Predicting hazardous Sb(III)/Sb(V) adsorption on biochar via an interpretable ML-DFT integrated BiMeSorb framework with experimental validation

Abstract

Antimony (Sb) is a persistent and highly toxic contaminant. Its environmental risk is dictated by its redox state, as Sb(III) and Sb(V) exhibit vastly different adsorption behaviors. Although previous machine learning (ML) studies have investigated Sb adsorption, this speciation-dependent behavior still hinders the rational design of biochars for effective Sb immobilization. Here, we establish a data-driven and mechanism-informed framework (BiMeSorb) that integrates interpretable ML with density functional theory (DFT) to quantitatively resolve interactions between Sb(III)/Sb(V) and biochars. Built on a curated database of 437 data points, our Gradient Boosting Decision Tree model achieves high predictive accuracy for adsorption capacity (test R² = 0.934). Interpretable ML analyses (SHAP and PDP) reveal that oxygen-containing functional groups, specific surface area, and Sb speciation dominate adsorption, while appropriate initial Sb concentration and adsorbent dosage are critical operational conditions for achieving high adsorption performance. DFT calculations confirm that Sb(III) and Sb(V) interact strongly with carboxyl and hydroxyl groups via hydrogen bonding, exhibiting distinct binding energetics. Targeted adsorption experiments with Fe-modified biochars further validated the ML-identified descriptor-performance relationships. By integrating prediction, mechanism, and validation, the BiMeSorb framework provides a quantitative and transferable strategy for rational design of biochar adsorbents to improve aqueous Sb adsorption performance under tested experimental conditions.

Graphical abstract

Keywords

Antimony / Biochar / Interpretable machine learning / Adsorption behavior / Density functional theory

Highlight

	● BiMeSorb links ML and DFT to study Sb adsorption on biochar.
	● GBDT predicts Sb adsorption with high accuracy (test R 2 = 0.934).
	● SHAP analysis highlights the importance of experimental conditions.
	● Highest Sb uptake occurs at pH = 5–7 and IHM/BC of 50–100 mg/g.
	● Fe-modified biochars support the predicted adsorption patterns.

References

[1] Alsulamy S . (2025). Predicting construction delay risks in Saudi Arabian projects: a comparative analysis of CatBoost, XGBoost, and LGBM. Expert Systems with Applications, 268: 126268

[2] Bai Y H , Jefferson W A , Liang J S , Yang T T , Qu J H . (2017). Antimony oxidation and adsorption by in-situ formed biogenic Mn oxide and Fe-Mn oxides. Journal of Environmental Sciences, 54: 126–134

[3] Bolan N , Kumar M , Singh E , Kumar A , Singh L , Kumar S , Keerthanan S , Hoang S A , El-Naggar A , Vithanage M . et al. (2022). Antimony contamination and its risk management in complex environmental settings: a review. Environment Inter-national, 158: 106908

[4] Chen J J , Li L P , Lv X W , Liu T , Zhang J , Luo L W , Tian Y . (2026a). Mechanistic insights into sulfamethoxazole removal in activated sludge systems through machine learning and microcosm experiments. Water Research, 292: 125331

[5] Chen Y L , Liu L , Gao D W . (2026b). Machine learning for small-data in aquatic environments: a review of challenges, methods, and optimization approaches. Engineering Environment, 20(6): 86

[6] Cheng J C , Hua X L , Zhang G H , Yu M Q , Wang Z , Zhang Y L , Liu W , Chen Y J , Wang H M , Luo Y D . et al. (2024). Synthesis of high-crystallinity Zeolite A from rare earth tailings: investigating adsorption performance on typical pollutants in rare earth mines. Journal of Hazardous Materials, 468: 133730

[7] Cheng Q , Li Q R , Huang X J , Li X Q , Wang Y Y , Liu W Z , Lin Z . (2022). The high efficient Sb(III) removal by cauliflower like amorphous nanoscale zero-valent iron (A-nZVI). Journal of Hazardous Materials, 436: 129056

[8] Dai Y J , Zhang N X , Xing C M , Cui Q X , Sun Q Y . (2019). The adsorption, regeneration and engineering applications of biochar for removal organic pollutants: a review. Chemosphere, 223: 12–27

[9] Delley B . (2000). From molecules to solids with the DMol³ approach. The Journal of Chemical Physics, 113(18): 7756–7764

[10] Duan Q N , Yan P W , Feng Y C , Wan Q R , Zhu X L . (2024). Machine learning assisted adsorption performance evaluation of biochar on heavy metal. Frontiers of Environmental Science & Engineering, 18(5): 55

