Photo-Catalytic Reduction of Nitrate by Ag-TiO₂/Formic Acid Under Visible Light: Selectivity of Nitrogen and Mechanism

Abstract

Ubiquitous nitrate (NO₃⁻) in groundwater sources is considered a hazard compound for human health. Photo-catalytic reduction by Ag-TiO₂/formic acid/visible light represents an emerging method for NO₃⁻ removal without secondary pollution. In this contribution, the removal of NO₃⁻ by photo-catalytic reduction and the selectivity of N₂ were systematically investigated under varied conditions, including concentrations of Ag-TiO₂, NO₃⁻, and formic acid (HCOOH). The removal efficiency of NO₃⁻ reached 84.47%, 82.68% of which was converted to N₂ under the optimal conditions: NO₃⁻ at 50 mg-N/L, Ag-TiO₂ at 1.0 g/L, HCOOH at 20.05 mmol/L, and reaction time at 120 min. The removal of NO₃⁻ was enhanced mainly by CO₂⁻ rather than by photo-generated electrons or HCOO⁻. The results of this study indicated that the production of ·CO₂⁻ by Ag-TiO₂ and HCOOH under visible light catalysis can achieve efficient NO₃⁻ removal.

1. Introduction

Nitrate (NO₃⁻) pollution is widespread in groundwater due to NO₃⁻ inevitably entering the environment through various routes, including over-fertilization and the discharge of industrial wastewater, etc. [1,2]. NO₃⁻ ingestion causes great damage to health, for example, methemoglobinemia, stomach cancer, or diabetes, with great effect on the health of babies, for example, “blue baby syndrome” or cyanosis [3,4]. High N causes water eutrophication and deteriorates natural aquatic and ecological system [5]. Therefore, it is imperative to develop effective and sustainable treatment technologies to remove NO₃⁻ and minimize associated risks.

The removal of NO₃⁻ can be realized by traditional physical–chemical treatment processes and biological denitrification [6,7]. However, conventional technologies have drawbacks, including low efficiency, secondary pollution, high cost, and residual waste streams of NO₃⁻, nitrite (NO₂⁻), and ammonia (NH₄⁺) [8,9,10]. In general, NO₃⁻ is easier to be converted to NH₄⁺, but the production of nontoxic forms like nitrogen (N₂) is preferred [11,12,13,14,15,16,17]. Increasing N₂ selectivity is a significant challenge in reducing NO₃⁻ in groundwater [11,18]. Many studies have found that photocatalysis can effectively remove pollutants from water [19], and carbon dioxide anion radical (·CO₂⁻) can efficiently reduce NO₃⁻ to N₂ [20]. Chen et al. achieved 60% selectivity of N₂ by ·CO₂⁻ under a UV/Fe(III)–oxalate system [10]. ·CO₂⁻ with a redox potential (−1.8 V) [21] similar to ·H (−2.3 V) [22] can be produced by the action of dissolving CO₂ and solvating electrons in water [23], or by the interaction of hydrogen free radical (·H) [24], sulfate radical (·SO₄⁻) [25], and hydroxyl radical (·OH) [26] with organic carboxylic acids, including formic acid (HCOOH) and oxalic acid [27]. ·OH has shown certain advantages over ·SO₄⁻ due to the production of SO₄²⁻, which can cause the secondary pollution. However, ·OH is usually generated by vacuum ultraviolet photolysis of water (λexc = 185 nm) or the activation of H₂O₂, which requires high energy or has storage and transportation limitations.

Efficient ·CO₂⁻ production is urgently required. Using visible light instead of ultraviolet light to produce ·CO₂⁻ has great application prospects. Combining visible light and modified TiO₂ to produce ·CO₂⁻ is a promising alternative. Modified TiO₂ would show visible-light activity after deposition of Ag due to the localized surface plasmon resonance and form holes (h⁺) [28,29,30,31,32,33]. h⁺ with strong oxidizing ability is inclined to react with H₂O to generate ·H, which could degrade the organic pollution efficiently. Ghafourian et al. achieved a high degradation efficiency of furan formaldehyde at 80.2% using visible/Ag-TiO₂ [34]. ·OH or h⁺ could react with formic acid substances (formate (HCOO⁻) or HCOOH) to generate ·CO₂⁻ [27,35]. Photo-catalytic NO₃⁻ reduction by ·CO₂⁻ is energy-efficient and has no secondary pollution.

