English

• Nitrate (NO₃^–) contamination of surface and groundwater has emerged as a pressing global concern, predominantly attributed to human activities, such as the excessive application of nitrogen-based fertilizers, improper treatment of domestic and industrial wastewater discharge, leaching from livestock manure, and landfill leachate [1,2]. Elevated levels of NO₃^– in water can induce eutrophication, trigger red tide, and deplete oxygen levels, resulting in suffocation and mortality of aquatic organisms [3,4]. Trace amounts of NO₃^– in drinking water can be converted into toxic NO₂^– within the human body, resulting in cyanosis (methemoglobinemia), fatigue, rapid breathing, cerebral hypoxia, and even fatality [5,6]. Additionally, NO₃^– in drinking water can also damage the human endocrine system, potentially causing liver damage and cancer [7,8]. Thus, urgent development of treatment processes and technologies for the harmless removal and resource utilization of NO₃^– in water is imperative.

  Electrochemical reduction presents a promising technique for the removal of NO₃^– from water. Only electrons and active hydrogen (*H) provided through electrical energy are utilized as reducing agents, eliminating the need for additional reducing agents [9]. This approach demonstrates high treatment efficiency, relatively low investment costs, mild operating conditions, and negligible generation of secondary hazardous substances [10]. However, the electrochemical reduction of NO₃^– entails a complex process that may lead to the formation of various nitrogen-containing species, with typical reaction intermediates and products including nitrite (NO₂^–), nitric oxide (NO), nitrous oxide (N₂O), nitrogen (N₂), hydrazine (N₂H₄), and ammonia (NH₃) [10,11], facilitated by adjustable potential/current (Eqs. 1-3) [12]. Hence, achieving efficient removal and highly selective conversion of NO₃^– in treating surface water, groundwater, or wastewater has long been a paramount objective. Among these products, N₂ and NH₃ are the most thermodynamically stable states of nitrogen, especially under neutral pH conditions [13]. Therefore, N₂ and NH₃ are commonly produced as end products in the electrocatalytic reduction of nitrates. Considering NH₃ tends to dissolve in water, resulting in the formation of ammonium ions (NH₄⁺) which may also be a potential pollutant, from an environmental perspective, the most desirable approach for NO₃^– removal with low concentration levels involves selectively reducing it to harmless N₂ [14]. Therefore, it is imperative to screen appropriate electrocatalysts to enhance N₂ selectivity in the electroreduction of NO₃^–.

  (1)

  (2)

  (3)

  In order to enhance the efficiency of NO₃^– conversion and N₂ selectivity, metal catalysts are widely employed in the *H-driven electrocatalytic reduction of NO₃^–. Noble metal electrodes, such as Pd, Pt, and Ru, are notable examples due to their good adsorption activation capabilities for *H [15,16]. Transition metals such as Cu, Fe, Co and Ni have also exhibited considerable catalytic activity in the nitrate reduction reaction [17]. The preferential conversion of NO₃^– to N₂ is commonly facilitated by Pd and Pt electrocatalysts, which are recognized as the most selective catalysts for producing N₂ [18]. Pd-based materials are favored by researchers due to the relatively higher cost-effectiveness to Pt. However, Pd catalysts may not undergo effective transformation of NO₃^–, which often restricts the overall production rate. In this regard, a majority of the related studies have investigated the bimetallic multiphase catalysts, which could offer advantages of not only high N₂ selectivity but also easily controllable catalytic active sites [5,19,20]. Introducing other metallic or nonmetallic elements through doping or substitution can alter the electronic structure of Pd and induce strain in the Pd lattice, which could significantly contribute to the enhanced electrocatalytic activity and/or stability of the catalyst [21]. For example, PdCu bimetallic catalysts exhibit excellent catalytic activity and N₂ selectivity due to Cu enhancing NO₃^– reduction kinetics and Pd promoting N₂ selectivity [22,23].

  To address the restriction of high cost associated with noble metal electrodes in practical applications, using inexpensive conductive materials as substrates to prepare composite electrodes is a feasible approach [5]. Much of the current research primarily involves depositing noble metal catalysts on carbon paper through drop-casting. The poor dispersibility of the catalyst on carbon paper and the weak interaction between the catalyst and the substrate may lead to lower catalyst utilization efficiency and poor stability. Highly ordered and photo-responsive TiO₂ nanotube arrays, distinguished by their unique nanotube structure and large surface area, as well as outstanding stability, optical, and electronic properties, are regarded as a catalyst support with high application potential [24,25]. Moreover, TiO₂ nanotube arrays, directly grown on a Ti substrate without the need for additional conductive additives and binders, offer advantageous for charge transfer. Notably, the properties of TiO₂ nanotubes, such as electrical conductivity, surface charge, and band gap, can be precisely controlled by constructing oxygen defects to meet specific requirements [26]. Therefore, it is reasonable to believe that TiO₂ nanotube arrays, as a carrier for metal catalysts, can be tailored to exhibit excellent catalytic activity.

