpour-Zeynali
{"title":"High current density in alkaline water splitting process at the stainless-steel electrode modified with ZIF-L/MgBiO3 nanocomposite","authors":"Sina Habibi, Mahsa Mehravaran, Karim Asadpour-Zeynali","doi":"10.1016/j.ceja.2026.101531","DOIUrl":"10.1016/j.ceja.2026.101531","url":null,"abstract":"<div><div>The current energy demand spurred attempts to cleanly produce hydrogen by electrochemical water splitting. Hence, developing cost-effective electrocatalysts capable of producing high current density at low overpotentials for the oxygen evolution reaction (OER) and the hydrogen evolution reaction (HER) plays a critical role in transforming the laboratory studies of the water splitting process into industrial scales. In this work, we have prepared ZIF-L/MgBiO<sub>3</sub> nanocomposite on the stainless steel (SS) as a cost-effective, and robust substrate to create an effective electrocatalyst for water splitting process. Additionally, several electrochemical studies show that ZIF-L/MgBiO<sub>3</sub>(0.5)/SS outperforms ZIF-L/MgBiO<sub>3</sub>(0.1)/SS, ZIF-L/MgBiO<sub>3</sub>(0.7)/SS, ZIF-L/SS, MgBiO<sub>3</sub>/SS, and bare SS in OER and HER. The results indicate that the ZIF-L/MgBiO<sub>3</sub>(0.5)/SS electrocatalyst required overpotentials of only 180 mV for OER and 225 mV for HER to attain a current density of 10 mA cm<sup>−2</sup>. The efficient electrocatalytic activity of the ZIF-L/MgBiO<sub>3</sub>(0.5)/SS in OER and HER can be due to the complementary effect of combining the two components of the MgBiO<sub>3</sub> perovskite and ZIF-L framework in the nanocomposite. Moreover, linear sweep voltammetry (LSV) tests in different temperatures show that the ZIF-L/MgBiO<sub>3</sub>(0.5)/SS is capable to generating high current density even at an upper temperature (60 °C). Furthermore, via implementing a two-electrode configuration by the ZIF-L/MgBiO<sub>3</sub>(0.5)/SS electrodes, the full-cell water splitting exhibits a potential of 1.65 V to accomplish a current density of 100 mA cm<sup>−2</sup> or 1.41 V at a current density of 10 mA cm<sup>−2</sup>. The characterization analyses confirm that the ZIF-L/MgBiO<sub>3</sub>(0.5)/SS exhibits good structural stability following a 30-hour chronoamperometry test.</div></div>","PeriodicalId":9749,"journal":{"name":"Chemical Engineering Journal Advances","volume":"28 ","pages":"Article 101531"},"PeriodicalIF":6.9,"publicationDate":"2026-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149347423","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Abdulhalim Musa Abubakar, Jerry Yakubu, Rashid Shamsuddin, Moses NyoTonglo Arowo, Suleiman A. Wali
{"title":"Optimization of biodiesel production from palm fatty acid distillate (PFAD) and synthesized Bambara nutshell catalyst using response surface methodology (RSM)","authors":"Abdulhalim Musa Abubakar, Jerry Yakubu, Rashid Shamsuddin, Moses NyoTonglo Arowo, Suleiman A. Wali","doi":"10.1016/j.ceja.2026.101542","DOIUrl":"10.1016/j.ceja.2026.101542","url":null,"abstract":"<div><div>This study investigated the optimization of biodiesel yield (BDY) from palm fatty acid distillate (PFAD) using a synthesized Bambara nutshell catalyst (BNC) through a single-step esterification process and Response Surface Methodology (RSM) based on a Central Composite Design (CCD). The effects of methanol-to-oil molar ratio (8:1–12:1), catalyst loading (1–5 wt.%), and reaction time (1–4 h) on BDY were evaluated. The synthesized BNC was comprehensively characterized using Fourier Transform Infrared Spectroscopy (FTIR), Scanning Electron Microscopy (SEM), Brunauer–Emmett–Teller (BET) surface area and pore structure analyzes, as well as total acid density (TAD) determination to establish the relationship between catalyst properties and biodiesel recovery. FTIR analysis revealed the presence of hydroxyl, oxygenated, aromatic and metal-associated functional groups that support catalytic activity, whereas SEM micrographs showed a rough and porous morphology favorable for reactant diffusion. BET analysis indicated a relatively high surface area of 459.60 m<sup>2</sup>/g, due to the abundant accessible catalytic sites present. RSM results demonstrated that methanol-to-oil ratio, catalyst loading, and reaction time exerted positive effects on BDY, where the optimum values occurred at approximately 12:1 methanol-to-oil ratio, 5.0 wt.