However, visible/Ag-TiO₂/HCOOH is rarely used to remove NO₃⁻, and the reduction efficiency, byproducts selectivity, and application potential of NO₃⁻ remain unclear. It should be noted that the realistic product of NO₃⁻ reduction is N₂ rather than NH₄⁺. Moreover, many studies have used HCOOH as a hole scavenger to reduce NO₃⁻, but few have focused on the role of HCOOH in the system. Duan et al. reported that photogenerated electrons capture NO₃⁻ and reduce it to N₂ [29]. However, Anderson et al. believed that hole-capture agents (such as formic acid) should be added to fill the holes and generate carbon dioxide free radicals to reduce NO₃⁻ to N₂ [28]. Thus, the mechanism of photo-catalytic reduction of NO₃⁻ requires further study.

Given these challenges and demands, the objectives of this study are as follows: (1) to determine the optimal experimental condition to achieve the highest NO₃⁻ removal efficiently and N₂ conversion efficiency and (2) to explore whether the main reducing substances are photogenerated electrons or ·CO₂⁻ and reveal the possible mechanisms of NO₃⁻ reduction by Ag-TiO₂/HCOOH under visible light.

2. Materials and Methods

2.1. Materials

All chemicals were of reagent grade without further purification. Titanium dioxide (TiO₂, >99%, Shanghai Aladdin Biochemical Technology Co., Ltd., Shanghai, China). Silver nitrate (AgNO₃, >99%, Shanghai Aladdin Biochemical Technology Co., Ltd., Shanghai, China), sodium formate (HCOONa, >99%, Shanghai Aladdin Biochemical Technology Co., Ltd., Shanghai, China), formic acid (HCOOH, >99%, Shanghai Aladdin Biochemical Technology Co., Ltd., Shanghai, China), potassium salt (KNO₃, KCl, K₂SO₄, KHCO₃, KH₂PO₄, and KI, all >99%, Shanghai Aladdin Biochemical Technology Co., Ltd., Shanghai, China), hydrochloric acid (HCl, 36%~38%, Shanghai Aladdin Biochemical Technology Co., Ltd., Shanghai, China), methanol (CH₃OH, >99%, Shanghai Aladdin Biochemical Technology Co., Ltd., Shanghai, China), acetic acid (CH₃COOH, >99%, Shanghai Aladdin Biochemical Technology Co., Ltd., Shanghai, China), ethylene diamine tetra acetic acid disodium (EDTA, >99%, Shanghai Aladdin Biochemical Technology Co., Ltd., Shanghai, China), and sodium hydroxide (NaOH, ≥96%, Sinopharm Chemical Reagent Co., Ltd., Shanghai, China).

2.2. Preparation and Characterization of Ag-TiO₂

Ag-TiO₂ was prepared by the ultraviolet light reduction deposition method [36]. The detailed preparation steps of Ag-TiO₂ are given in Text S1 (Supplementary Materials).

Ag-TiO₂ was characterized by X-ray diffraction (XRD), ultraviolet visible diffuse reflectance spectra (UV–vis DRS), X-ray photoelectron spectroscopy (XPS), and photoluminescence spectra (PL). The crystal structure of Ag-TiO₂ was analyzed by XRD (Dandong Aolong Radiation Instrument Group Co., Ltd., Dandong, China) with the instrument parameters for Cu Ka radiation (λ = 1.54056 Å) at 20–40 kV and 10–450 mA at a scanning rate of 8°/min over the 2θ range of 5–80°. The characterization of UV–vis DRS was operated through a Japanese Shimazu U-3010 UV–vis diffusion reflectance spectrometer (Qingdao Xingzhongyi Intelligent Technology Co., Ltd., Qingdao, China) at a scanning rate of 300 nm/min and a scanning wavelength range of 200–800 nm. The surface chemical states was analyzed by XPS run on ESCALAB 250Xi spectrometer (Dens Instrument Technology (Wuxi) Co., Ltd., Wuxi, China) using the K ray of Al target as the source of radiation (1486.6 eV), at a scanning range of 1100–0 eV; the spot size was 500 μm, and the voltage and current were 15 kV and 10 mA. The separation efficiency of the photo-generated e⁻ and h⁺ of Ag-TiO₂ was analyzed through the characterization of PL by Fluorescence spectrophotometer (F-7000, Zhejiang Lichen Instrument Technology Co., Ltd., Shaoxing, China). Detailed characterization results are given in Figures S1–S5 (Supplementary Materials).