  Furthermore, despite significant progress in electrocatalytic NO₃⁻ reduction technology, its substantial overpotential remains a major obstacle to achieving efficient NO₃⁻ conversion [27]. Introducing external energy, such as light, in electrocatalysis is expected to accelerate electrochemical interface reactions and reduce overpotential [28]. Researches have reported that, compared to standalone electrocatalysis, photo-assisted electrocatalysis exhibits high energy efficiency in the activation of small molecules at low overpotential [29-31]. Simultaneously, light irradiation may interfere with the electronic properties of electrocatalysts, such as electron migration, band bending, charge distribution, Fermi level, and intermediate desorption energy. These factors could alter catalytic pathways and performance [32]. Therefore, an in-depth exploration of the performance changes and mechanisms for photo-assisted electrocatalytic NO₃^– reduction system is required.

  In this study, defective TiO_x nanotube arrays were utilized as the substrate, and PdCu/TiO_x composite catalysts were prepared via electrochemical deposition. The morphology, structure, and PdCu loading of the catalyst were characterized. The NO₃^– reduction efficiency and N₂ selectivity with different Pd-Cu ratios were investigated under both electrocatalysis and photo-assisted electrocatalysis conditions. The mechanism of NO₃^– reduction in the photo-assisted electrocatalysis system was explored. Finally, the application potential and economic feasibility of the photo-assisted electrocatalytic technology for selectively reducing NO₃^– to N₂ were analyzed and evaluated.

  The synthesis and characterization of PdCu/TiO_x composite catalysts was stated in detail in the materials and method part in Supporting information. After electrochemical reduction, the color of the electrode changed from silver-gray to deep blue (Fig. S1 in Supporting information), indicating the formation of oxygen vacancy defects on the surface of TiO₂ nanotubes [33]. A top-view scanning electron microscopy (SEM) image of the Pd₁Cu₁/TiO_x nanotube arrays (Fig. 1a) showed a highly ordered morphology with open-ended nanotubes, similar to the TiO₂ nanotube arrays. The nanotubes had an approximate length of 5 µm as observed in the cross-sectional image (Fig. 1b). Transmission electron microscopy (TEM) images revealed a nanotube diameter of approximately 70–100 nm in the Pd₁Cu₁/TiO_x samples (Figs. 1c and d). However, no PdCu nanoparticles were observed, indicating that PdCu might be loaded on the nanotubes as sub-nanometer clusters or individually dispersed atoms. High-resolution TEM (HRTEM) analysis confirmed the presence of the (101) crystal plane of anatase TiO₂ with a lattice fringe spacing of 0.35 nm in the Pd₁Cu₁/TiO_x (Fig. 1e) [34]. High-angle annular dark-field scanning transmission electron microscopy (HADDF-STEM) and energy-dispersive X-ray spectroscopy (EDS) elemental mapping images further verified the successful deposition of Pd and Cu metals, which was evenly distributed on the nanotubes (Figs. 1f–j). The uniform distribution of PdCu on the nanotubes implies an increased exposure of active sites, thus benefiting catalytic reactions [35]. The atomic percentages of Pd and Cu on the Pd₁Cu₁/TiO_x nanotube arrays were determined to be 0.15 at% and 0.17 at%, respectively (Table S1 in Supporting information).

  Figure 1

  

  Figure 1.  Characterization of Pd₁Cu₁/TiO_x nanotube arrays. (a) SEM top view, (b) SEM cross-sectional views, (c, d) TEM images, (e) HRTEM image, (f) HAADF-STEM image and EDS element mapping: (g) Ti, (h) O, (i) Pd, and (j) Cu.

  DownLoad: Full-Size Img PowerPoint

  The structure of the catalysts was revealed through X-ray characterization techniques. The X-ray diffraction (XRD) patterns of TiO₂ and Pd₁Cu₁/TiO_x nanotube arrays exhibit similar diffraction peaks at around 2θ = 25.5°, 38.1°, 48.2°, 54.2°, 55.2° (Fig. 2a), corresponding to the (101), (004), (200), (105), (211) crystal planes of anatase TiO₂ (PDF #75–1537) [36]. Additionally, two additional diffraction peaks at 2θ = 62.5° and 71.6° correspond to the (110) and (103) crystal planes of Ti [37]. Notably, no peaks corresponding to the crystal planes of palladium oxide or copper oxide were detected, indicating the successful incorporation of Pd and Cu ions into the TiO₂ lattice. This can be attributed to the presence of oxygen vacancy defects on the surface of the TiO_x, which facilitates the capture and stabilization of metal atoms [38]. Electron paramagnetic resonance spectroscopy (EPR) spectra show that the TiO_x exhibited a prominent signal peak at a g-value of 2.005 (Fig. 2b), indicating the abundant formation of oxygen defect sites on the surface of the nanotube arrays after electrochemical reduction [39]. The O 1s signals also confirm the presence of oxygen vacancies in TiO_x (Fig. S2 in Supporting information). The O 1s peak of TiO_x can be deconvoluted into three peaks at 530.29, 530.90 and 532.17 eV, corresponding to lattice oxygen (O-metal), oxygen vacancy (O_V) and adsorbed oxygen (Ads. O) [40,41]. Fig. 2c demonstrates the high-resolution X-ray photoelectron spectroscopy (XPS) spectra of TiO₂ and Pd₁Cu₁/TiO_x nanotube arrays in the Ti 2p region. The TiO₂ nanotube arrays exhibit two characteristic peaks at 458.68 eV and 464.46 eV, corresponding to Ti 2p_3/2 and Ti 2p_1/2, respectively. In contrast, the Pd₁Cu₁/TiO_x nanotube arrays show a shift in the Ti 2p_3/2 and Ti 2p_1/2 characteristic peaks towards lower binding energies by 0.25 and 0.27 eV, respectively. This shift is attributed to the presence of Ti³⁺ and oxygen vacancies on the surface of the sample [34].