% catalyst loading and 4 h reaction time. Optimum predicted BDY obtained is 93.08% close to the experimental value of 94.65% (Run 14). Model adequacy evaluation through Box–Cox transformation assessment, residual diagnostics, normal probability plots, and predicted-versus-actual comparisons confirmed the robustness of the developed quadratic model. The low residual magnitude, conformity to normality, absence of outliers, and close agreement between actual and predicted yields provide strong evidence that the model was statistically adequate for BDY prediction and optimization. In accordance with the study, BNC is a promising plant-derived heterogeneous catalyst capable of replacing conventional sulfuric acid catalysts for PFAD conversion, in support of a more sustainable and environmentally benign biodiesel production.</div></div>","PeriodicalId":9749,"journal":{"name":"Chemical Engineering Journal Advances","volume":"28 ","pages":"Article 101542"},"PeriodicalIF":6.9,"publicationDate":"2026-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149346885","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Comparative machine learning models for predicting reactive blue 19 degradation in ultraviolet light emitting diode/chlorine process","authors":"Abolfazl Soleymani, Gholamreza Asadollahfardi, Alireza Gholizadeh, Reza Rezaei","doi":"10.1016/j.ceja.2026.101539","DOIUrl":"10.1016/j.ceja.2026.101539","url":null,"abstract":"<div><div>UV-LED/chlorine AOP effectively removes persistent pollutants, but its nonlinear behavior complicates accurate degradation prediction. This study develops an integrated experimental–machine learning (ML) framework to predict the degradation efficiency of the textile dye Reactive Blue 19 (RB19) during UV-LED/chlorine oxidation. A dataset of 467 experimental samples covering a broad range of operational conditions, interfering anions, and radical scavengers was used to train and evaluate five ML models: Multiple Linear Regression (MLR), K-Nearest Neighbors (KNN), Random Forest (RF), Extreme Gradient Boosting (XGB), and Artificial Neural Network (ANN). Model performance was assessed using R², RMSE, MAE, and Index of Agreement (IA), while feature importance and SHapley Additive exPlanations (SHAP) were employed to interpret model predictions. Among the tested models, XGB achieved the best performance, with a test R² of 0.9899, RMSE of 3.24%, MAE of 2.17%, and IA of 0.9974, indicating excellent predictive capability without overfitting. UV-LED fluence (23.6%) and reaction time (20.8%) had the highest XGB feature-importance values, followed by sodium sulfite (13.4%) and pH (13.2%). SHAP analysis indicated that higher values of irradiation-related parameters were generally associated with higher predicted degradation efficiency, whereas higher initial dye concentration and sulfite concentration were associated with lower predicted degradation efficiency. Overall, XGB provides a reliable tool for process optimization, control, and reducing experimental effort in textile wastewater treatment.</div></div>","PeriodicalId":9749,"journal":{"name":"Chemical Engineering Journal Advances","volume":"28 ","pages":"Article 101539"},"PeriodicalIF":6.9,"publicationDate":"2026-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149346886","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Matteo Tollemeto, Sonia Sarnelli, Laura de Vittorio, Marta Rubio-Huertas, Lucia Baldino, Anja Boisen
{"title":"Supercritical CO<sub>2</sub> engineered niosomes as functional materials for oral drug delivery.","authors":"Matteo Tollemeto, Sonia Sarnelli, Laura de Vittorio, Marta Rubio-Huertas, Lucia Baldino, Anja Boisen","doi":"10.1016/j.ceja.2026.101357","DOIUrl":"10.1016/j.ceja.2026.101357","url":null,"abstract":"<p><p>Achieving scalable and controlled production of nanovesicles remains a major bottleneck in translating nanocarrier systems to industry. Conventional batch methods, such as thin-film hydration and ethanol injection, often show low reproducibility, residual solvents, and multistep workflows that hinder scale-up. To address these challenges, we developed SuperSomes, a supercritical CO<sub>2</sub> (SC-CO<sub>2</sub>)-based process enabling continuous, solvent-free fabrication of nanovesicles. The method relies on atomization of an aqueous phase within a gas-expanded medium, promoting rapid self-assembly of surfactant or lipid molecules under mild conditions and allowing precise control over vesicle size and composition. As a proof of concept, we used SuperSomes to generate Span 80/Tween 80 or Span 80/Capric acid (C10) niosomes for oral drug delivery. By systematically tuning surfactant ratios, we identified how hydrophilic-lipophilic balance and packing parameters govern vesicle formation and stability. The resulting niosomes showed high resistance to pH and temperature variations. <i>In vitro</i> Caco-2 studies revealed enhanced intracellular uptake and paracellular transport, indicating a dual absorption mechanism. <i>In vivo</i>, orally administered fluorescent niosomes exhibited minimal gastric retention and improved intestinal permeation. Overall, SuperSomes offers a reproducible, scalable, and environmentally friendly platform for producing nanocarriers tailored for oral delivery, overcoming key limitations of conventional manufacturing approaches.