[11] Fakhar A , Galgo S J C , Canatoy R C , Rafique M , Sarfraz R , Farooque A A , Khan M I . (2025). Advancing modified biochar for sustainable agriculture: a comprehensive review on charac-terization, analysis, and soil performance. Biochar, 7(1): 8

[12] Fan L , Wang D X , Yu H L , Gong Z , He Y , Guo J Y . (2025a). Application of machine learning to predict the fluoride removal capability of MgO. Journal of Environmental Chemical Engineering, 13(1): 115317

[13] Fan W , Shen X L , Zhu Z L , Liu X C , Zhang H , Qiu Y L , Yin D Q . (2025b). Research progress on biochar materials for new pollutants removal in the aquatic environment: a mini-review. Frontiers of Environmental Science & Engineering, 19(3): 33

[14] Fu W L , Yao X , Zhang L S , Zhou J E , Zhang X Y , Yuan T , Lv S Y , Yang P , Fu K R , Huo Y Q . et al. (2025). Design optimization of bimetal-modified biochar for enhanced phosphate removal performance in livestock wastewater using machine learning. Bioresource Technology, 418: 131898

[15] Gao B R , Wu L R , Li S K , Yang J J , Dou M M , Chang G W , Li X . (2025). Rapid capture of perfluorooctanoic acid and perfluorooctane sulfonate at environmentally relevant concen-trations via the ‘mesh trap’ of triazine-based polymer network: mechanism and photocatalytic regeneration. Journal of Hazardous Materials, 483: 136698

[16] Gu X , Guo P H , Li Z S , Xu X , Cao Y N , Yang G E , Kuang C T , Li X G , Qing Y , Wu Y Q . (2024). A multifunctional coconut shell biochar modified by titanium dioxide for heavy metal removal in water/soil and tetracycline degradation. Journal of Cleaner Production, 482: 144192

[17] Han B , Butterly C , Zhang W , He J Z , Chen D L . (2021). Adsorbent materials for ammonium and ammonia removal: a review. Journal of Cleaner Production, 283: 124611

[18] Hassan M , Wang B , Wu P , Wang S S . (2024). Engineered biochar for in-situ and ex-situ remediation of contaminants from soil and water. Science of the Total Environment, 957: 177384

[19] Hassan R , Behtouei Z , Baghban A . (2025). Advanced machine learning for precise prediction of biochar’s heavy metal sorption efficiency. Journal of Hazardous Materials Advances, 18: 100739

[20] He M S , Zhang Z , Wang M , Liang C Y , Wang H J , Cheng C , Li Y Y , Wang Y K , Zhang Z . (2025). A review of hydroxyapatite synthesis for heavy metal adsorption assisted by machine learning. Journal of Hazardous Materials, 481: 136525

[21] Hu J , Yin R , Pan Y , Wang J F , Ren H Q . (2025). Multimodal learning-assisted identification of effluent water quality and toxicity in wastewater treatment plants. Environmental Science & Technology, 59(27): 13857–13869

[22] Hu X J , Huo J Y , Yue J Y , Hu Z , Xie H J , Zhang J . (2026). Substrate-mediated trade-offs between pollutant removal efficiency and ecological risks in constructed wetlands treating multi-antibiotic wastewater. Water Research, 293: 125414

[23] Huang J S , Ashraf W M , Ansar T , Abbas M M , Tlija M , Tang Y Y , Guo Y X , Zhang W . (2025). Optimisation led energy-efficient arsenite and arsenate adsorption on various materials with machine learning. Water Research, 271: 122815

[24] Huang J Y , Zheng Y X , Luo L B , Feng Y , Zhang C L , Wang X , Liu X Y . (2016). Facile preparation of highly hydrophilic, recyclable high-performance polyimide adsorbents for the removal of heavy metal ions. Journal of Hazardous Materials, 306: 210–219

[25] Huo Z Y , Wang X X , Huang X , Elimelech M . (2024). Intensifying electrified flow-through water treatment technologies via local environment modification. Frontiers of Environmental Science & Engineering, 18(6): 69

[26] Jia X C , Zhou J W , Liu J , Liu P , Yu L , Wen B , Feng Y . (2020). The antimony sorption and transport mechanisms in removal experiment by Mn-coated biochar. Science of the Total Environment, 724: 138158

[27] Joshi M , Bhatt D , Srivastava A . (2023). Enhanced adsorption efficiency through biochar modification: a comprehensive review. Industrial & Engineering Chemistry Research, 62(35): 13748–13761

[28] Kang K , Kim Y . (2025). Solubility prediction of edible insect proteins: a comparative study of machine learning and response surface methodology. Food Chemistry, 488: 144856