  Figure 2

  

  Figure 2.  Structure of catalysts. (a) XRD spectra, (b) EPR spectra, (c) high-resolution Ti 2p XPS spectra, (d) high-resolution Pd 3d XPS spectra, (e) high-resolution Cu 2p XPS spectra, and (f) UV–vis absorption spectra.

  DownLoad: Full-Size Img PowerPoint

  The oxidation states of PdCu loaded on nanotubes were investigated using high-resolution XPS analysis in the Pd 3d region and Cu 2p region separately. Fig. 2d shows distinct peaks at 334.8 eV and 340.2 eV, corresponding to Pd 3d_5/2 and Pd 3d_3/2, respectively [35]. Further peak fitting analysis reveals characteristic peaks of Pd⁰ at binding energies of 334.7 and 340.2 eV, indicating that the majority of Pd was loaded on the support in the metallic form. Additionally, small signal peaks at binding energies of 336.1 and 342.2 eV correspond to Pd²⁺, indicating the presence of a small amount of Pd in the form of PdO [42]. Fig. 2e shows the Cu 2p XPS spectrum of Pd₁Cu₁/TiO_x. The oxidation state of Cu is often determined by the position of the Cu 2p_3/2 peak [43]. The Cu 2p_3/2 spectrum was deconvoluted into signal peaks at 932.2 and 934.5 eV. The peak at 932.2 eV was attributed to Cu⁰ or Cu¹⁺, while the peak at 934.5 eV was attributed to Cu²⁺ [44]. In addition, satellite peaks of Cu²⁺ were observed at 941.1 eV, 943.7 eV, and 962.4 eV, indicating that the deposited Cu species on the nanotubes primarily existed in the form of elemental Cu and its oxides. The atomic percentages of Pd and Cu on the Pd₁Cu₁/TiO_x nanotube arrays obtained from XPS results were 0.17 at% and 0.21 at%, respectively (Table S1), slightly higher than the EDS results. This difference can be attributed to the higher sensitivity of XPS to the sample surface [45]. Fig. 2f shows the ultraviolet–visible spectroscopy (UV–vis) absorption spectra of TiO₂ and TiO_x nanotube arrays. The reduced TiO_x exhibited enhanced light absorption in the 400–600 nm visible range compared to TiO₂ nanotubes, which is attributed to the formation of a high concentration of oxygen vacancies on the nanotube surface and the induction of Ti³⁺ centers through electroreduction treatment. These defects lead to the generation of defect levels below the conduction band, imparting strong visible light absorption capability to the TiO_x nanotube arrays [46]. In summary, defective TiO_x nanotube arrays were successfully prepared. The abundant oxygen vacancies on the surface served as capture sites for PdCu metal atoms, ensuring their uniform distribution on the nanotube support. Additionally, these defects enhanced the responsiveness to visible light, facilitating catalytic reactions driven by light energy.

  To screen for the catalyst with optimal activity, the NO₃^– electrocatalytic reduction efficiency and product selectivity of TiO₂, TiO_x, Cu/TiO_x, Pd/TiO_x, Pd₁Cu₂/TiO_x, Pd₁Cu₁/TiO_x and Pd₂Cu₁/TiO_x nanotube arrays were investigated, respectively (Fig. S3 in Supporting information). Individual TiO₂ and TiO_x nanotube arrays exhibited weak kinetics for NO₃^– reduction. However, the introduction of Cu onto the TiO_x significantly enhanced the rate of NO₃^– reduction. Notably, the concentration of NO₂^– increased visibly during the reaction in the Cu/TiO_x system, suggesting that Cu sites primarily facilitate the conversion of NO₃^– to NO₂^– [23,47]. On the other hand, the Pd/TiO_x nanotube arrays displayed the weakest NO₃^– reduction performance, indicating that individual Pd is not efficient in promoting the reduction of NO₃^–. This could be ascribed to the exceptional hydrogen spill-over capabilities of Pd [48,49]. However, when Pd and Cu were simultaneously loaded onto the TiO_x nanotube arrays, PdCu/TiO_x composite catalysts exhibited faster kinetics of NO₃^– reduction with minimal generation of NO₂^– throughout the entire experimental process (<1%). The incorporation of Cu can modify the electronic configuration and the d-band center of Pd, thereby enhancing the catalytic efficacy [50]. As shown in Fig. 3a, after 75 min of reaction, the Pd₁Cu₁/TiO_x catalyst showed the highest NO₃^– conversion rate (81.2%) and N₂ selectivity (67.2%) compared to other catalysts. Therefore, Pd₁Cu₁/TiO_x catalyst was selected for further study on the photo-assisted electrocatalytic reduction of NO₃^–.

