*2.3. Leaching Experiments*

The main reason for the unstable efficiency of phosphorus removal in bioretention facilities is the leaching of phosphorus from filter additives, leading to eutrophication in urban rivers. To avoid this pollution and evaluate the phosphorus leaching quantities, polyurethane-biochar crosslinked material (PCB) and hardwood biochar (HB) were continuously rinsed with deionized water (DW) or artificial stormwater (AS) and the release characterizations of phosphorus and metal ions were analyzed. AS was a mixture solution of 120 mg/L CaCl2 and 3 mg/L PO4-P (Na2HPO4) at pH 7.0, referring to the recognized makeup of synthetic urban runoff [4]. In order to reduce the influence of other factors, oils, heavy metals and nitrogen were not added to the AS.

Five grams of material, dried at 60 ◦C for 48 h, was added to a conical flask containing 100 mL of DW or AS (ISO 21268:2019). At 20 ± 2 ◦C, the material oscillated at a frequency of 150 rpm for 24 h. After settlement for 30 min, supernatants were aspirated into a centrifuge tube and centrifuged at 5000 rpm (Hitachi CR21 III, Tokyo, Japan) for 20 min. Another 100 mL of DW or AS was added to the conical flask and the leaching-settlingcentrifuging steps were repeated another seven times. The supernatants were filtered with 0.45 μm filters and analyzed for phosphate (PO4-P), total phosphorus (TP-P), metal ions (Na+, K+, Mg2+, Ca2+) and water conductivity. The San++ Continuous Flow Analyzer (Skalar Dutch) was used for the testing of PO4-P and TP-P. The samples were mixed with H8MoN2O4, C8H4K2O12Sb2 and C6H8O6 and a colorimetric analysis was carried out at an 880 nm wavelength for the detection of PO4-P with a detection limit of 0.001 mg/L (ISO 15681:2018). The method for the detection of TP-P was the same as PO4-P, only needing a pretreatment (oxidized by K2S2O8 solution and digested by UV) before mixing. Metal ions were tested by using a NexION300X inductively coupled plasma mass spectrometer (PerkinElmer, Waltham, MA, USA) with a detection limit of 0.001 mg/L. Conical bottles containing only DW or AS without other materials were used as the control groups. We repeated two sets of tests for each material. After leaching experiments, the DW-rinsed materials were denoted as PCB-DW and HB-DW respectively and then put into a desiccator for later tests.