[29] Lai L , Liu X T , Ren W B , Zhou Z , Zhao X W , Zeng X F , Lin C Y , He M C , Ouyang W . (2023). Efficient removal of Sb(Ⅲ) from water using β-FeOOH-modified biochar: synthesis, performance and mechanism. Chemosphere, 311(Pt 1): 137057

[30] Lan J, Palizhati A, Shuaibi M, Wood B M, Wander B, Das A, Uyttendaele M, Zitnick C L, Ulissi Z W (2023). AdsorbML: a leap in efficiency for adsorption energy calculations using generalizable machine learning potentials. npj Computational Materials, 9(1): 172

[31] Li K, Ma S L, Xu S J, Fu H C, Li Z Q, Li Y, Liu S L, Du J G (2021). The mechanism changes during bisphenol A degradation in three iron functionalized biochar/peroxymonosulfate systems: the crucial roles of iron contents and graphitized carbon layers. Journal of Hazardous Materials, 404(Pt A): 124145

[32] Lin Y H , Wu Y L , Jiang X J , Wang Q , Liu Y B , Jin L M . (2026). Advancing bidirectional hydrogen spillover in heterostructure catalysts for electrocatalytic antimony removal and recovery from wastewater. Journal of Hazardous Materials, 501: 140662

[33] Liu C, Balasubramanian P, An J X, Li F Y (2025). Machine learning prediction of ammonia nitrogen adsorption on biochar with model evaluation and optimization. npj Clean Water, 8(1): 13

[34] Liu X , Zhou J W , Zhou W Q , Feng Y , Finfrock Y Z , Liu Y Y , Liu P , Shu C . (2021). Efficiency and mechanisms of Sb(III/V) removal by Fe-modified biochars using X-ray absorption spectroscopy. Journal of Environmental Chemical Engineering, 9(6): 106741

[35] Lu J Y , Wu X Y , Yi Z R , Xiang K , Mahmud S , Lu Z , Lian D H , Wang W T , Feng Y M , Wang Z . et al. (2025). Robust β-FeOOH@Polytetrafluoroethylene nanocomposite membrane with synergistic adsorption–photocatalysis for Antimony(III) removal from wastewater. ACS Applied Nano Materials, 8(44): 21238–21249

[36] Lundberg S M , Erion G , Chen H , Degrave A , Prutkin J M , Nair B , Katz R , Himmelfarb J , Bansal N , Lee S I . (2020). From local explanations to global understanding with explainable AI for trees. Nature Machine Intelligence, 2(1): 56–67

[37] Luo D , Wang L Y , Nan H Y , Cao Y J , Wang H , Kumar T V , Wang C Q . (2023). Phosphorus adsorption by functionalized biochar: a review. Environmental Chemistry Letters, 21(1): 497–524

[38] Lyu H , Xu Z M , Zhong J , Gao W H , Liu J X , Duan M . (2024). Machine learning-driven prediction of phosphorus adsorption capacity of biochar: insights for adsorbent design and process optimization. Journal of Environmental Management, 369: 122405

[39] Mei Y C , Zhuang S T , Wang J L . (2025). Adsorption of heavy metals by biochar in aqueous solution: a review. Science of the Total Environment, 968: 178898

[40] Mubarak M F , Mohamed A M G , Keshawy M , Elmoghny T A , Shehata N . (2022). Adsorption of heavy metals and hardness ions from groundwater onto modified zeolite: batch and column studies. Alexandria Engineering Journal, 61(6): 4189–4207

[41] Pinto M J , Sharath S S , Sudhakar K , Priya S S , Thirunavukkarasu I . (2025). Predictive modeling and feature attribution of CO₂ adsorption on LDH-derived materials using machine learning approach. Results in Engineering, 28: 108225

[42] Piros P , Ferenci T , Fleiner R , Andréka P , Fujita H , Főző L , Kovács L , Jánosi A . (2019). Comparing machine learning and regression models for mortality prediction based on the Hungarian Myocardial Infarction Registry. Knowledge-Based Systems, 179: 1–7

[43] Qiu Y W , Xu X Y , Xu Z B , Liang J , Yu Y L , Cao X D . (2020). Contribution of different iron species in the iron-biochar composites to sorption and degradation of two dyes with varying properties. Chemical Engineering Journal, 389: 124471

[44] Rajput P , Yadav S , Liu C , Balasubramanian P . (2025). Predicting biochar adsorption capacity for methylene blue removal using machine learning. Journal of Water Process Engineering, 69: 106749