2.3. Photocatalytic Experiments

Photo-catalytic reduction of NO₃⁻ was performed in quartz glass (100 mL) that was continuously purged with N₂ to remove dissolved oxygen; the suspension was then magnetically stirred to ensure a uniform contact between Ag-TiO₂ and NO₃⁻ during the whole process. Next, 1.0 g/L of Ag-TiO₂ and 100 mL of KNO₃ (50 mg-N/L) was put into the reactor, 20.05 mmol/L of HCOOH was added, and the obtained suspension liquid was stirred for 30 min in the dark. Then, it was irradiated by a 500 W xenon lamp, which provided a visible light source, with continuous stirring for 3 h; a quartz cold hydrazine was used to control the temperature of the quartz glass. After the reaction was completed, the reaction fluid was filtrated through a filtration membrane (0.45 μm), and the supernatant was diluted to determine the concentration of NO₃⁻, NO₂⁻, and ammonia (NH₄⁺) by ultraviolet spectrophotometer [10,37].

2.4. Analytical Methods

The analysis of N-containing compounds was detected using a UV–visible spectrophotometer (UV-1800PC) and ion chromatography (MIC, Zurich, Switzerland). NO₂⁻, NH₄⁺, and N₂ are the main reduction products in the process of visible light catalytic reduction of NO₃⁻. The product selectivity to NO₂⁻, NH₄⁺, and N₂ was calculated according to Equations (1)–(3).

$$\mathrm{S}\left(\mathrm{N}{\mathrm{O}}_2^{-}\right)={\displaystyle \frac{{\mathrm{C}}_{\mathrm{t}}\left(\mathrm{N}{\mathrm{O}}_2^{-}\right)}{{\mathrm{C}}_0\left(\mathrm{N}{\mathrm{O}}_3^{-}\right)-{\mathrm{C}}_{\mathrm{t}}\left(\mathrm{N}{\mathrm{O}}_3^{-}\right)}}$$

(1)

$$\mathrm{S}\left(\mathrm{N}{\mathrm{H}}_4^{+}\right)={\displaystyle \frac{{\mathrm{C}}_{\mathrm{t}}\left(\mathrm{N}{\mathrm{H}}_4^{+}\right)}{{\mathrm{C}}_0\left(\mathrm{N}{\mathrm{O}}_3^{-}\right)-{\mathrm{C}}_{\mathrm{t}}\left(\mathrm{N}{\mathrm{O}}_3^{-}\right)}}$$

(2)

$$\mathrm{S}\left({\mathrm{N}}_2\right)={\displaystyle \frac{{\mathrm{C}}_0\left(\mathrm{N}{\mathrm{O}}_3^{-}\right)-{\mathrm{C}}_{\mathrm{t}}\left(\mathrm{N}{\mathrm{O}}_3^{-}\right)-{\mathrm{C}}_{\mathrm{t}}\left(\mathrm{N}{\mathrm{O}}_2^{-}\right)-{\mathrm{C}}_{\mathrm{t}}\left(\mathrm{N}{\mathrm{H}}_4^{+}\right)}{{\mathrm{C}}_0\left(\mathrm{N}{\mathrm{O}}_3^{-}\right)-{\mathrm{C}}_{\mathrm{t}}\left(\mathrm{N}{\mathrm{O}}_3^{-}\right)}}$$

(3)

where C₀ is the initial concentration of the substance, and C_t is the concentration of the substance in the test solution at reaction time t [38].

