Niobium Oxide Photocatalytically Oxidizes Ammonia in Water at Ambient Conditions

Abstract

Ammonia contamination in water is a signifcant environmental issue since it is toxic and leads to eutrophication. Photocatalysis has been investigated as a strategy for ammonia degradation but can potentially form toxic nitrite (NO₂^–) and nitrate (NO₃^–) byproducts. This work reports on the ability of niobium oxide (Nb₂O₅) to photocatalytically oxidize aqueous-phase ammonia (NH₃). Whereas as-synthesized Nb₂O₅ showed little catalytic activity (< 1% NH₃ conversion after 6 h of UV-C irradiation, at room temperature and atmospheric pressure, and under O₂ headspace), Nb₂O₅ treated in basic solution (OH-Nb₂O₅) was able to photocatalytically degrade NH₃ at ca. 9% conversion after six hours, with ca. 70% selectivity to the desired N₂, with a frst-order rate constant of ca. 12 times higher than the as synthesize catalyst (1.6 × 10^–3 min^–1 vs. 2.0 × 10^–2 min^–1). Raman spectroscopic analysis indicated the presence of terminal Nb=O species after base treatment of Nb₂O₅, implicating them as catalytically active sites. These results underscore how a simple structural modifcation can signifcantly affect photocatalytic activity for aqueous ammonia oxidation.

Keywords:
ammonia oxidation; photocatalysis; niobium oxide; aqueous ammonia

Introduction

Ammonia contamination plays a significant role in the eutrophication and reduction of dissolved oxygen in ground and surface waters, which can result in toxicity to aquatic life at very low concentrations (<1 mg NH₃ L^–1).¹ The largest source of ammonia-containing wastewater is the direct discharge of nutrient-rich animal waste, with other sources such as fertilizer manufacturing, food processing, landfill leachate, and wastewater treatment facilities also contributing to the problem.^{2 3 4} Excess ammonia can also pose a problem during drinking water treatment by compromising the efficiency of disinfection processes or by conversion into the more toxic species such as nitrite and nitrate.⁵

Current technologies for the remediation of aqueous ammonia pollution include biological denitrification,⁶ breakpoint chlorination,⁷ and catalytic degradation.⁸ Biological denitrification is typically very slow, limited to low ammonia concentrations, and produces sewage sludge as a byproduct. Breakpoint chlorination is widely used in drinking water disinfection; however, it is only practical for low influent concentrations of ammonia and can cause the formation of toxic disinfection byproducts.⁵,⁷ Catalytic treatment approaches are promising due to their lower byproduct formation, nontoxicity, and flexible operation. Most current aqueous ammonia oxidation catalysts require significant energy input in the form of high pressure, temperature, or electricity. Further, current NH₃ oxidation catalysts suffer from poisoning at low reaction temperatures due to unreactive dehydrogenated N* and NH_x* species on the surface.^{8 9 10}

Advanced photocatalytic oxidation (AOPs) has emerged as a promising and effective technique for the degradation of water contaminants, particularly ammonia.^{11 12 13 14} The method has several advantages, including a broad operational pH range spanning from 2 to 11, utilization of environmentally friendly chemicals, operational simplicity, cost-effectiveness, and exceptional efficiency in mineralization. However, the inherent challenge lies in the substantial band gap of commonly employed semiconductors, such as titania oxide (TiO₂, with an anatase phase band gap of 3.2 eV), which curtails their capacity to absorb visible light. This constraint restricts the use of visible light for catalytic activation. Additionally, certain photocatalysts necessitate UV-C and UV-A light sources to generate crucial reactive species, including hydroxyl radicals $\left(\text{HO}·\right)$, sulfate radicals $\left({\text{SO}}_4^{·-}\right)$, singlet oxygen $\left({}^1\text{O}{}_2\right)$, and superoxide anions $\left({\text{O}}_2^{·-}\right)$.^{15 16}, ¹⁷ Furthermore, the availability of natural sunlight is confined to a limited timeframe, thereby imposing a temporal constraint on the use of this technology.

AOPs for ammonia removal from water has been a subject of thorough investigation, particularly concerning catalysts predominantly derived from TiO₂ and zinc oxide (ZnO). The reaction pH is a critical parameter in affecting the yield and selectivity to dinitrogen (N₂) during NH₃ photo-oxidation.¹⁴ It has been reported to affect NH₃ adsorption on the catalyst surface for TiO₂-based catalysts, where at acidic pH the elevated concentration of hydronium ions increases the surface exchange competition with ammonium (NH₄⁺), with the ideal pH to be between 9.3 (i.e., the pK_a of NH₄⁺) and 10, due to the more favorable adsorption of neutral NH₃. At pH values higher than 10, reaction performance decreases are attributed to the competitive adsorption between hydroxyl anions and NH₃.^{11 12 13 14}

Adding transition metals such as platinum (Pt) to traditional photocatalysts (e.g. TiO₂, ZnO) can also enhance reaction yield and/or selectivity.¹⁴ These enhancements are proposed to be associated with the ability of transition metals to capture e^– from the conduction band,¹⁴,^{18 19 20} and/or better interaction with nitrogenated species.²¹ Ou et al.²² showed that the photodegradation of aqueous ammonia over platinized titanate nanotubes (TNTs) increases from 5 to 85% with the addition of 20 wt.% of Pt, which was proposed to be associated with the stronger absorption affinity of Pt with NH₃/NH₄⁺ than to the unmodified TNTs. While the addition of transition metals generally improves ammonia photooxidation performance, it also increases photocatalyst cost, especially with the elevated loadings of noble metals required.

