*3.1. Characterization of ZB12-500*

The SEM analysis results of ZB12-500 before, during, and after the reaction are shown in Figure 2a–c. Figure 2a shows the ZB12-500 material before the reaction. The particle size of the composite material is about 30 µm, roughly spherical, smooth surface, and rich honeycomb channels. nZVI particles are evenly distributed on the surface and pores of biochar particles, such as a spider web, and nZVI particles in different pores are separated by a carbon skeleton. This is because of the rich porous structure and large specific surface area of biochar [48]. This image shows that the ZB12-500 composite was well prepared by the sodium borohydride method. Figure 2b shows ZB12-500 during the reaction. Compared with before the reaction, the nZVI particles on the surface of the ZB12-500 were obviously corroded, and the corrosion of nZVI particles was mostly surface corrosion. Some of the porous structures on the surface of the ZB12-500 were blocked because the nZVI lost electrons, and the iron ions diffused from the inner core of the nZVI to the outer core, forming iron oxides on the surface of the particles. In this process, there may be

mechanisms such as iron dissolution, migration, and reagglomeration of nZVI particles, and migration of iron ions between different particles [49]. Figure 2c shows ZB12-500 after the reaction. The surface morphology of the material was highly crystallized, and the porous structure on the surface of the composite material completely disappeared and was covered by iron oxide. core, forming iron oxides on the surface of the particles. In this process, there may be mechanisms such as iron dissolution, migration, and reagglomeration of nZVI particles, and migration of iron ions between different particles [49]. Figure 2c shows ZB12-500 after the reaction. The surface morphology of the material was highly crystallized, and the porous structure on the surface of the composite material completely disappeared and was covered by iron oxide.

where Ct (mg/L) is the residual concentration of nitrate, C0 (mg/L) is the initial concentrations of nitrate; k1 and k2 (1/h) are the reaction rate constants of first and second order reaction kinetics, respectively; ff and fs are the proportion of fast and slow compartment removal in the total removal, respectively, and ff + fs = 1; kf and ks (1/h) are the fast and

The SEM analysis results of ZB12-500 before, during, and after the reaction are shown in Figure 2a–c. Figure 2a shows the ZB12-500 material before the reaction. The particle size of the composite material is about 30 μm, roughly spherical, smooth surface, and rich honeycomb channels. nZVI particles are evenly distributed on the surface and pores of biochar particles, such as a spider web, and nZVI particles in different pores are separated by a carbon skeleton. This is because of the rich porous structure and large specific surface area of biochar [48]. This image shows that the ZB12-500 composite was well prepared by the sodium borohydride method. Figure 2b shows ZB12-500 during the reaction. Compared with before the reaction, the nZVI particles on the surface of the ZB12-500 were obviously corroded, and the corrosion of nZVI particles was mostly surface corrosion. Some of the porous structures on the surface of the ZB12-500 were blocked because the nZVI lost electrons, and the iron ions diffused from the inner core of the nZVI to the outer

*Water* **2022**, *14*, 2877 7 of 23

slow compartment reaction rate constants, respectively.

**3. Results**

*3.1. Characterization of ZB12-500*

The SSA analysis results of ZB12-500 before, during, and after the reaction are shown in Table 3. After the reaction, the SSA of the material increased. The possible reason is that the iron oxides generated by the nZVI reaction were redistributed on the biochar support, which may cover the original nZVI particles or other active sites, resulting in a larger gap between the iron oxides and an increase in SSA. The SSA analysis results of ZB12-500 before, during, and after the reaction are shown in Table 3. After the reaction, the SSA of the material increased. The possible reason is that the iron oxides generated by the nZVI reaction were redistributed on the biochar support, which may cover the original nZVI particles or other active sites, resulting in a larger gap between the iron oxides and an increase in SSA.

**Table 3.** SSA of ZB12-500 before, during, and after the reaction. **Table 3.** SSA of ZB12-500 before, during, and after the reaction.