</p>","PeriodicalId":9749,"journal":{"name":"Chemical Engineering Journal Advances","volume":"27 ","pages":"None"},"PeriodicalIF":6.9,"publicationDate":"2026-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13545070/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148896544","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
A.Annam Renita , J Aravind Kumar , Anstey Vathani T , Sai Sruthi V , Sathish T , Basem Abu Zneid
{"title":"Green valorisation of waste cooking oil into biodiesel using copper metal-organic framework (Cu-MOF) catalyst in a packed-bed continuous process","authors":"A.Annam Renita , J Aravind Kumar , Anstey Vathani T , Sai Sruthi V , Sathish T , Basem Abu Zneid","doi":"10.1016/j.ceja.2026.101056","DOIUrl":"10.1016/j.ceja.2026.101056","url":null,"abstract":"<div><div>Waste cooking oil (WCO) has the benefit of conversion to biodiesel, which has both ecological and economic benefits. This paper presents a copper-based metal-organic framework (Cu-MOF) catalyst that was used in the WCO transesterification reaction and then the separation of the product in a packed bed column. The thermogravimetric analysis was used in the determination of catalyst stability and showed a degradation temperature of 250°C. Other structural and morphological characterisation was done through FESEM, FTIR, and XRD. Optimisation of the process was conducted based on the response surface using a central composite design that showed a maximum biodiesel yield of 88% when the ratio of methanol to oil was 6:1 and 5.28 g of the Cu-MOF was used. Sustained activity in reusability tests was demonstrated up to the tenth cycle, with a 70% yield observed until the yield dropped. The results indicate high catalytic efficiency and recyclability of Cu-MOF, which has a strong potential for producing sustainable biodiesel by using WCO.</div></div>","PeriodicalId":9749,"journal":{"name":"Chemical Engineering Journal Advances","volume":"26 ","pages":"Article 101056"},"PeriodicalIF":7.1,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146172095","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"High-performance boron nitride-halloysite nanocomposite as a recyclable adsorbent for methylene blue removal","authors":"Shirin Daneshnia , Majid Abdouss , Sadegh Sahraei , Zohreh Arefi Khorrami , Rasoul Esmaeely Neisiany","doi":"10.1016/j.ceja.2026.101058","DOIUrl":"10.1016/j.ceja.2026.101058","url":null,"abstract":"<div><div>The development of efficient and renewable adsorbents for industrial wastewater decolorization remains a critical environmental challenge. In this study, we report the synthesis of a novel boron nitride-halloysite (h-BN-HNT) nanocomposite via a facile solution mixing method and its outstanding performance in removing methylene blue (MB) from aqueous solutions. Morphology, chemical structure, and crystallinity of the nanocomposite were characterized by scanning electron microscopy (SEM), Fourier transform infrared spectroscopy (FTIR), and X-ray diffraction (XRD), confirming successful functionalization and revealing a hierarchical porous structure with high potential for dye adsorption. The nanocomposite material demonstrated excellent performance in water decolorization, achieving complete MB removal within 40 min for all concentrations studied (5–50 mg/L), with a maximum adsorption capacity of 473.6 mg/g at 50 mg/L initial concentration after 40 min. Kinetic studies revealed that the adsorption process follows pseudo-second-order kinetics with excellent correlation (R² = 0.9999, for the initial concentration, 20 mg/l), indicating chemisorption as the dominant mechanism. Statistical error analysis confirmed the model validity with low chi-squared (χ² = 0.93), sum of squared errors (SSE = 339.45), and average relative error (ARE = 0.11 %) values. Notably, the nanocomposite also demonstrated excellent recyclability, maintaining over 95% removal efficiency after five consecutive adsorption-desorption cycles, confirming its practical reusability. These results establish the h-BN-HNT nanocomposite as a highly promising candidate for practical applications in industrial wastewater treatment, offering a sustainable and cost-effective solution for the removal of cationic dyes.</div></div>","PeriodicalId":9749,"journal":{"name":"Chemical Engineering Journal Advances","volume":"26 ","pages":"Article 101058"},"PeriodicalIF":7.1,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146172092","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Balarabe B. Ismail , Hassan A. Saeed , Umar Abdulbaki Danhassan , Songfeng Yu , Yunfeng Pu , Wenjun Wang , Donghong Liu
