Boron is a special pollutant. Because of its relatively small molecular weight, it can freely penetrate the reverse osmosis membrane in the same way that water molecules can in reverse osmosis during seawater desalination, which affects the effluent quality of desalinated seawater. In this study, a new magnetic adsorption material, MNP-NMDG, was synthesized by combining magnetic nanoparticles (MNPs) of Fe3O4 with N-methyl-d-glucamine with a high selectivity to boron, and MNP-NMDG was characterized by scanning electron microscopy (SEM), Fourier transform infrared spectroscopy (FT-IR), and X-ray diffraction (XRD). The adsorption properties of the MNP-NMDG for boron during seawater desalination under static and dynamic conditions was studied from the aspects of pH, adsorbent dosage, adsorption kinetics, and isotherms. The results showed that according to the breakthrough curve of dynamic adsorption, MNP-NMDG had a high boron-adsorption capacity, and the static adsorption capacity was 9.21 mg/g. The adsorption performance was the best at pH = 9, and the adsorption equilibrium was achieved within 40 min. Boron adsorption conformed to the Freundlich adsorption isotherm and to the pseudo-second-order kinetic model. This composite material not only provides an effective and rapid way to remove boron from desalinated seawater, but also has a shorter removal time and makes it more easily separated using the external magnetic field.

Affected by excessive fertilizer application and livestock breeding, the problem of nitrate pollution in the groundwater in the Mihe alluvial–diluvial fan area is becoming increasingly prominent, which poses a great threat to human production and life. Given this, the risk of nitrate pollution in the shallow groundwater of the Mihe alluvial–diluvial fan is evaluated by introducing a data envelopment analysis (DEA) method. Using this model, 28 groundwater sampling points are selected as the decision-making unit (DMU); the nitrogen and pesticide application rate, livestock and poultry stock, groundwater burial depth, aquifer water abundance, and vegetable planting area are taken as the model input; and the nitrate content is taken as the model output to quantitatively calculate the pollution risk index to form a spatial distribution map of pollution risk. The calculation using the model shows that the average pollution risk index of the study area is 0.382, the spatial variation is 1.12, the pollution risk index gradually decreases from south to north, and agricultural planting and livestock and poultry breeding are the main pollution sources. The calculation of nitrate pollution risk using this model not only enriches the nitrate pollution evaluation model but also provides a basis for further implementing the action of reducing fertilizer use by increasing its efficiency and strengthening the prevention of agricultural diffused pollution.