The XRD analysis results of ZB12-500 before, during, and after the reaction are shown in Figure 2d. ZB12-500B, ZB12-500D, and ZB12-500A represent ZB12-500 samples before,

The results of FTIR analysis of ZB12-500 before, during, and after the reaction are shown in Figure 3. As can be seen from the figure, these three samples contain abundant functional groups, among which a characteristic peak with high strength appears near 3430 cm−1, which corresponds to the tensile vibration of the –OH bond [56–59]. The characteristic peak intensity at 1620 cm−1 is slightly smaller and is related to the C=O and –OH bonds of carbonyl or carboxyl groups. The intensity of characteristic peaks at 1320 cm−1

and 82.60° represent planes (110), (200), and (211) of the body-centered cubic crystal structure in nZVI particles, respectively [50]. Before the reaction, the ZB12-500B sample showed a diffraction peak at 2θ = 44.8°, which corresponds to the body-centered cubic α-Fe0 (110) crystal plane, indicating that the prepared ZB12-500 material contains α-Fe0, and nZVI particles are successfully loaded on the surface of biochar support [51,52]. The peak intensity of Fe0 in the ZB12-500D sample decreases significantly, while the diffraction peak of Fe0 in the ZB12-500A sample completely disappears, indicating that in the process of reducing NO3−-N by ZB12-500, nZVI was all consumed or covered by reaction products. The characteristic peaks of Fe2O3 and Fe3O4 located at 2θ = 35.5° appear in sample ZB12-500D, which is widened to a certain extent in sample ZB12-500A, while these two peaks are not found in sample ZB12-500B, indicating that the main iron oxides generated after the reduction reaction are Fe2O3 and Fe3O4. Some scholars have found that Fe3O4 is formed at the interface between nZVI and iron oxide, while Fe2O3 is formed at the interface between

iron oxide and water [53–55].

The XRD analysis results of ZB12-500 before, during, and after the reaction are shown in Figure 2d. ZB12-500B, ZB12-500D, and ZB12-500A represent ZB12-500 samples before, during, and after the reaction, respectively. The characteristic peaks at 2θ = 44.8◦ , 65.32◦ , and 82.60◦ represent planes (110), (200), and (211) of the body-centered cubic crystal structure in nZVI particles, respectively [50]. Before the reaction, the ZB12-500B sample showed a diffraction peak at 2θ = 44.8◦ , which corresponds to the body-centered cubic α-Fe<sup>0</sup> (110) crystal plane, indicating that the prepared ZB12-500 material contains α-Fe<sup>0</sup> , and nZVI particles are successfully loaded on the surface of biochar support [51,52]. The peak intensity of Fe<sup>0</sup> in the ZB12-500D sample decreases significantly, while the diffraction peak of Fe<sup>0</sup> in the ZB12-500A sample completely disappears, indicating that in the process of reducing NO<sup>3</sup> −-N by ZB12-500, nZVI was all consumed or covered by reaction products. The characteristic peaks of Fe2O<sup>3</sup> and Fe3O<sup>4</sup> located at 2θ = 35.5◦ appear in sample ZB12- 500D, which is widened to a certain extent in sample ZB12-500A, while these two peaks are not found in sample ZB12-500B, indicating that the main iron oxides generated after the reduction reaction are Fe2O<sup>3</sup> and Fe3O4. Some scholars have found that Fe3O<sup>4</sup> is formed at the interface between nZVI and iron oxide, while Fe2O<sup>3</sup> is formed at the interface between iron oxide and water [53–55].