{"title":"Mechanism of enhanced punicalagin adsorption by ultrasound-chitosan functionalized baobab green adsorbents","authors":"Balarabe B. Ismail , Hassan A. Saeed , Umar Abdulbaki Danhassan , Songfeng Yu , Yunfeng Pu , Wenjun Wang , Donghong Liu","doi":"10.1016/j.ceja.2026.101069","DOIUrl":"10.1016/j.ceja.2026.101069","url":null,"abstract":"<div><div>Developing sustainable low-cost adsorbents with high polyphenol adsorption capacity is crucial for advancing green separation technologies. This study integrates experimental and computational techniques to elucidate the synergistic effects of ultrasonication (US) and chitosan (CS) functionalization on the adsorption performance of baobab fruit shell (BFS) green adsorbents for punicalagin adsorption. Characterization via BET, FTIR, XRD, and SEM-EDS analyzes confirmed that US-CS functionalization increased surface area, pore volume, and accessible functional groups. Modified BFS (MBFS) adsorbents treated under high-intensity short-time (HIST, 687.9 W cm<sup>−2</sup>, 5 min) US treatment showed a three-fold increase in punicalagin adsorption capacity (211.21 mg g<sup>−1</sup>) and rapid adsorption equilibrium within 5 min. US generally intensified mass transfer during adsorption, as evidenced by the significant increase in external mass transfer coefficient from 1.21 × 10<sup>−4</sup> to 2.82 × 10<sup>−4</sup> ms<sup>−1</sup> following HIST treatment. Based on XPS results, hydrogen bonding, hydrophobic, and electrostatic interactions are involved in punicalagin adsorption on MBFS. The pseudo-first-order kinetic and Langmuir isotherm models (R² > 0.99) best described the adsorption mechanism, confirming both physisorption and a propensity for monolayer adsorption on homogeneous surfaces. HIST treatment preserved structural integrity under transient mechanical stress, enabling MBFS to retain over 85 % efficiency after four adsorption-desorption cycles. Machine learning models, particularly XGBoost (R² > 0.95; RMSE = 6.47), accurately predicted the punicalagin adsorption capacity and identified HIST and CS crosslinking as the key variables. Overall, US- and CS-functionalized BFS adsorbents show strong potential as a green and sustainable alternatives for polyphenol purification.</div></div>","PeriodicalId":9749,"journal":{"name":"Chemical Engineering Journal Advances","volume":"26 ","pages":"Article 101069"},"PeriodicalIF":7.1,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146172106","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Domenico Licursi , Vincenzo Russo , Benedetta Bertini , Federica Orabona , Francesco Taddeo , Martino Di Serio , Anna Maria Raspolli Galletti , Claudia Antonetti
{"title":"Sustainable synthesis of p,p'-diphenolic acid from renewable levulinic acid with heterogeneous acid resins","authors":"Domenico Licursi , Vincenzo Russo , Benedetta Bertini , Federica Orabona , Francesco Taddeo , Martino Di Serio , Anna Maria Raspolli Galletti , Claudia Antonetti","doi":"10.1016/j.ceja.2026.101080","DOIUrl":"10.1016/j.ceja.2026.101080","url":null,"abstract":"<div><div>Diphenolic acid (<em>p,p’</em>-DPA) is a very interesting substitute for bisphenol A. In this work, the optimization of its synthesis from phenol and bio-based levulinic acid was proposed, comparing homogeneous (e.g. mineral acids) and heterogeneous catalysts (commercial acid resins), the latter being preferred due to their greater industrial relevance. Some unresolved bottlenecks still hinder <em>p,p’</em>-DPA industrial development, in particular 1) lack of systematic optimization studies; 2) preference for lab-scale synthesized catalysts, rather than commercial ones, the latter industrially available and well-characterized; 3) sporadic evaluations on catalyst recycling; 4) undeveloped purification of the product from the reaction medium with non-toxic solvents; 5) complete lack of kinetic studies on heterogeneous systems. The present work aims to address these issues, optimizing the <em>p,p’</em>-DPA synthesis with commercial acid resins. Aquivion P87S was identified as the best heterogeneous catalyst, with the highest yield and selectivity to <em>p,p'</em>-DPA (both 91.7%), as well as good stability over time (4 cycles), all remarkable achievements, well beyond those of the state of the art. Ethanol was identified as the most compatible solvent with Aquivion P87S, enabling its effective swelling and mild regeneration. For the intensification development, <em>p,p'</em>-DPA purification was successfully achieved (purity degree of 94%), developing a new sustainable procedure with water as the crystallization solvent. Green metrics and LCA energy analysis demonstrated the potential sustainability of this process. The activation energies of all the tested resins were calculated with the Arrhenius equation and the Weisz Prater criterion was utilized to assess potential internal diffusion limitations within the systems.