  Figure 3

  

  Figure 3.  NO₃^– removal rate and N₂ selectivity after 75 min of reaction: (a) by different electrocatalysts, (b) by Pd₁Cu₁/TiO_x at different voltages in photo-assisted electrocatalytic (PEC) system, and (c) under Pd₁Cu₁/TiO_x photocatalytic (PC), electrocatalytic (EC) and PEC modes. (d) Durability test of PEC of Pd₁Cu₁/TiO_x. Experimental condition: [NO₃^––N] = 25.2 mg/L, voltage = –0.45 V vs. reversible hydrogen electrode (RHE), irradiation wavelength (λ) > 320 nm, initial pH 6.8.

  DownLoad: Full-Size Img PowerPoint

  Considering the excellent light-responsive properties of TiO_x, the introduction of light holds promise in altering the electronic performance of Pd₁Cu₁/TiO_x catalyst, thereby changing its intrinsic activity and selectivity [28]. Fig. S4 (Supporting information) illustrates the photo-assisted electrocatalytic NO₃^– reduction performance and product generation of the Pd₁Cu₁/TiO_x catalyst at voltages ranging from –0.35 V to –0.55 V vs. reversible hydrogen electrode (RHE). Similar to the electrocatalytic NO₃^– reduction, the final products under the photo-assisted electric field still consist mainly of N₂ and NH₃. It is noteworthy that the introduction of light remarkably improved the removal rate of NO₃^– and the selectivity for N₂. At a voltage of –0.35 V vs. RHE, the photo-assisted electrocatalytic NO₃^– conversion rate and the N₂ selectivity could reach 81.7% and 75.8% respectively (Fig. 3b), which was superior to what is achieved under electrocatalysis at –0.45 V vs. RHE, indicating that introducing light can contribute to reducing the reaction overpotential. At the voltage of –0.45 V vs. RHE, the photo-assisted electrocatalysis (PEC) exhibited the highest NO₃^– conversion rate of 95.1% and a N₂ selectivity of approximately 90%, much better than that under either sole photocatalysis (PC) or electrocatalysis (EC) modes (Fig. 3c). The significantly enhanced NO₃^– reduction performance indicates that the electron transfer capability and intrinsic activity of Pd₁Cu₁/TiO_x can be improved through photo-assisted methods [51]. In addition to its excellent NO₃^– reduction activity and N₂ selectivity, the Pd₁Cu₁/TiO_x nanotube arrays also demonstrated good durability. In nine repeated PEC tests, the Pd₁Cu₁/TiO_x nanotube arrays did not show any signs of deactivation (Fig. 3d). Further, the XRD patterns and SEM image of Pd₁Cu₁/TiO_x evidenced no significant alterations was found on the structure and surface characteristics of Pd₁Cu₁/TiO_x before and after the durability test, as depicted in Figs. S5 and S6 (Supporting information), respectively.

  To investigate the mechanism behind the improved performance of NO₃^– reduction in PEC system, a photoelectrochemical analysis was conducted. The linear sweep voltammetry (LSV) curve indicates that the current density under PEC conditions was higher than that under EC conditions (Fig. 4a). Furthermore, the current density was further enhanced when NO₃^– was added to the electrolyte, suggesting the occurrence of a noticeable NO₃^– reduction reaction on the cathode [52]. The transient photocurrent curve of Pd₁Cu₁/TiO_x exhibited corresponding increases and decreases in photocurrent with the periodic opening and closing of light illumination (Fig. 4b), indicating its excellent photo-responsive capability. Moreover, the photocurrent response was more pronounced with the addition of NO₃^–, implying the participation of photogenerated electrons in the NO₃^– reduction reaction. Comparing the electrochemical impedance spectroscopy (EIS) of Pd₁Cu₁/TiO_x under EC and PEC conditions, it is evident that the introduction of an external light field reduced the interfacial charge transfer resistance of the catalyst (Fig. 4c). These results indicate that the introduction of a light field not only significantly reduced the reaction overpotential but also accelerated the electron transfer rate at the catalyst interface, thereby improving the removal efficiency of NO₃^–.

  Figure 4

  

  Figure 4.  Mechanism of NO₃^– reduction in PEC system with Pd₁Cu₁/TiO_x catalyst. (a) LSV curves, (b) photocurrent curves, (c) EIS spectra, (d) EPR spectra, (e) effects of different tert-butyl alcohol (TBA) concentrations, and (f) comparison of the efficiency in deionized water and secondary effluent of a wastewater treatment plant (WWTP) by Pd₁Cu₁/TiO_x after 75 min of PEC reaction.