The results of FTIR analysis of ZB12-500 before, during, and after the reaction are shown in Figure 3. As can be seen from the figure, these three samples contain abundant functional groups, among which a characteristic peak with high strength appears near 3430 cm−<sup>1</sup> , which corresponds to the tensile vibration of the –OH bond [56–59]. The characteristic peak intensity at 1620 cm−<sup>1</sup> is slightly smaller and is related to the C=O and –OH bonds of carbonyl or carboxyl groups. The intensity of characteristic peaks at 1320 cm−<sup>1</sup> and 1100 cm−<sup>1</sup> is small, which represents the tensile vibration of the C–O bond. The characteristic peak intensity at 670 cm−<sup>1</sup> is weak, which is the characteristic peak of the Fe–OH bond [60–62]. The number and types of functional groups will greatly affect the physical and chemical properties of materials, especially the –OH functional group, which will affect the absorption of other substances by biochar [63]. Compared with ZB12-500B, the intensity corresponding to the –OH characteristic peak of ZB12-500A and ZB12-500D at 3430 cm−<sup>1</sup> increased, while the intensity of the characteristic peaks at 1620 cm−<sup>1</sup> , 1320 cm−<sup>1</sup> and 670 cm−<sup>1</sup> decreased slightly, which may be due to the formation of C–O–Fe, Fe–O–Fe, and C–N–Fe. *Water* **2022**, *14*, 2877 9 of 23 and 1100 cm−1 is small, which represents the tensile vibration of the C–O bond. The characteristic peak intensity at 670 cm−1 is weak, which is the characteristic peak of the Fe–OH bond [60–62]. The number and types of functional groups will greatly affect the physical and chemical properties of materials, especially the –OH functional group, which will affect the absorption of other substances by biochar [63]. Compared with ZB12-500B, the intensity corresponding to the –OH characteristic peak of ZB12-500A and ZB12-500D at 3430 cm−1 increased, while the intensity of the characteristic peaks at 1620 cm−1, 1320 cm−1 and 670 cm−1 decreased slightly, which may be due to the formation of C–O–Fe, Fe–O–Fe, and C–N–Fe.

**Figure 3.** FTIR spectra of ZB12-500 before, during, and after the reaction. **Figure 3.** FTIR spectra of ZB12-500 before, during, and after the reaction.

Figure 4a–c shows the full spectrum scan of ZB12-500 material before, during, and after the reaction. The peaks of binding energies at 710 eV, 530 eV, and 284 eV correspond to the absorption peaks of Fe 2p, O 1s, and C 1s, respectively [60]. The main elements in the three samples are Fe, O, and C, but the content of the elements is different. Among Figure 4a–c shows the full spectrum scan of ZB12-500 material before, during, and after the reaction. The peaks of binding energies at 710 eV, 530 eV, and 284 eV correspond to the absorption peaks of Fe 2p, O 1s, and C 1s, respectively [60]. The main elements in the three samples are Fe, O, and C, but the content of the elements is different. Among

them, the Fe element is mainly from nZVI and iron oxide, the O element is mainly from oxygen-containing functional groups in biochar, and C element is mainly from biochar.

increased continuously after the beginning of the reaction. This is because although nZVI was consumed during the reaction, the increased content of Fe oxide exceeded the consumed nZVI content. One hour after the reaction between ZB12-500 and NO3−-N solution, the content of the Fe element increased greatly, and in the following 23 h, the content of the Fe element increased slightly, indicating that the reaction was nearly complete after one hour of reaction, which is consistent with the change of the iron element in XRD. The content of the O element in the ZB12-500 increased first and then decreased in the reaction process, which may be related to the consumption of nZVI and the formation of iron oxides in the reaction process. The change in the –OH group content in the FTIR spectra also showed the same trend. The content of element C decreased by 12.98% in 1 h after the reaction, and then decreased by 0.74% in the following 23 h, indicating that the functional group containing C participated in the reaction. This is consistent with the variation of the

C element in the XRD and FTIR spectra.

them, the Fe element is mainly from nZVI and iron oxide, the O element is mainly from oxygen-containing functional groups in biochar, and C element is mainly from biochar. By comparing the scanning spectra, it can be found that the content of the Fe element increased continuously after the beginning of the reaction. This is because although nZVI was consumed during the reaction, the increased content of Fe oxide exceeded the consumed nZVI content. One hour after the reaction between ZB12-500 and NO<sup>3</sup> −-N solution, the content of the Fe element increased greatly, and in the following 23 h, the content of the Fe element increased slightly, indicating that the reaction was nearly complete after one hour of reaction, which is consistent with the change of the iron element in XRD. The content of the O element in the ZB12-500 increased first and then decreased in the reaction process, which may be related to the consumption of nZVI and the formation of iron oxides in the reaction process. The change in the –OH group content in the FTIR spectra also showed the same trend. The content of element C decreased by 12.98% in 1 h after the reaction, and then decreased by 0.74% in the following 23 h, indicating that the functional group containing C participated in the reaction. This is consistent with the variation of the C element in the XRD and FTIR spectra. *Water* **2022**, *14*, 2877 10 of 23

