Hydrous ferric oxide (HFO) is most effective with high treatment capacity on arsenate [As(V)] sorption although its transformation and aggregation nature need further improvement. Here, HFO nanoparticles with carboxymethyl cellulose (CMC) or starch as modifier was synthesized for the purpose of stability improvement and As(V) removal from water. Comparatively, CMC might be the optimum stabilizer for HFO nanoparticles because of more effective physical and chemical stability. The large-pore structure, high surface specific area, and the non-aggregated nature of CMC-HFO lead to increased adsorption sites, and thus high adsorption capacities of As(V) without pre-treatment (355 mg·g⁻¹), which is much greater than those reported in previous studies. Second-order equation and dual-mode isotherm model could be successfully used to interpret the sorption kinetics and isotherms of As(V), respectively. FTIR, XPS and XRD analyses suggested that precipitation and surface complexation were primary mechanisms for As(V) removal by CMC modified HFO nanoparticles. A surface complexation model (SCM) was used to simulate As adsorption over pH 2.5–10.4. The predominant adsorbed arsenate species were modeled as bidentate binuclear surface complexes at low pH and as monodentate complexes at high pH. The immobilized arsenic remained stable when aging for 270 d at room temperature.

Toxic heavy metals are ubiquitous in the aquatic environment and show a significant danger to human health. Carbon nanotubes have been extensively used in treating the contamination of groundwater due to their porous multi-layer nature. Batch tests revealed that oxidized multiwalled carbon nanotubes (O-MWCNT_S) offer better removal of Pb(II). The removal rate of Pb(II) was 90.15% at pH 6 within 24 h, which was ~58% more than that of Cd(II). The removal rate decreased to 55.59% for Pb(II) and to 16.68% for Cd(II) when the initial concentration of Pb(II)/Cd(II) ranged from 5 to 15 mg·g⁻¹. The removal rate in the competitive tests was about 60.46% for Pb(II) and 9.70% for Cd(II). The Langmuir model offered better description of the adsorptive data for both ions. And the Q_m of Pb(II) was 5.73 mg·g⁻¹, which was 2.39 mg·g⁻¹ more than that of Cd(II) in a single-icon system, while Q_m was 7.11 mg·g⁻¹ with Pb(II) and 0.78 mg·g⁻¹ with Cd(II) in competitive water. And thermodynamic tests further indicated that the activating energy of Pb(II) and Cd(II) was 83.68 and 172.88 kJ·mol⁻¹, respectively. Lead and cadmium adsorbed on the surface of O-MWCNT_S are antagonistic in the competitive system. Based on XPS analyses, it was concluded that the absorbed lead/cadmium species on O-MWCNT_S were (-COO)₂Pb, (-COO)Pb(-O)/(-COO)₂Cd, and (-COO)Cd(-O). Additionally, they offered theoretical evidence supporting the practicality of using nanocomposite membranes as a means to remove cadmium and lead.

With the rapid advancement of deep learning technology, the residual networks technique (ResNet) has made significant strides in the field of image processing, and its application in soil science has been steadily increasing. ResNet outperforms traditional methods by effectively mitigating the vanishing gradient problem, enabling deeper network training, enhancing feature extraction, and improving accuracy in complex pattern recognition tasks. ResNet, as an efficient deep learning model, can automatically extract features from complex soil image data, enabling accurate soil classification and assessment of soil health. Recent research is increasingly applying ResNet to various fields, including soil type classification and health assessment. Firstly, this manuscript outlines various methods for collecting soil data, highlighting the significance of employing diverse data sources to comprehensively understand soil characteristics. These methods include the acquisition of soil microscopic images, which provide high-resolution insights into the soil’s particulate structure at the cellular level; remote sensing images, which offer valuable information regarding large-scale soil properties and spatial variations through satellite or drone-based technologies; and high-definition images, which capture fine-scale details of soil features, enabling more precise and detailed analysis. By integrating these techniques, a solid foundation is established for subsequent soil image analysis, thereby enhancing the accuracy of soil classification, health assessments, and environmental impact evaluations. Furthermore, this approach contributes to advancements in precision agriculture, land use planning, soil erosion monitoring, and contamination detection, ultimately supporting sustainable soil management and ecological conservation efforts. Then, the advantages of using ResNet in soil science are analyzed, and its performance across different soil image processing tasks is explored. Finally, potential future development directions are proposed.