3. Results and Discussion

3.1. Removal of Nitrate

In actual wastewater, the initial concentration of NO₃⁻ fluctuates over time. Therefore, it is significant and necessary to investigate the photo-catalytic performance under the different NO₃⁻ concentrations. The reduction of NO₃⁻ was evaluated under a visible light system. Figure 1a shows the removal of NO₃⁻ in different initial concentrations (15–100 mg-N/L). After 30 min of reaction, the removal efficiency of NO₃⁻ at 15 mg-N/L reached 77.63%, and almost all of the NO₃⁻ was reduced after 60 min. The removal efficiency reached 84.51% at 50 mg-N/L after 120 min. The removal efficiency of NO₃⁻ was 70.59% after 210 min when NO₃⁻ was 100 mg-N/L. The removal efficiency of NO₃⁻ was slightly decreased with increasing NO₃⁻ concentration. Although the photo-catalytic efficiency decreased with increasing initial concentration, the actual removal of NO₃⁻ from 15 mg-N/L increased to 70.59 mg-N/L. Under each NO₃⁻ concentration investigated, NO₃⁻ reduction could be fitted by Langmuir Hinshelwood kinetic model described by Equation (4).

$$\mathrm{r}=-{\displaystyle \frac{\mathrm{dC}}{\mathrm{d}\mathrm{t}}}={\displaystyle \frac{{\mathrm{k}}_{\mathrm{r}}{\mathrm{K}}_{\mathrm{a}\mathrm{d}}\mathrm{C}}{1+{\mathrm{K}}_{\mathrm{a}\mathrm{d}}\mathrm{C}}}$$

(4)

where r is the removal efficiency of NO₃⁻, C is the concentration of NO₃⁻ at reaction time t, Kr is the intrinsic rate, and K_ad is the adsorption equilibrium constant.

As the adsorption is weaker, and the initial concentration of NO₃⁻ is lower, K_adC can be ignored. In this case, the model can be simplified into a first-order kinetic equation (Equation (5)). For the initial reaction efficiency (t = 0, C = C₀), the model can be simplified as follows (Equation (6)).

$$\mathrm{r}={\mathrm{k}}_{\mathrm{r}}{\mathrm{K}}_{\mathrm{a}\mathrm{d}}\mathrm{C}=\mathrm{kC}$$

(5)

$$\ln {\displaystyle \frac{\mathrm{c}}{{\mathrm{c}}_0}}=-\mathrm{kt}$$

(6)

where C₀ is the initial concentration of NO₃⁻, k is the pseudo-first-order rate constant that can be obtained by linear regression of the plot of ln(C/C₀) versus t [39,40], and the fitting parameters are shown in Table S1 (Supplementary Data). As illustrated in Figure 1b, a linear relationship between k and concentration of NO₃⁻ was established. The k was decreased with increasing NO₃⁻ concentration; that is, the photo-catalytic reduction efficiency of NO₃⁻ was decreased. When the NO₃⁻ concentration gradually increases, the active sites provided by the fixed dosage of Ag-TiO₂ are insufficient [41,42]. The excess NO₃⁻ was adsorbed and accumulated on the surface of Ag-TiO₂, which would inhibit the adsorption of HCOO⁻ and the production of ·CO₂⁻. Meanwhile, a high initial concentration of NO₃⁻ could reduce light penetration and photon energy and lead to a lower removal efficiency [43]. In fact, NO₃⁻ levels in drinking water sources in many countries still exceed the maximum contaminant level of 50 mg-N/L [44]. Therefore, 50 mg-N/L was selected as the initial NO₃⁻ concentration through the whole experiment.

3.2. Selectivity of Nitrogen

3.2.1. Effect of Ag-TiO₂ Dosage

The photo-catalyst is an important part of photo-catalytic reaction system, and a photo-catalytic reaction cannot carry on in its absence. Figure 2 illustrates the effects of Ag-TiO₂ dosage on NO₃⁻ removal efficiency and product selectivity. The removal efficiency of NO₃⁻ increased from 58.79% to 78.94% and then decreased to 54.77% with the increase of Ag-TiO₂ from 0.5 to 2.0 g/L, as shown in Figure 2a. N₂ selectivity showed the same trend with NO₃⁻ removal efficiency, which firstly increased from 53.10% to 69.32% and then decreased to 59.46%. The opposite trend was found for NO₂⁻ selectivity: it decreased from 40.98% to 21.84% and then increased to 29.78%. The selectivity of NH₄⁺ increased slightly with the increase of Ag-TiO₂ dosage but remained at a low concentration of <10.76%. The N₂ selectivity was consistent with Sowmya et al.’s observation [40].