**Figure 4.** XPS survey spectra of ZB12-500 (**a**) before, (**b**) during, and (**c**) after the reaction. Highresolution XPS scan spectra over Fe 2p of ZB12-500 (**d**) before, (**e**) during, and (**f**) after the reaction. **Figure 4.** XPS survey spectra of ZB12-500 (**a**) before, (**b**) during, and (**c**) after the reaction. Highresolution XPS scan spectra over Fe 2p of ZB12-500 (**d**) before, (**e**) during, and (**f**) after the reaction.

Figure 4d–f shows the high-resolution XPS scan spectra over Fe 2p of ZB12-500 before, during, and after the reaction. The peaks at the binding energy of 707 eV, 710 eV, and 712 eV represent the absorption peaks of Fe0, Fe2+ and Fe3+ in Fe 2p3/2, respectively. The

of Fe0, Fe2+ and Fe3+ in Fe 2p1/2, respectively [64,65]. The relative contents of the three different valence iron elements changed obviously before, during, and after the reaction of ZB12-500, especially the peak of Fe0 at 707 eV binding energy, and the intensity of this peak almost disappeared after the reaction. Table 4 shows the relative contents of Fe in different valence states in ZB12-500 before, during, and after the reaction. Compared with the ZB12-500 before the reaction, the content of Fe0 in the composite during and after reaction decreased by 3.84% and 5.00%, respectively, indicating that nZVI was continuously consumed during the whole reaction. The increased of Fe3+ content mainly occurred in the first hour of the reaction process, and then decreased in the next 23 h. Meanwhile, the

Figure 4d–f shows the high-resolution XPS scan spectra over Fe 2p of ZB12-500 before, during, and after the reaction. The peaks at the binding energy of 707 eV, 710 eV, and 712 eV represent the absorption peaks of Fe<sup>0</sup> , Fe2+ and Fe3+ in Fe 2p3/2, respectively. The peaks at the binding energy of 721 eV, 723 eV, and 725 eV represent the absorption peaks of Fe<sup>0</sup> , Fe2+ and Fe3+ in Fe 2p1/2, respectively [64,65]. The relative contents of the three different valence iron elements changed obviously before, during, and after the reaction of ZB12-500, especially the peak of Fe<sup>0</sup> at 707 eV binding energy, and the intensity of this peak almost disappeared after the reaction. Table 4 shows the relative contents of Fe in different valence states in ZB12-500 before, during, and after the reaction. Compared with the ZB12-500 before the reaction, the content of Fe<sup>0</sup> in the composite during and after reaction decreased by 3.84% and 5.00%, respectively, indicating that nZVI was continuously consumed during the whole reaction. The increased of Fe3+ content mainly occurred in the first hour of the reaction process, and then decreased in the next 23 h. Meanwhile, the increased of Fe2+ content mainly occurred in the last 23 h of the reaction process, indicating that Fe3+ was converted to Fe2+ in the later reaction. *Water* **2022**, *14*, 2877 11 of 23 increased of Fe2+ content mainly occurred in the last 23 h of the reaction process, indicating that Fe3+ was converted to Fe2+ in the later reaction.

> **Table 4.** The relative content of Fe valence states measured with XPS. **Table 4.** The relative content of Fe valence states measured with XPS.


*3.2. Effect of the Pyrolysis Temperature of ZB12-500 on Biochar 3.2. Effect of the Pyrolysis Temperature of ZB12-500 on Biochar*

Figure 5 shows ZB12-500 was the most effective reactant, followed, in decreasing order, by ZB12-350, ZB12-650, ZB12-800, and nZVI with removal efficiencies of 93.49%, 86.92%, 86.65%, 84.90%, and 40.14%, respectively. In the five different composites, the N<sup>2</sup> conversion ratios from large to small were ZB12-500 > ZB12-350 > ZB12-800 > ZB12-650 > nZVI, corresponding to 27.34%, 24.12%, 24.11%, 16.83%, and 5.71%, respectively. Figure 5 shows ZB12-500 was the most effective reactant, followed, in decreasing order, by ZB12-350, ZB12-650, ZB12-800, and nZVI with removal efficiencies of 93.49%, 86.92%, 86.65%, 84.90%, and 40.14%, respectively. In the five different composites, the N2 conversion ratios from large to small were ZB12-500 > ZB12-350 > ZB12-800 > ZB12-650 > nZVI, corresponding to 27.34%, 24.12%, 24.11%, 16.83%, and 5.71%, respectively.