[45] Russo S , Besmer M D , Blumensaat F , Bouffard D , Disch A , Hammes F , Hess A , Lürig M , Matthews B , Minaudo C . et al. (2021). The value of human data annotation for machine learning based anomaly detection in environmental systems. Water Research, 206: 117695

[46] Safeer R, Liu G J, Yousaf B, Ashraf A, Haider M I S, Cheema A I, Ijaz S, Rashid A, Sikandar A, Pikoń K (2024). Insights into the biogeochemical transformation, environmental impacts and biochar-based soil decontamination of antimony. Environmental Research, 251(Pt 2): 118645

[47] Shen X Y , Ma S Y , Xu S Q . (2025). Adsorption performance and mechanism of low antimonate concentrations in water by zirconium-modified biochar. ACS Omega, 10(27): 29098–29108

[48] Shen Y K , Zhao E C , Zhang W , Baccarelli A A , Gao F . (2022). Predicting pesticide dissipation half-life intervals in plants with machine learning models. Journal of Hazardous Materials, 436: 129177

[49] Singh J , Jadeja R . (2025). Recent advances in agricultural waste derived magnetic biochar for removal of heavy metal ions: mechanistic insights and technological innovation. Journal of Molecular Structure, 1325: 141005

[50] Song X J , Zhao J , Jia M K , Yin R , Ren H Q . (2026). Elucidating the role of reactive nitrogen species in micropollutant degradation using machine learning. Water Research, 292: 125345

[51] Thiyagalingam J , Shankar M , Fox G , Hey T . (2022). Scientific machine learning benchmarks. Nature Reviews Physics, 4(6): 413–420

[52] Ullah H , Khan S , Zhu X Y , Chen B L , Rao Z P , Wu N C , Idris A M . (2025). Machine learning-aided biochar design for the adsorptive removal of emerging inorganic pollutants in water. Separation and Purification Technology, 362: 131421

[53] van den Brink S , Kleijn R , Sprecher B , Mancheri N , Tukker A . (2022). Resilience in the antimony supply chain. Resources, Conservation and Recycling, 186: 106586

[54] Wang B Y , Lan J M , Bo C M , Gong B L , Ou J J . (2023). Adsorption of heavy metal onto biomass-derived activated carbon: review. RSC Advances, 13(7): 4275–4302

[55] Wei M , Wang B , Wu P , Zhang X Y , Chen M , Wang S S . (2024). Electrolytic manganese residue-biochar composite for simul-taneous removal of antimony and arsenic from water: adsorption performance and mechanisms. Journal of Cleaner Production, 437: 140623

[56] Wu X T , Zhou Q X , Mu L , Hu X G . (2022). Machine learning in the identification, prediction and exploration of environmental toxicology: challenges and perspectives. Journal of Hazardous Materials, 438: 129487

[57] Xiang Z Q , Huang C K , Huang J C , Yan Y , Liu G B , Yu X M , Liu W J , Cao H J , Liu A Z . (2024). Mechanism for the synergistic removal of Sb(III) and Sb(V) from printing and dyeing wastewater by polyferric sulfate. Journal of Environmental Chemical Engineering, 12(2): 112492

[58] Xiao W Z , Wang Z X , Gui Y G . (2024). Adsorption properties of metal atom (Co, V, W, Zr)-modified MoTe₂ for CO, CH₃CHO, and C₆H₆ gases: a DFT study. Molecules, 29(21): 5086

[59] Xiao X , Chen B L . (2017). A direct observation of the fine aromatic clusters and molecular structures of biochars. Environmental Science & Technology, 51(10): 5473–5482

[60] Xie S J , Wang Y J , Zhang R , Huang L , Yan J , Chen Z X , Zhang H G , Li M , Zeng H B . (2026). Fe-N-C-supported LaFeO₃ heterojunction electrodes for efficient copper and fluoride removal in symmetric capacitive deionization. Water Research, 297: 125744

[61] Xu L T , Zhang Z Y , Isah M E , Matsubae K . (2025a). Exploring antimony material flow in the context of energy transition: a scenario-based analysis. Resources, Conservation and Recycling, 222: 108432

[62] Xu S M , Zheng K , Boruntea C R , Cheng D G , Chen F Q , Ye G H , Zhou X G , Coppens M O . (2023). Surface barriers to mass transfer in nanoporous materials for catalysis and separations. Chemical Society Reviews, 52(12): 3991–4005

[63] Xu Z S , Li J , Cheng S B , Rui X , Zhao Y , He H J , Guan H Y , Sharma A , Erxleben M , Chang R Y . et al. (2025b). Deep learning for wildfire risk prediction: integrating remote sensing and environmental data. ISPRS Journal of Photogrammetry and Remote Sensing, 227: 632–677