Further, the selection intensity of the product can be expressed by multiplying the NO₃⁻ removal efficiency by the product selectivity. The N₂ selection intensity was the highest when the Ag-TiO₂ dosage was at 1.0 g/L, as shown in Figure 2b. The reasons are as follows. When the dosage of Ag-TiO₂ increased from 0.5 to 1.0 g/L, ·CO₂⁻ generation was increased to a sufficient value, and the turbidity induced by the addition of Ag-TiO₂ did not affect the visible light. However, when Ag-TiO₂ was further increased from 1.0 to 2.0 g/L, the active sites were covered by Ag-TiO₂ particulate, which was not conducive to the transformation of NO₃⁻ and led to NO₂⁻ accumulation [45,46]. The turbidity also affected the visible light. The production process of ·CO₂⁻ in the photo-catalytic reduction of nitrate by Ag-TiO₂/HCOOH is clearly illustrated by Equations (7)–(11). Ag-TiO₂ generated hot electrons and h⁺ under the irradiation of visible light; meanwhile, the hot electrons are injected into the adjacent semiconductor TiO₂. The h⁺ can react with H₂O to produce ·OH Equations (7)–(9) [34]. When HCOOH is used as a hole scavenger, ·OH and h⁺ react with formic acid to produce ·CO₂⁻ (Equations (10) and (11)) [47].

Ag + hv(visible) → Ag*

(7)

Ag + TiO₂ → Ag⁺ (h⁺) + TiO₂ (e⁻)

(8)

H₂O + h⁺ → H⁺ + ·OH

(9)

·OH + HCOO⁻ → ·CO₂⁻ + H₂O

(10)

h⁺ + HCOO⁻ → H⁺ + ·CO₂⁻

(11)

3.2.2. Effect of Nitrate Concentrations

Figure 3a illustrates the influence of initial NO₃⁻ concentration from 15 to 50 and 100 mg-N/L on the NO₃⁻ removal efficiency and product selectivity. With initial NO₃⁻ concentrations increasing from 15 to 50 and 100 mg-N/L, the NO₃⁻ removal efficiency decreased from 100% to 84.46% and 54.98% at a Ag-TiO₂ dosage of 1.0 g/L. The removal efficiency of NO₃⁻ was negatively correlated with its concentration. The low NO₃⁻ removal efficiency at 100 mg-N/L was due to the low Ag-TiO₂ dosage, which could not provide sufficient h⁺. The excess NO₃⁻ competed with HCOO⁻ for h⁺, which decreased ·CO₂⁻ production and further inhibited NO₃⁻ transformation to N₂, and thus, NO₂⁻ accumulated. As shown in Figure 3a, the selectivity of NO₂⁻ increased from 0.41% to 21.63%. The selectivity of N₂ was 70.46%, 82.68%, and 77.84% at NO₃⁻ concentrations of 15, 50, and 100 mg-N/L. The NH₄⁺ selectivity decreased from 29.13% to 3.36% and 0.53% (Figure 3a). At a low NO₃⁻ of 15 mg-N/L, NO₂⁻ absorbed on the surface of Ag-TiO₂ and was converted to N₂ or NH₄⁺. As the NO₃⁻ concentration increased, more ·OH was generated, as shown in Equations (12) and (13). The ·OH transformed NO₂⁻ to HOONO and reduced the selectivity of N₂ and NH₄⁺ (Equations (14) and (15)) [43]. The maximum N₂ selection intensity was obtained when the NO₃⁻ concentration was 50 mg-N/L, as shown in Figure 3b. Based on NO₃⁻ removal efficiency and N₂ selection intensity, the optimal NO₃⁻ concentration is 50 mg-N/L.

NO₃⁻ + hv → ·NO₂ + ·O⁻

(12)

·O⁻ + H⁺ ↔·OH

(13)

·OH + NO₂⁻ → ·NO₂ + OH⁻

(14)

·OH + ·NO₂ →HOONO

(15)

3.2.3. HCOOH Concentrations

The effect of the concentration of HCOOH on the photo-catalytic NO₃⁻ reduction is illustrated in Figure 4. Figure 4a shows the removal efficiency of NO₃⁻ increased from 47.17% to 81.21% along with the increasing dosage of HCOOH from 4.01 mmol/L to 20.05 mmol/L and reached the highest at 20.05 mmol/L (M(HCOOH:NO₃⁻) = 5.6). When the HCOOH concentration further increased to 40.11 mmol/L and 80.22 mmol/L, the removal efficiency of NO₃⁻ decreased to 56.30% and 38.28%. The highest selectivity of N₂ was achieved at 70.33%.