### *3.3. Effect of the Dosage of ZB12-500 3.3. Effect of the Dosage of ZB12-500*

with the increase in dosage.

Figure 6 shows how the dosage of ZB12-500 affected the removal efficiency of nitrate. When the dosages of ZB120-500 were 2, 3, 4, 5, 6, and 7 g/L, the removal efficiencies were 54.62%, 72.62%, 89.49%, 93.98%, 94.78%, and 96.04%, respectively. A positive correlation was observed between the dosage of ZB12-500 and the efficiency of nitrate removal. In terms of the selectivity of nitrogen products, when the dosage was less than 5 g/L, the proportion of various reduction products basically did not change, and the N2 conversion Figure 6 shows how the dosage of ZB12-500 affected the removal efficiency of nitrate. When the dosages of ZB120-500 were 2, 3, 4, 5, 6, and 7 g/L, the removal efficiencies were 54.62%, 72.62%, 89.49%, 93.98%, 94.78%, and 96.04%, respectively. A positive correlation was observed between the dosage of ZB12-500 and the efficiency of nitrate removal. In terms of the selectivity of nitrogen products, when the dosage was less than 5 g/L, the proportion of various reduction products basically did not change, and the N<sup>2</sup> conversion

ratio was about 27%, while when the dosage was more than 5 g/L, the N2 conversion ratio decreased to about 24%, and the conversion ratio of NH4+-N and NO2<sup>−</sup> increased slightly

ratio was about 27%, while when the dosage was more than 5 g/L, the N<sup>2</sup> conversion ratio decreased to about 24%, and the conversion ratio of NH<sup>4</sup> + -N and NO<sup>2</sup> − increased slightly with the increase in dosage. *Water* **2022**, *14*, 2877 12 of 23 *Water* **2022**, *14*, 2877 12 of 23

**Figure 6.** Effect of ZB12-500 dosage (initial NO3−-N concentration: 30 mg/L, pH 6, ZB12 sample: ZB12-500). **Figure 6.** Effect of ZB12-500 dosage (initial NO<sup>3</sup> −-N concentration: 30 mg/L, pH 6, ZB12 sample: ZB12-500). **Figure 6.** Effect of ZB12-500 dosage (initial NO3−-N concentration: 30 mg/L, pH 6, ZB12 sample: ZB12-500).

### *3.4. Effect of pH 3.4. Effect of pH 3.4. Effect of pH*

Figure 7 shows the effect of the initial pH on nitrate removal by ZB12-500. Obviously, the removal efficiency was negatively correlated with the initial pH of the solution. When the initial pH increased from 5 to 10, the removal efficiency decreased from 97.29% to 89.04%. The N2 conversion ratio was the highest when pH = 5, followed by pH = 6, and the lowest when pH = 7, which were 27.13%, 26.38%, and 21.92%, respectively. In particular, ZB12-500 exhibited similar nitrate reduction results at initial pH values of 5 and 6. Figure 7 shows the effect of the initial pH on nitrate removal by ZB12-500. Obviously, the removal efficiency was negatively correlated with the initial pH of the solution. When the initial pH increased from 5 to 10, the removal efficiency decreased from 97.29% to 89.04%. The N<sup>2</sup> conversion ratio was the highest when pH = 5, followed by pH = 6, and the lowest when pH = 7, which were 27.13%, 26.38%, and 21.92%, respectively. In particular, ZB12-500 exhibited similar nitrate reduction results at initial pH values of 5 and 6. Figure 7 shows the effect of the initial pH on nitrate removal by ZB12-500. Obviously, the removal efficiency was negatively correlated with the initial pH of the solution. When the initial pH increased from 5 to 10, the removal efficiency decreased from 97.29% to 89.04%. The N2 conversion ratio was the highest when pH = 5, followed by pH = 6, and the lowest when pH = 7, which were 27.13%, 26.38%, and 21.92%, respectively. In particular, ZB12-500 exhibited similar nitrate reduction results at initial pH values of 5 and 6.