  DownLoad: Full-Size Img PowerPoint

  Research has shown that electrocatalytic reduction of NO₃^– can primarily proceed through two pathways: direct electron transfer reduction and a *H-mediated indirect reduction pathway [53-55]. The direct electron reduction pathway primarily occurs on non-precious metal catalyst sites, such as Cu and Cu/Ni, due to their strong NO₃^– adsorption capability [56]. On the other hand, the *H-mediated NO₃^– reduction pathway is more likely to occur on precious metals, such as Pd, Pt, and Rh, which possess a strong hydrogen adsorption capability, thus inhibiting the occurrence of hydrogen evolution side reactions [7]. To validate the hypothesis, the generation of *H and its involvement in the NO₃^– reduction process was investigated using EPR spectroscopy. As shown in Fig. 4d, the EPR spectra of Pd₁Cu₁/TiO_x system exhibited nine distinct characteristic signal peaks with intensity ratio approaching 1:1:2:1:2:1:2:1:1 and the calculated hyperfine coupling constants as α_N = 16.3 G and α_H = 22.5 G, which corresponds to the adduct of 5,5-dimethyl-1-pyrroline-N-oxide (DMPO) with atomic *H [53,57]. This suggests that atomic *H generated during the reaction desorbed from the catalyst surface and was captured by DMPO in the reaction solution [57]. To further investigate the contribution of *H in the NO₃^– reduction, tert-butyl alcohol (TBA) was applied as a radical scavenger [58]. The results showed that increasing the TBA concentration in the reaction solution greatly suppressed NO₃^– reduction (Fig. 4e). The removal rate of NO₃^– decreased by approximately 21% when TBA was added at 1 mmol/L, and the removal rate further decreased by 37% when TBA concentration was increased to 10 mmol/L. However, increasing the TBA concentration to 20 mmol/L showed little difference in the NO₃^– removal rate, indicating that *H in the solution may be almost completely inhibited [7]. It is worth noting that while TBA as a scavenger inhibits the reaction of NO₃^– and free *H which desorbs from the catalyst into the solution, it may not completely eliminate the *H adsorbed on the surface of the Pd₁Cu₁/TiO_x catalyst. This adsorbed *H still plays a crucial role in the removal of NO₃^–. Therefore, it is reasonable to speculate that, in the PEC reduction of NO₃^–, at least 37% of the NO₃^– reduction occurs through the *H-mediated indirect pathway.

  Based on the obtained results and discussions, the reduction mechanisms of NO₃^– on the Pd₁Cu₁/TiO_x catalyst can be proposed. The Pd active sites dissociate nearby water molecules through the Volmer mechanism to generate *H (Eq. 4), therefore the pH of the solution in the cathode chamber gradually increased from 6.80 to 11.69 as the reaction proceeds (Fig. S7 in Supporting information). Meanwhile, Cu sites, as strong Lewis acid sites, change the electronic structure of the catalyst surface and initially adsorb NO₃^– ions in the solution [59]. Then the overflow of *H at Pd sites facilitates the reduction of *NO₃^– to *NO₂^– (Eq. 5). The high reduction ability of *H (E ⁰ (H⁺/H) = –2.31 V vs. SHE) enables the cleavage of the N—O bond in NO₂^–, leading to the gradual transformation into *N on the Pd surface (Eqs. 6 and 7). Finally, N₂ is formed through N—N coupling (Eq. 8) [49,60].

  (4)

  (5)

  (6)

  (7)

  (8)

  The economic feasibility of the PEC system for selectively reducing NO₃^– was assessed. High removal efficiency and N₂ selectivity make the designed PEC system with the Pd₁Cu₁/TiO_x nanotube arrays as the cathode promising for treating NO₃^–-containing wastewater due to its superiority over other systems equipped with Pd or Cu-based electrocatalysts (Fig. S8 in Supporting information). Taking into account that in practical applications, light energy can be derived from sunlight, which would be no cost involved, the total energy cost was mainly attributed to the electrical energy cost to achieve the highest efficiency in converting NO₃^–-N to N₂ and the catalyst preparation [61]. Using the formula Eqs. S5 and S6 in Supporting information, the energy consumption required for Pd₁Cu₁/TiO_x nanotube arrays to reduce 25.2 mg/L of NO₃^–-N in water to N₂ was calculated to be 2.74 kWh/mol, with a corresponding E_EO of 0.79 kWh/m³, which is comparable to the systems using other electrocatalysts (Fig. S9 in Supporting information). Considering the potential utilization of inexpensive renewable energy sources such as solar or wind power, studies have indicated that the price of wind power in the next 10 years is estimated to be around $0.02 per kilowatt-hour (kWh) [62]. Therefore, the corresponding total electrical energy cost for the Pd₁Cu₁/TiO_x nanotube arrays was $0.0944/m³, presenting a significantly competitive or economically advantageous option compared to other water treatment technologies (Table 1) [8]. Importantly, this reaction system does not necessitate the addition of Cl^– to enhance the selectivity of N₂, thus reducing water purification costs while eliminating the production of toxic by-products containing Cl.

  Table 1

  

  Table 1.  Cost of nitrate removal by different water treatment technologies [8].

  DownLoad: CSV

  In order to investigate the efficacy of PdCu/TiO_x catalysts in selectively reducing NO₃^– to N₂ in complex real water, the potential impact of alkali metal ions and organic matter present in the water on NO₃^– reduction must be considered. Consequently, the secondary effluent from a wastewater treatment plant (WWTP) was subjected to treatment using PdCu/TiO_x under the PEC mode. The main water quality parameters of this water are shown in Table S3 (Supporting information). For consistency with previous experimental results, suspended solids in the water were initially filtered using a 0.45 µm membrane, and NO₃^– was added to elevate the concentration to 25.2 mg/L. As shown in Fig. 4f, under the PEC mode, the Pd₁Cu₁/TiO_x catalyst demonstrated commendable treatment performance for the secondary effluent. Following 75 min of reaction, the removal efficiency of NO₃^– reached 95.3%, with a selectivity for N₂ of 83.7%. These results closely resemble those observed in deionized water, indicating minimal impact of the presence of organic matter or inorganic ions in the secondary effluent on NO₃^– reduction.