When the HCOOH concentration is low, less ·CO₂⁻ is produced, and NO₃⁻ removal and N₂ selectivity are low. Nevertheless, a high concentration of HCOOH hampers the adsorption of NO₃⁻ on the surface of Ag-TiO₂ due to the competitive adsorption between negatively charged HCOO⁻ and NO₃⁻, which also inhibits NO₃⁻ removal.

NH₄⁺ was maintained at a low selectivity of 5% at all HCOOH concentrations. The selectivity of NO₂⁻ was high at 76.60% when the HCOOH concentration was 4.01 mmol/L. NO₂⁻ decreased at first and then increased with an increase in HCOOH concentration, and N₂ showed the opposite trend. Zhang et al. explained that the insufficiency of HCOOH could promote the formation of NO₂⁻ (Equations (16) and (17)) [48,49]; hence, HCOOH was easily transformed into NO₂⁻ when NO₃⁻ was insufficient. Figure 4b shows that the maximum N₂ selectivity at the concentration of HCOOH was 20.05 mmol/L. In addition, considering the limit of COD proposed by the Chinese environmental protection agency, 20.05 mmol/L of NO₃⁻ concentration was selected as the optimal dosage of hole scavengers.

2NO₃⁻ + 5HCOO⁻ + 7H⁺ → N₂ + 5CO₂ + 6H₂O

(16)

NO₃⁻ + HCOO⁻ + H⁺ → NO₂⁻ + CO₂ + H₂O

(17)

3.3. Mechanism of Nitrate Reduction

3.3.1. Choice of Hole Scavengers

Hole scavengers play a significant role in the photo-catalytic process. In order to maintain the electrical neutrality of materials, the generated holes need to be transferred by hole scavengers. Hole scavengers can irreversibly combine with photo-generated h⁺ and effectively prevent the combination of photo-generated electrons and h⁺, thereby promoting the efficiency of the photo-catalytic reduction of NO₃⁻. Different hole scavengers have different binding abilities with photo-generated h⁺ and affect the efficiency of the photo-catalytic reduction of NO₃⁻. Hence, it is very important to select the appropriate hole scavengers. In order to determine the optimal hole scavengers with high efficiency in the photo-catalytic NO₃⁻ reduction process, we compared HCOOH with its substructural analogs, including HCOONa, CH₃OH, and CH₃COOH.

As shown in Figure 5a, the removal efficiency of NO₃⁻ followed the order of HCOOH > HCOONa > CH₃OH > CH₃COOH due to the difference in initial pH (

Table 1. The initial solution pH corresponding to the different hole scavengers.

Hole Scavengers	HCOOH	HCOONa	CH3OH	CH3COOH
Initial pH	2.7	6.6	5.2	3.5

). The initial pH can affect the surface charge of TiO₂. There is a zwitterionic substance titanium hydroxyl group (TiOH) on the surface of TiO₂, as shown in reaction Equations (18) and (19) [50]. The isoelectric point of titanium dioxide is 6.25. When the solution pH is below 6.25, the surface titanium hydroxyl group of TiO₂ will undergo protonation, and the surface of TiO₂ will carry a positive charge. This is conducive to adsorbing the negatively charged ions by electrostatic adsorption.

The pH of 2.7 of the HCOOH solution was lower than the isoelectric point ≈ 6.25 of TiO₂, which was conducive to adsorption of NO₃⁻. The pH of 6.63 of the HCOONa system was higher than 6.25 (isoelectric point of TiO₂), which made the surface of Ag-TiO₂ negatively charged. And due to electrostatic repulsion, this was not conducive to the adsorption of NO₃⁻ and HCOO⁻, resulting in a low NO₃⁻ removal efficiency. Although the pH in both cases was lower than 6.25, the removal efficiency of NO₃⁻ was unsatisfactory with CH₃OH or CH₃COOH as hole scavengers. Generally, CH₃OH was used as a hole scavenger, and it was oxidized to formaldehyde by holes rather than ·CO₂⁻. Moreover, the adsorption of CH₃OH on the catalyst surface was extremely weak and could not capture the h⁺ effectively. Ag-TiO₂ has a strong adsorption capacity for CH₃COOH, which inhibits the adsorption of NO₃⁻ on the surface of Ag-TiO₂. It was thus concluded that HCOOH is the optimal hole scavenger.