**Figure 7.** Effect of initial pH (initial NO3−-N concentration: 30 mg/L, Dosage: 5 g/L, ZB12 sample: ZB12-500). **Figure 7.** Effect of initial pH (initial NO3−-N concentration: 30 mg/L, Dosage: 5 g/L, ZB12 sample: ZB12-500). **Figure 7.** Effect of initial pH (initial NO<sup>3</sup> −-N concentration: 30 mg/L, Dosage: 5 g/L, ZB12 sample: ZB12-500).

### *3.5. Effect of Initial Nitrate Concentration 3.5. Effect of Initial Nitrate Concentration 3.5. Effect of Initial Nitrate Concentration*

Figure 8 shows the effect of the initial concentration of nitrate (NO3−-N) on the removal of nitrate by ZB12-500. When the initial concentration of NO3−-N were 30 mg/L, 50 mg/L, 70 mg/L and 100 mg/L, the removal efficiencies were 93.94%, 82.60%, 75.3%, and 51.42%, respectively, and the N2 conversion ratios were 26.82%, 28.45%, 29.74%, and 31.66%, respectively. Figure 8 shows the effect of the initial concentration of nitrate (NO3−-N) on the removal of nitrate by ZB12-500. When the initial concentration of NO3−-N were 30 mg/L, 50 mg/L, 70 mg/L and 100 mg/L, the removal efficiencies were 93.94%, 82.60%, 75.3%, and 51.42%, respectively, and the N2 conversion ratios were 26.82%, 28.45%, 29.74%, and 31.66%, respectively. Figure 8 shows the effect of the initial concentration of nitrate (NO<sup>3</sup> −-N) on the removal of nitrate by ZB12-500. When the initial concentration of NO<sup>3</sup> −-N were 30 mg/L, 50 mg/L, 70 mg/L and 100 mg/L, the removal efficiencies were 93.94%, 82.60%, 75.3%, and 51.42%, respectively, and the N<sup>2</sup> conversion ratios were 26.82%, 28.45%, 29.74%, and 31.66%, respectively.

**Figure 8.** Effect of initial nitrate (NO3−-N) concentration (pH 6, dosage: 5 g/L, ZB12 sample: ZB12- 500). **Figure 8.** Effect of initial nitrate (NO<sup>3</sup> −-N) concentration (pH 6, dosage: 5 g/L, ZB12 sample: ZB12-500). 500).

### *3.6. Effect of Co-Existing Ions 3.6. Effect of Co-Existing Ions*

*3.6. Effect of Co-Existing Ions* Figure 9 shows the effect of the co-existing ions on nitrate removal by ZB12-500. Common ions (Na+, Mg2+, Ca2+, Cl−, SO42−, and HCO3−) in groundwater were investigated. According to the experimental data in the previous sections, without adding co-existing ions, the removal efficiency of NO3—N exceeded 93%, and the conversion ratio of N2 exceeded 25% under the same reaction conditions. However, after adding co-existing ions Na+, Mg2+, Ca2+, Cl−, SO42−, and HCO3<sup>−</sup> into the solution, the removal efficiencies of NO3−-N were 91.14%, 95.20%, 94.33%, 92.88%, 80.20%, and 57.00%, and the conversion ratios of N2 were Figure 9 shows the effect of the co-existing ions on nitrate removal by ZB12-500. Common ions (Na<sup>+</sup> , Mg2+, Ca2+, Cl−, SO<sup>4</sup> <sup>2</sup>−, and HCO<sup>3</sup> −) in groundwater were investigated. According to the experimental data in the previous sections, without adding co-existing ions, the removal efficiency of NO<sup>3</sup> <sup>−</sup>-N exceeded 93%, and the conversion ratio of N<sup>2</sup> exceeded 25% under the same reaction conditions. However, after adding co-existing ions Na<sup>+</sup> , Mg2+, Ca2+, Cl−, SO<sup>4</sup> <sup>2</sup>−, and HCO<sup>3</sup> − into the solution, the removal efficiencies of NO<sup>3</sup> −-N were 91.14%, 95.20%, 94.33%, 92.88%, 80.20%, and 57.00%, and the conversion ratios of N<sup>2</sup> were 26.30%, 26.01%, 25.02%, 25.82%, 29.49%, and 37.01%, respectively. Figure 9 shows the effect of the co-existing ions on nitrate removal by ZB12-500. Common ions (Na+, Mg2+, Ca2+, Cl−, SO42−, and HCO3−) in groundwater were investigated. According to the experimental data in the previous sections, without adding co-existing ions, the removal efficiency of NO3—N exceeded 93%, and the conversion ratio of N2 exceeded 25% under the same reaction conditions. However, after adding co-existing ions Na+, Mg2+, Ca2+, Cl−, SO42−, and HCO3<sup>−</sup> into the solution, the removal efficiencies of NO3−-N were 91.14%, 95.20%, 94.33%, 92.88%, 80.20%, and 57.00%, and the conversion ratios of N2 were 26.30%, 26.01%, 25.02%, 25.82%, 29.49%, and 37.01%, respectively.