  The TiO_x nanotube arrays feature enriched oxygen defect sites to capture PdCu metals, ensuring uniform dispersion on the carrier surface. Consequently, the Pd₁Cu₁/TiO_x composite catalyst showed optimal electrocatalytic activity, and the catalytic activity was further enhanced when light was introduced due to the remarkably reduced overpotential for the interface active *H formation and accelerating electron migration. The cost-effectiveness of the photo-assisted Pd₁Cu₁/TiO_x electrocatalytic system demonstrated competitiveness and economic advantages compared to other water treatment technologies.

  Although with high efficiency and selectivity, electrochemical nitrate reduction is currently still limited to small-scale laboratory experiments and has not yet been adapted for large-scale industrial or commercial uses, nor for practical implementation in real-world aquatic settings. As the stability and cost of the electrocatalysts are highly required in the industrial and commercial applications, it is of greater importance to do longer stability tests on the studied electrocatalysts to make such electrocatalysts more suitable for large-scale industrial utilization. Methods to synthesize electrocatalysts also need to improve for larger-scale usage and lower costs, such as using milder synthesis techniques, designing self-supporting and integrated catalytic materials. In future research, consideration should be given to the investigation of corrosion, leaching, and surface poisoning that may occur during the long-term operation of the catalyst, which may lead to a decrease in reaction activity and potential adverse effects on the environment. Meanwhile, the removal of nitrates in surface and groundwater with low conductivity may require additional electrolyte supplementation, which needs to be overcome in the future. Additionally, for water with more complex qualities, the combined use with other water treatment processes may be necessary to increase practicality.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  CRediT authorship contribution statement

  Yuxin Zeng: Writing – original draft, Visualization, Formal analysis, Data curation. Yan Luo: Resources, Methodology, Investigation. Yao He: Resources, Methodology, Investigation. Kaihang Zhang: Writing – review & editing, Validation, Conceptualization. Binbin Zhu: Writing – review & editing, Validation, Conceptualization. Yuanzheng Zhang: Writing – review & editing, Supervision, Project administration, Funding acquisition, Conceptualization. Junfeng Niu: Writing – review & editing, Supervision, Project administration.

  Acknowledgments

  This work was funded partly by the National Natural Science Foundation of China (No. 52300084), China Postdoctoral Science Foundation (No. 2023M741151), and the Fundamental Research Funds for the Central Universities (No. 2024MS063).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2024.110514.

  
  
• 1. [1]

     X. Zhang, Y. Wang, C. Liu, et al., Chem. Eng. J. 403 (2021) 126269. doi: 10.1016/j.cej.2020.126269

  2. [2]

     Y. Xiong, Y. Wang, J. Zhou, et al., Adv. Mater. 36 (2024) 2304021. doi: 10.1002/adma.202304021

  3. [3]

     J. Lim, C.Y. Liu, J. Park, et al., ACS Catal. 11 (2021) 7568–7577. doi: 10.1021/acscatal.1c01413

  4. [4]

     X. Zou, C. Chen, C. Wang, et al., Sci. Total Environ. 800 (2021) 149645. doi: 10.1016/j.scitotenv.2021.149645

  5. [5]

     S. Yuan, Y. Xue, R. Ma, et al., Sci. Total Environ. 866 (2023) 161444. doi: 10.1016/j.scitotenv.2023.161444

  6. [6]

     M. Fu, Y. Mao, H. Wang, et al., Chin. Chem. Lett. 35 (2024) 108341. doi: 10.1016/j.cclet.2023.108341

  7. [7]

     Y.J. Shih, Z.L. Wu, C.Y. Lin, et al., Appl. Catal. B: Environ. 273 (2020) 119053. doi: 10.1016/j.apcatb.2020.119053

  8. [8]

     H. Liu, J. Park, Y. Chen, et al., ACS Catal. 11 (2021) 8431–8442. doi: 10.1021/acscatal.1c01525

  9. [9]

     Y. Ren, S. You, Y. Wang, et al., Environ. Sci. Technol. 58 (2024) 2144–2152. doi: 10.1021/acs.est.3c09759

  10. [10]

      J. Zhou, S. Gao, G. Hu, Energy Fuel. 38 (2024) 6701–6722. doi: 10.1021/acs.energyfuels.4c00415

  11. [11]

      M.S. Iqbal, Z.B. Yao, Y.K. Ruan, et al., Rare Metal. 42 (2023) 1075–1097. doi: 10.1007/s12598-022-02215-7

  12. [12]

      Y. Zeng, C. Priest, G. Wang, et al., Small Method. 4 (2020) 2000672. doi: 10.1002/smtd.202000672

  13. [13]

      S. Meng, Y. Ling, M. Yang, et al., J. Environ. Chem. Eng. 11 (2023) 109418. doi: 10.1016/j.jece.2023.109418

  14. [14]

      A.S. Fajardo, P. Westerhoff, C.M. Sanchez-Sanchez, et al., Appl. Catal. B: Environ. 281 (2021) 119465. doi: 10.1016/j.apcatb.2020.119465

  15. [15]