TiOH + H⁺ → TiOH₂⁺  pH < 6.25

(18)

TiOH → TiO⁻ + H⁺   pH > 6.25

(19)

Based on previous studies, the effect of HCOOH as hole scavenger can be summarized as follows: (1) HCOOH can capture holes, reduce the combination probability of holes and electrons, and improve the yield of the photo-quanta of the catalyst; (2) HCOOH can yield an initial pH of 2.7, lower than the isoelectric point of TiO₂, which is conducive to adsorbing NO₃⁻ and HCOO⁻ on the surface of TiO₂ for catalysis and promoting the reaction of the photo-catalytic reduction of NO₃⁻; (3) HCOOH has reducibility (E₀ (CO₂/HCOO⁻) = −0.2 eV) and can also be used as a reducing agent for NO₃⁻ reduction in theory; and (4) HCOOH can be oxidized by holes to generate ·CO₂⁻ (E₀ (CO₂/·CO₂⁻) = −1.8 eV) in the process of trapping holes. In the HCOOH and HCOONa photo-catalytic reduction system, the reducing substances include photo-generated electrons (E₀ (e) = −0.3 V) generated by the catalyst excited by incident light and ·CO₂⁻ from HCOO⁻ (E₀ (CO₂/HCOO⁻) = −0.2 eV), which contribute to efficient NO₃⁻ removal. Some scholars believe that, in the process of the photo-catalytic reduction of NO₃⁻, ·CO₂⁻ plays a major role instead of photo-generated electrons [29]. However, fewer studies have provided controversial evidence that photo-generated electrons are more important [10,28]. This study was conducted to verify whether photo-generated e⁻ or ·CO₂⁻ is predominant in the photo-catalytic reduction of nitrate by Ag-TiO₂/HCOOH under visible light.

3.3.2. Effective Reducing Substances

In this study, the major reducing agent was explored by controlling the concentration of HCOOH to control the formation of ·CO₂⁻ and combining with the free radical quenching experiment to indirectly control the production path of photo-generated e⁻, HCOO⁻, and ·CO₂⁻.

In order to determine the effective reductants, the removal of NO₃⁻ was compared with different hole scavengers, including H₂O, HCOOH, KI, and EDTA + HCOOH. As shown in Figure 5b, the NO₃⁻ removal efficiency follows the order of HCOOH > KI > EDTA + HCOOH > H₂O. Different reducing substances were produced using different hole scavengers, as shown in

Table 2. Main reducing substances in different hole scavengers.

Hole Scavenger	Main Reducing Substances
HCOOH	·CO2−, photo-generated e−, HCOO−, ·OH
H2O	·OH
KI	Photo-generated e−
EDTA+HCOOH	Photo-generated e−, HCOO−

. The reducing substances were photo-generated e⁻, ·CO₂⁻, HCOO⁻, and ·OH with HCOOH as the hole scavenger in the photo-catalytic reduction of NO₃⁻ by Ag-TiO₂ under visible light, and the highest NO₃⁻ removal efficiency was achieved at 82.7%. When H₂O was used as the hole scavenger, the reducing substance was ·OH, and NO₃⁻ removal efficiency was only 1.3%. KI is commonly used as a hole scavenger. When photo-generated e⁻ was the reducing substance, NO₃⁻ removal efficiency was only 7.2%, as show in

Table 3. The removal efficiency of NO₃⁻ with different concentrations of KI.

KI (mmol/L)	Removal Efficiency of NO3− (%)
1	6.8
2	7.1
20	7.2

. EDTA is a frequently used radical quencher, which can quench ·CO₂⁻ and ·OH. Compared with the HCOOH system, the main reductants in the EDTA + HCOOH system are photo-generated e⁻ and HCOO⁻. NO₃⁻ removal efficiency was 5.2% in the EDTA + HCOOH system. Combining the reducing substances and NO₃⁻ removal efficiency under a different hole scavenger system, the main reductant is ·CO₂⁻.