26.30%, 26.01%, 25.02%, 25.82%, 29.49%, and 37.01%, respectively.

**Figure 9.** Effect of co-existing ions (initial NO3−-N concentration: 30 mg/L, pH 6, Dosage: 5 g/L, ZB12 sample: ZB12-500). **Figure 9.** Effect of co-existing ions (initial NO3−-N concentration: 30 mg/L, pH 6, Dosage: 5 g/L, ZB12 sample: ZB12-500). **Figure 9.** Effect of co-existing ions (initial NO<sup>3</sup> −-N concentration: 30 mg/L, pH 6, Dosage: 5 g/L, ZB12 sample: ZB12-500).

### *3.7. Kinetics 3.7. Kinetics*

*3.7. Kinetics* The kinetics data were fitted using first-order kinetic, second-order kinetic, and twocompartment first-order kinetic equations, as shown in Figure 10. The whole removal process was obviously divided into two stages: 0–2 h for the rapid removal stage and 2–24 h for the slow removal stage. The kinetics parameters fitted by the kinetics model are shown in Table 5. The results show that the R2 values of the three kinetics models were all above 0.99, indicating that both adsorption and reduction reactions existed in the removal pro-The kinetics data were fitted using first-order kinetic, second-order kinetic, and twocompartment first-order kinetic equations, as shown in Figure 10. The whole removal process was obviously divided into two stages: 0–2 h for the rapid removal stage and 2–24 h for the slow removal stage. The kinetics parameters fitted by the kinetics model are shown in Table 5. The results show that the R2 values of the three kinetics models were all above 0.99, indicating that both adsorption and reduction reactions existed in the removal process. The R2 of the two-compartment first-order kinetic model was the highest, which was The kinetics data were fitted using first-order kinetic, second-order kinetic, and twocompartment first-order kinetic equations, as shown in Figure 10. The whole removal process was obviously divided into two stages: 0–2 h for the rapid removal stage and 2–24 h for the slow removal stage. The kinetics parameters fitted by the kinetics model are shown in Table 5. The results show that the R<sup>2</sup> values of the three kinetics models were allabove 0.99, indicating that both adsorption and reduction reactions existed in the removal process. The R<sup>2</sup> of the two-compartment first-order kinetic model was the highest, which

cess. The R2 of the two-compartment first-order kinetic model was the highest, which was

was 0.997, indicating that it is more reasonable to use the two-compartment first-order kinetic model to explain the process of NO<sup>3</sup> −-N removal by the ZB12-500 composite. From the two-compartment first-order kinetic parameters, the main stage in the whole removal process was the fast compartment reaction stage, accounting for 92.5%, while the slow compartment reaction stage only accounted for 7.5%. The fast compartment reaction rate constant was 3.093 h−<sup>1</sup> , and the slow compartment reaction rate constant was 0.038 h−<sup>1</sup> . 0.997, indicating that it is more reasonable to use the two-compartment first-order kinetic model to explain the process of NO3−-N removal by the ZB12-500 composite. From the two-compartment first-order kinetic parameters, the main stage in the whole removal process was the fast compartment reaction stage, accounting for 92.5%, while the slow compartment reaction stage only accounted for 7.5%. The fast compartment reaction rate constant was 3.093 h−1, and the slow compartment reaction rate constant was 0.038 h−1.