      Y. Wang, C. Wang, M. Li, et al., Chem. Soc. Rev., 50 (2021) 6720. doi: 10.1039/d1cs00116g

  16. [16]

      F. Ni, Y. Ma, J. Chen, et al., Chin. Chem. Lett. 32 (2021) 2073–2078. doi: 10.1016/j.cclet.2021.03.042

  17. [17]

      Y. Liu, Y. Zheng, Y. Ren, et al., Environ. Sci. Technol. 58 (2024) 7228–7236. doi: 10.1021/acs.est.4c01464

  18. [18]

      J. Lim, Y. Chen, D.A. Cullen, et al., ACS Catal. 13 (2023) 87–98. doi: 10.1021/acscatal.2c04841

  19. [19]

      S. Jung, S. Bae, W. Lee, Environ. Sci. Technol. 48 (2014) 9651–9658. doi: 10.1021/es502263p

  20. [20]

      L. Zhang, Y. Wu, Z. Zhu, et al., Biochar 6 (2024) 8. doi: 10.1007/s42773-023-00298-8

  21. [21]

      C.J. Li, G.C. Shan, C.X. Guo, et al., Rare Metal. 42 (2023) 1778–1799. doi: 10.1007/s12598-022-02234-4

  22. [22]

      Z. Shen, D. Liu, G. Peng, et al., Sep. Purif. Technol. 241 (2020) 116743. doi: 10.1016/j.seppur.2020.116743

  23. [23]

      L. Fang, S. Wang, S. Lu, et al., Chin. Chem. Lett. 35 (2024) 108864. doi: 10.1016/j.cclet.2023.108864

  24. [24]

      I. Weidinger, K.H. Ly, H.I. Oener, et al., Angew. Chem. Int. Ed. 57 (2018) 7225–7229. doi: 10.1002/anie.201802597

  25. [25]

      M. Tang, Q. Tong, Y. Li, et al., Chin. Chem. Lett. 34 (2023) 108410. doi: 10.1016/j.cclet.2023.108410

  26. [26]

      Z. Ren, Q. Chen, J. Zhang, et al., Mater. Today Phys. 36 (2023) 101162. doi: 10.1016/j.mtphys.2023.101162

  27. [27]

      M. Duca, M.T.M. Koper, Energy Environ. Sci. 5 (2012) 9726–9742. doi: 10.1039/c2ee23062c

  28. [28]

      D. Yang, G. Wang, X. Wang, Sci. China Mater. 62 (2019) 1369–1373. doi: 10.1007/s40843-019-9455-3

  29. [29]

      H. Yang, D. Yang, Y. Zhou, et al., J. Am. Chem. Soc. 143 (2021) 13721–13730. doi: 10.1021/jacs.1c05580

  30. [30]

      B. Wang, F. Yang, Y. Dong, et al., Chem. Eng. J. 396 (2020) 125255. doi: 10.1016/j.cej.2020.125255

  31. [31]

      M. Zhang, X. Xuan, W. Wang, et al., Adv. Funct. Mater. 30 (2020) 2005983. doi: 10.1002/adfm.202005983

  32. [32]

      D. Yang, H. Yu, T. He, et al., Nat. Commun. 10 (2019) 1–10. doi: 10.1038/s41467-018-07882-8

  33. [33]

      Y. Yang, M.R. Hoffmann, Environ. Sci. Technol. 50 (2016) 11888–11894. doi: 10.1021/acs.est.6b03540

  34. [34]

      T. Wu, H. Zhao, X. Zhu, et al., Adv. Mater. 32 (2020) 2000299. doi: 10.1002/adma.202000299

  35. [35]

      M. Nie, H. Sun, J. Liao, et al., Int. J. Hydrog. Energy 46 (2021) 6441–6447. doi: 10.1016/j.ijhydene.2020.11.135

  36. [36]

      Z. Liu, X. Zhang, S. Nishimoto, et al., J. Phys. Chem. C 112 (2008) 253–259. doi: 10.1021/jp0772732

  37. [37]

      W. Zhou, Z. Yin, Y. Du, et al., Small 9 (2013) 140–147. doi: 10.1002/smll.201201161

  38. [38]

      J. Wan, W. Chen, C. Jia, et al., Adv. Mater. 30 (2018) 1705369. doi: 10.1002/adma.201705369

  39. [39]

      S.A. Rawool, K.K. Yadav, V. Polshettiwar, Chem. Sci. 12 (2021) 4267–4299. doi: 10.1039/d0sc06451c

  40. [40]

      S. Wang, T. He, P. Chen, et al., Adv. Mater. 32 (2020) 2001385. doi: 10.1002/adma.202001385

  41. [41]

      J. Huang, Z. Chen, J. Cai, et al., Nano Res. 15 (2022) 5987–5994. doi: 10.1007/s12274-022-4279-5

  42. [42]

      Y. Li, C. Zhang, J. Ma, et al., Appl. Catal. B: Environ. 217 (2017) 560–569. doi: 10.1016/j.apcatb.2017.06.023

  43. [43]

      Y. Wang, Z. Chen, P. Han, et al., ACS Catal. 8 (2018) 7113–7119. doi: 10.1021/acscatal.8b01014

  44. [44]