On the basis of the above discussion, the mechanism of photo-catalytic reduction of NO₃⁻ by Ag-TiO₂/HCOOH under visible light was revealed. (1) When Ag-TiO₂ is irradiated with visible light, high light quantum yield is produced. (2) The hot electrons generated by Ag can be effectively transferred to the energy band of TiO₂, contributing to a high separation between e⁻ and h. (3) h⁺ reacts with H₂O to generate ·OH; then, h⁺ reacts with ·OH and HCOOH to generate ·CO₂⁻. (4) ·CO₂⁻ is captured by NO₃⁻, and NO₃⁻ is reduced to NO₃⁻ and finally to N₂ and/or NH₄⁺.

3.4. Effect of Co-Anions

In groundwater, anions like sulphate (SO₄²⁻), chloride (Cl⁻), bicarbonate (HCO₃⁻), and phosphate (H₂PO₄⁻) are present. This work investigated the influence of these co-anions on photo-catalytic NO₃⁻ reduction and selectivity of N₂ under visible light.

The NO₃⁻ removal efficiency was effected by SO₄²⁻, Cl⁻, and H₂PO₄⁻, while HCO₃⁻ did not affect NO₃⁻ removal efficiency, as shown in Figure 6. The higher the concentration of SO₄²⁻, Cl⁻, and H₂PO₄⁻, the stronger the inhibition effect on NO₃⁻ reduction. This result is consistent with Jessica et al.’s work [51]. This result aligns with the surface anionization effects suggested by Sheng et al. that SO₄²⁻ and H₂PO₄⁻ adsorbed h⁺, which blocked the adsorption and degradation of HCOO⁻ [52].

Figure 7a–d show the selectivity of the reduction products (NO₂⁻, NH₄⁺, and N₂) of NO₃⁻ under different co-anions. The existence of Cl⁻ was adverse for the reduction of NO₃⁻ but was conducive to the reduction of NO₃⁻ to N₂.

H₂PO₄⁻ led to an increase in NO₂⁻, as shown in Figure 7c. The reason is that H₂PO₄⁻ can release H⁺, and PO₄²⁻ and PO₄²⁻ compete with NO₃⁻ for electrons, and NO₃⁻ is reduced to NO₂⁻ rather N₂ [40]. SO₄²⁻ can act as an electron acceptor and competed withs NO₃⁻ for electrons, resulting in a decreased NO₃⁻ removal efficiency. HCO₃⁻, as the hole scavenger, can capture h⁺ and ·OH to form ·CO₂⁻, which improves the removal efficiency of NO₃⁻. Meanwhile, the existence of HCO₃⁻ contributes to a high buffer capacity. When the pH is greater than 6.25, TiO₂ will be deprotonated, resulting in a negative charge on the TiO₂ surface. Due to electrostatic repulsion, the adsorption of NO₃⁻ and HCOO⁻ on the surface of TiO₂ will be hindered, resulting in the reduction of ·CO₂⁻ and reduced NO₃⁻ removal efficiency.

4. Conclusions

The photocatalytic reduction of NO₃⁻ by Ag-TiO₂/HCOOH under visible light was efficient, and the reduction reaction conformed to the first-order kinetic model. The removal efficiency of NO₃⁻ firstly increased and then decreased with the increase in the concentration of Ag-TiO₂, NO₃⁻, and HCOOH. The concentration of Ag-TiO₂, NO₃⁻, and HCOOH affected the removal efficiency of NO₃⁻ and the selectivity of NH₄⁺, NO₂⁻, and N₂. All of them influenced the generation of ·CO₂⁻, which was confirmed in the major reductants. When the concentration of Ag-TiO₂ was 1.0 g/L, the concentration of HCOOH was 20.05 mmol/L, the removal efficiency of NO₃⁻ (50 mg-N/L) was the highest at 84.47%, and N₂ conversion efficiency reached 82.68% in 120 min. Except for HCO₃⁻, the existence of co-anions affected NO₃⁻ reduction, and the higher the concentration of anions, the stronger the inhibition effect. These results demonstrated that the method of producing ·CO₂⁻ by Ag-TiO₂ and HCOOH could realize efficient removal of NO₃⁻ and selectivity of N₂ in groundwater under visible light.