**Figure 10.** Kinetics of ZB12-500 for NO3−-N. (initial NO3−-N concentration: 30 mg/L, pH 6, Dosage: 5 g/L, ZB12 sample: ZB12-500). **Figure 10.** Kinetics of ZB12-500 for NO<sup>3</sup> −-N. (initial NO<sup>3</sup> −-N concentration: 30 mg/L, pH 6, Dosage: 5 g/L, ZB12 sample: ZB12-500).

**Table 5.** Kinetic parameters of NO3−-N removal by ZB12-500. **Table 5.** Kinetic parameters of NO<sup>3</sup> −-N removal by ZB12-500.


### *4.1. Effect of the Pyrolysis Temperature of ZB12-500 on Biochar* **4. Discussion**

### For nZVI, the removal efficiency of nitrate is greatly affected by agglomeration. The *4.1. Effect of the Pyrolysis Temperature of ZB12-500 on Biochar*

low N2 selectivity of pure nZVI is also affected by the natural defects of the materials. In the reduction process, after the surface active site of the aggregate is inactivated by the reaction, the internal active site is also inactivated by being covered. For ZB12, the removal efficiency of NO3−-N firstly increased and then decreased with the increase of biochar pyrolysis temperature. However, the selectivity of the nitrogen products had no obvious rule with the pyrolysis temperature of the biochar carrier. Although increasing the temperature of pyrolysis increases the electronic conductivity and the degree of graphitization of biochar, it also leads to the loss of functional groups [37,66]. Some scholars have pointed out that with an increase in pyrolysis temperature, the number of functional groups of wood-based biochar (mainly –OH and aliphatic C–H functional groups) and grass-based biochar (mainly C–O functional groups) decreases [67]. The trend of nitrate removal efficiency and nitrogen product selectivity with pyrolysis temperature proves to some extent that when the pyrolysis temperature of biochar is 500 °C, the types and number of functional groups related to reduced nitrate in ZB12 reach an ideal equilibrium state with their electrical conductivity and graphitization degree. The NO2−-N produced by the reaction of the five kinds of samples with nitrite nitrogen was very small, indicating that NO2−-N was the intermediate product of the reaction. Briefly, ZB12-500 was the best reactant for nitrate removal and was selected for the following studies. For nZVI, the removal efficiency of nitrate is greatly affected by agglomeration. The low N<sup>2</sup> selectivity of pure nZVI is also affected by the natural defects of the materials. In the reduction process, after the surface active site of the aggregate is inactivated by the reaction, the internal active site is also inactivated by being covered. For ZB12, the removal efficiency of NO<sup>3</sup> −-N firstly increased and then decreased with the increase of biochar pyrolysis temperature. However, the selectivity of the nitrogen products had no obvious rule with the pyrolysis temperature of the biochar carrier. Although increasing the temperature of pyrolysis increases the electronic conductivity and the degree of graphitization of biochar, it also leads to the loss of functional groups [37,66]. Some scholars have pointed out that with an increase in pyrolysis temperature, the number of functional groups of woodbased biochar (mainly –OH and aliphatic C–H functional groups) and grass-based biochar (mainly C–O functional groups) decreases [67]. The trend of nitrate removal efficiency and nitrogen product selectivity with pyrolysis temperature proves to some extent that when the pyrolysis temperature of biochar is 500 ◦C, the types and number of functional groups related to reduced nitrate in ZB12 reach an ideal equilibrium state with their electrical conductivity and graphitization degree. The NO<sup>2</sup> −-N produced by the reaction of the five kinds of samples with nitrite nitrogen was very small, indicating that NO<sup>2</sup> −-N was the intermediate product of the reaction. Briefly, ZB12-500 was the best reactant for nitrate removal and was selected for the following studies.