      T. Zhang, J. Low, X. Huang, et al., ChemCatChem 9 (2017) 3054–3062. doi: 10.1002/cctc.201700512

  45. [45]

      X. Zhou, I. Hwang, O. Tomanec, et al., Adv. Funct. Mater. 31 (2021) 2102843. doi: 10.1002/adfm.202102843

  46. [46]

      B. Jga, B. Qsa, C. Rg, et al., J. CO₂ Util. 35 (2020) 205–215. doi: 10.1016/j.jcou.2019.09.016

  47. [47]

      A.C.A.d. Vooys, R.A.v. Santen, J.A.R.v. Veen, J. Mol. Catal. A: Chem. 154 (2000) 203–215. doi: 10.1016/S1381-1169(99)00375-1

  48. [48]

      S. Lee, Y. Lee, W. Choi, Appl. Catal. B: Environ. 327 (2023) 122432. doi: 10.1016/j.apcatb.2023.122432

  49. [49]

      H. Shin, S. Jung, S. Bae, et al., Environ. Sci. Technol. 48 (2014) 12768–12774. doi: 10.1021/es503772x

  50. [50]

      Q.-W. Ding, Q. Luo, L. Lin, et al., Rare Metal. 41 (2022) 2595–2605. doi: 10.1007/s12598-022-01997-0

  51. [51]

      Y. Zhou, L. Zheng, D. Yang, et al., Small Method. 5 (2020) 2000991.

  52. [52]

      Y. Wang, W. Zhou, R. Jia, et al., Angew. Chem. Int. Ed. 59 (2020) 5350–5354. doi: 10.1002/anie.201915992

  53. [53]

      W. Zheng, L. Zhu, Z. Yan, et al., Environ. Sci. Technol. 55 (2021) 13231–13243.

  54. [54]

      S. Garcia-Segura, M. Lanzarini-Lopes, K. Hristovski, et al., Appl. Catal. B: Environ. 236 (2018) 546–568. doi: 10.1016/j.apcatb.2018.05.041

  55. [55]

      J. Gao, B. Jiang, C. Ni, et al., Appl. Catal. B: Environ. 254 (2019) 391–402. doi: 10.1016/j.apcatb.2019.05.016

  56. [56]

      Y.J. Shih, Z.L. Wu, Y.H. Huang, et al., Chem. Eng. J. 383 (2020) 123157. doi: 10.1016/j.cej.2019.123157

  57. [57]

      J. Li, G. Zhan, J. Yang, et al., J. Am. Chem. Soc. 142 (2020) 7036–7046. doi: 10.1021/jacs.0c00418

  58. [58]

      C. Ni, Y. Li, X. Meng, et al., Chem. Eng. J. 411 (2021) 128485. doi: 10.1016/j.cej.2021.128485

  59. [59]

      X. Li, Z. Shen, J. Lv, et al., ACS EST Eng. 3 (2023) 2202–2212. doi: 10.1021/acsestengg.3c00052

  60. [60]

      C. Chen, K. Li, C. Li, et al., Environ. Sci. Technol. 53 (2019) 13868–13877. doi: 10.1021/acs.est.9b04447

  61. [61]

      R. Chauhan, V.C. Srivastava, Chem. Eng. J. 386 (2020) 122065. doi: 10.1016/j.cej.2019.122065

  62. [62]

      C. Liu, Y. Wang, R. Zhu, Resour. Conserv. Recycl. 121 (2017) 33–39. doi: 10.1016/j.resconrec.2016.10.001

• 

  Figure 1  Characterization of Pd₁Cu₁/TiO_x nanotube arrays. (a) SEM top view, (b) SEM cross-sectional views, (c, d) TEM images, (e) HRTEM image, (f) HAADF-STEM image and EDS element mapping: (g) Ti, (h) O, (i) Pd, and (j) Cu.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Structure of catalysts. (a) XRD spectra, (b) EPR spectra, (c) high-resolution Ti 2p XPS spectra, (d) high-resolution Pd 3d XPS spectra, (e) high-resolution Cu 2p XPS spectra, and (f) UV–vis absorption spectra.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  NO₃^– removal rate and N₂ selectivity after 75 min of reaction: (a) by different electrocatalysts, (b) by Pd₁Cu₁/TiO_x at different voltages in photo-assisted electrocatalytic (PEC) system, and (c) under Pd₁Cu₁/TiO_x photocatalytic (PC), electrocatalytic (EC) and PEC modes. (d) Durability test of PEC of Pd₁Cu₁/TiO_x. Experimental condition: [NO₃^––N] = 25.2 mg/L, voltage = –0.45 V vs. reversible hydrogen electrode (RHE), irradiation wavelength (λ) > 320 nm, initial pH 6.8.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Mechanism of NO₃^– reduction in PEC system with Pd₁Cu₁/TiO_x catalyst. (a) LSV curves, (b) photocurrent curves, (c) EIS spectra, (d) EPR spectra, (e) effects of different tert-butyl alcohol (TBA) concentrations, and (f) comparison of the efficiency in deionized water and secondary effluent of a wastewater treatment plant (WWTP) by Pd₁Cu₁/TiO_x after 75 min of PEC reaction.

  下载: 全尺寸图片 幻灯片

  Table 1.  Cost of nitrate removal by different water treatment technologies [8].

  下载: 导出CSV
•