</div></div>","PeriodicalId":9749,"journal":{"name":"Chemical Engineering Journal Advances","volume":"26 ","pages":"Article 101080"},"PeriodicalIF":7.1,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146172152","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Rizqy Ahsana Putri , Riyanarto Sarno , Wahyu Prasetyo Utomo , Fadlilatul Taufany , Kelly Rossa Sungkono , Taufiq Choirul Amri , Alya Kamilah , Rini Handayani , Sang-Seok Lee , A. Min Tjoa , Arif Abdullah Sagran
{"title":"Cross-Attention Feature Fusion network for robust estimation of Cd2+ and Pb2+ in water samples using Cyclic Voltammetry","authors":"Rizqy Ahsana Putri , Riyanarto Sarno , Wahyu Prasetyo Utomo , Fadlilatul Taufany , Kelly Rossa Sungkono , Taufiq Choirul Amri , Alya Kamilah , Rini Handayani , Sang-Seok Lee , A. Min Tjoa , Arif Abdullah Sagran","doi":"10.1016/j.ceja.2026.101065","DOIUrl":"10.1016/j.ceja.2026.101065","url":null,"abstract":"<div><div>Voltammetry is a promising technique for estimating heavy metal pollution such as Cadmium (Cd<span><math><msup><mrow></mrow><mrow><mn>2</mn><mo>+</mo></mrow></msup></math></span>) and Lead (Pb<span><math><msup><mrow></mrow><mrow><mn>2</mn><mo>+</mo></mrow></msup></math></span>) in water. Its advantages include rapid analysis and cost-effectiveness over established methods like Atomic Absorption Spectroscopy (AAS) and Inductively Coupled Plasma - Mass Spectrometry (ICP-MS). However, current analysis often depends only on peak data, ignoring the rest of the voltammetric signal which may contain useful information that could potentially improve measurement accuracy. To address this limitation, the Cross-Attention Feature Fusion (CAFF) network is proposed to analyze Cyclic Voltammetry (CV) signals acquired using a 3-electrode setup with a Glassy Carbon Electrode (GCE) as the working electrode, Platinum as the counter, and Ag/AgCl as the reference. Unlike standard self-attention mechanisms or simple concatenation fusion methods, CAFF introduces a novel dual-stream architecture that dynamically captures the inter-dependencies between raw CV signals and extracted peak data—an approach previously unexplored in electrochemical sensing. The model integrates an Improved Beluga Whale Optimization (IBWO) algorithm that automatically determines the optimal hyperparameters, resulting in a more robust model. Robustness was assessed using Chemically-Informed Degradation Simulation (CIDS). As a result, the proposed CAFF-IBWO model demonstrated superior performance, achieving R<span><math><msup><mrow></mrow><mrow><mn>2</mn></mrow></msup></math></span> values of 0.97 for Cd<span><math><msup><mrow></mrow><mrow><mn>2</mn><mo>+</mo></mrow></msup></math></span> and 1.00 for Pb<span><math><msup><mrow></mrow><mrow><mn>2</mn><mo>+</mo></mrow></msup></math></span>. It also significantly reduced the Mean Absolute Percentage Error (MAPE) by 65.79% for Cd<span><math><msup><mrow></mrow><mrow><mn>2</mn><mo>+</mo></mrow></msup></math></span> and 72.50% for Pb<span><math><msup><mrow></mrow><mrow><mn>2</mn><mo>+</mo></mrow></msup></math></span> compared to single-input attention networks. Furthermore, CAFF-IBWO exhibited remarkable resilience against signal degradation, maintaining stable prediction performance across varying noise conditions. While the study focuses specifically on Cd<span><math><msup><mrow></mrow><mrow><mn>2</mn><mo>+</mo></mrow></msup></math></span> and Pb<span><math><msup><mrow></mrow><mrow><mn>2</mn><mo>+</mo></mrow></msup></math></span> and requires further validation for broader generalization, the demonstrated performance is highly promising. These findings underscore the model’s potential for real-world environmental sensing applications.</div></div>","PeriodicalId":9749,"journal":{"name":"Chemical Engineering Journal Advances","volume":"26 ","pages":"Article 101065"},"PeriodicalIF":7.1,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146098718","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}