[64] Xue X T , Yao M , Yu J X , Lu H Y , Liu Y B , Wang H B , Gao H S . (2025). Removal of Sb(V) from textile wastewater with iron-based coagulants: influence factors, precipitate speciation and mechanism analysis. Journal of Water Process Engineering, 79: 108924

[65] Yamde A A , Lade V G , Bindwal A B , Tiwari M S , Birmod R P . (2025). Machine learning approaches for the prediction of hydrogen uptake in metal-organic-frameworks: a comprehensive review. International Journal of Hydrogen Energy, 98: 1131–1154

[66] Yang K , Yang J J , Jiang Y , Wu W H , Lin D H . (2016). Correlations and adsorption mechanisms of aromatic compounds on a high heat temperature treated bamboo biochar. Environmental Pollution, 210: 57–64

[67] Yao B , Li Y X , Zeng W Q , Yang G , Zeng J H , Nie J , Zhou Y Y . (2023). Synergistic adsorption and oxidation of trivalent antimony from groundwater using biochar supported magnesium ferrite: performances and mechanisms. Environmental Pollution, 323: 121318

[68] Yuan L , Wang J L , Zhong Z Q , Li J G , Deng H . (2023). Immobilization of antimony in soil and groundwater using ferro-magnesium bimetallic organic frameworks. Journal of Environmental Sciences, 125: 194–204

[69] Zhang C Q , Wu M Y , Wu K , Li H S , Zhang G S . (2023). Efficient removal of antimonate and antimonite by a novel lanthanum-manganese binary oxide: performance and mechanism. Journal of Hazardous Materials, 442: 130132

[70] Zhang L , Wen J , Li Y F , Chen J L , Ye Y Y , Fu Y Y , Livingood W . (2021). A review of machine learning in building load prediction. Applied Energy, 285: 116452

[71] Zhang L F , Lou S C , Hao X R , Zhang H Y , Zhang W Q , Liu X , Huang J F . (2022b). Highly-porous and excellent-capacity zirconium-chitosan composite with superior Sb(III)/Sb(V) removal performance. Separation and Purification Technology, 303: 122253

[72] Zhang L P , Dong Y , Liu J F , Liu W , Lu Y R , Lin H . (2022a). Promotion of higher synthesis temperature for higher-efficient removal of antimonite and antimonate in aqueous solution by iron-loaded porous biochar. Bioresource Technology, 363: 127889

[73] Zhang Y J , Han B X , Song Q Q , Jin S S , Lu Z , Lu S , Zhang X , Zhang Y T , Li D Y . (2025a). Revealing key structural and operating parameters on salt/dye separation of loose nano-filtration membrane by ensemble machine learning. Journal of Membrane Science, 732: 124274

[74] Zhang Y J , He J L , Li D Y , Wang J , Song Q Q , Wang X L , Zhu J Y , Feng X Q , Duan K , Gao X . et al. (2025b). Evaluation of machine learning applied in membrane-based water desalination: a review. Desalination, 613: 119041

[75] Zhao S Y , Guo J Y , Tang Y , Zhou Y B . (2025). Applications of machine learning in heavy metal adsorption modeling: a review. Separation and Purification Technology, 377: 134168

[76] Zhou W Q , Liu P , Liu X S , Ye Z H , Zhou Z Y , Dong Z C , Zhou J W . (2025). Mineralogical control on antimony mobility via weathering products from waste rock in the Xikuangshan mine, China. Science of the Total Environment, 1004: 180766

[77] Zhou Z H . (2018). A brief introduction to weakly supervised learning. National Science Review, 5(1): 44–53

[78] Zhu H , An Q , Syafika Mohd Nasir A , Babin A , Lucero Saucedo S , Vallenas A , Li L , Baldwin S A , Lau A , Bi X T . (2023). Emerging applications of biochar: a review on techno-environmental-economic aspects. Bioresource Technology, 388: 129745

[79] Zhu Y M , Li Z F , Shen J , Wu K Y , Zhao P P , Wu Z H , Liu Z Q , Yang J G , Liu H , Rensing C . et al. (2022). Toxicity of different forms of antimony to rice plants: photosynthetic electron transfer, gas exchange, photosynthetic efficiency, and carbon assimilation combined with metabolome analysis. Journal of Hazardous Materials, 437: 129433

[80] Zhuang Y L , Haderlein S B , Lutze H V , Sun C , Fink F , Paul A , Spahr S . (2025). Persulfate activation by biochar for trace organic contaminant removal from urban stormwater. Water Research, 284: 123921

RIGHTS & PERMISSIONS

Higher Education Press 2026