Nb₂O₅ has a similar band gap to the most widely reported photocatalyst TiO₂ (3.5 eV vs. 3.4 eV, respectively)²³,²⁴ and has been explored as a photocatalyst for carbon dioxide (CO₂) reduction²⁵ as well as water oxidation.²⁴ However, to our knowledge, this semiconductor material has never been studied for aqueous ammonia photooxidation. However, it has shown promise in heterogeneous gas-phase ammonia oxidation at elevated temperatures.²⁶ In this work, we prepared Nb₂O₅ and tested it for aqueous ammonia photocatalysis. While the as-prepared material had low activity, we found that a simple base treatment substantially improved the ammonia photooxidation activity. Through Raman spectroscopy, we identifed structural features to which we attribute the much higher photooxidation activity of the base-treated material.

Experimental

Materials

NH₄Cl, NH₄[NbO(C₂O₄)₂(H₂O)₂]·nH₂O, TiO₂-P25, and NaOH were all purchased from Sigma-Aldrich (St. Louis, MO, USA). Oxygen (99.999%) was purchased from Airgas (Houston, TX, USA). Deionized water (DI water, 18.2 MΩ cm at 25 °C, Thermo Fisher Scientific, Waltham, MA, USA) was used for all experiments.

Catalyst synthesis

The Nb₂O₅ was synthesized following previous reports.²⁷ Briefly, 4.8 g of ammonium niobium oxalate (NH₄[NbO(C₂O₄)₂(H₂O)₂]·nH₂O) were dissolved in 70 mL of DI water, and placed in a Parr Teflon-lined stainless steel autoclave (Moline, IL, USA) for 24 h at 175 °C. The resultant powder was filtered, washed with 1 L of DI H₂O per gram of material, dried overnight at 80 °C, and then calcined under air at 400 °C for 2 h. This as-synthesized Nb₂O₅ was used to prepare a modified Nb₂O₅ (OH-Nb₂O₅) by mixing 2 g of Nb₂O₅ with 10 mL of 1 M NaOH under vigorous stirring for 2 h at room temperature. The resulting solid was washed with DI water (1 L of DI water per gram of material) and then dried overnight at 80 °C.

Ammonium photooxidation experiments

Ammonium photooxidation experiments were carried out in an enclosed photoreactor fabricated in-house and described in previous publications,^{28 29 30 31} equipped with six UV-C Ushio G4T5 low-pressure mercury lamps (Cypress, CA, USA; primary emission wavelength at 254 nm; input 4 W and output 0.8 W). A 120-mL quartz round-bottom flask was used as a vessel for the reaction. The threads were lined with Teflon tape and then capped with a silicone septum, into which an oxygen purge needle and a gas vent needle were inserted.

In a typical batch experiment, 0.4 g L^–1 of catalyst is loaded into the quartz round bottom bottle, and left to stir for 15 min. The pH of the solution was then adjusted to 9 by the addition of aliquots of NaOH stock solution (1 M and/or 10 M). Next, the system was purged with oxygen at 100 mL min^–1 for ca. 15 min before a concentrated solution of NH₄Cl (10 g NH₄⁺ L^–1) was injected to obtain a final solution concentration of 75 mg NH₄⁺ L^–1 (= 4.2 mmol N L^–1). The reaction vessel was left to stir in the dark for one hour, after which the UV-C lamps were turned on and the reaction was carried out for a total of 6 h. 1 mL aliquots were taken via syringe over time.

Measurement of nitrogen species concentrations

The NH₄⁺ concentration in the aqueous phase was measured by ion chromatography (IC, Thermo Fisher Scientific, Waltham, MA, USA) with suppressed conductivity detection (Dionex Aquion, 4 × 250 mm IonPac CS12, CDRS Suppressor at 59 mA) using 20 mM methane sulphonic acid eluent at 1 mL min^–1 flow rate. The limit of detection for this quantification method was 0.1 ppm NH₄⁺. The aqueous NO₃^– and NO₂^– concentrations were also measured by IC (Thermo Fisher Scientific, Waltham, MA, USA; Dionex Aquion, 4 × 250 mm IonPac AS23, AERS 500 Carbonate Suppressor at 32 mA) using a carbonate/ bicarbonate eluent (10 mL of 0.45 M sodium carbonate, 1 mL of 0.45 M sodium bicarbonate and 989 mL of DI water) at 1 mL min^–1 flow rate. The limits of detection for this quantification method were 0.1 ppm NO₂^–, and 0.1 ppm NO₃^–.

Materials characterization

Powder X-ray diffraction (XRD) patterns were collected on a Rigaku D/Max Ultima II diffractometer (Tokyo, Japan; 40 kV, 40 mA) using CuKα radiation (1.5418 Å). The data was collected from 15° to 90° at a rate of 2° min^–1 and step size of 0.04°. Scanning electron microscopy (SEM) images were taken on a JEOL 6500F microscope (Tokyo, Japan). The samples were loaded on a carbon tape and analyzed at 15 kV high tension. Brunauer-Emmett-Teller (BET) surface area measurements were performed on an Autosorb-iQ-MP Quantachrome (Boynton Beach, FL, USA) system measured at liquid-nitrogen temperature (77 K). Prior to the measurements, the samples were degassed for 12 h under vacuum (ca. 2 mmHg) at 200 °C. Raman spectra were taken under ex-situ conditions (i.e., room temperature, ambient air) using a Renishaw inVia microRaman spectrometer (New Mills, UK) with a 532 nm excitation laser and a 40× long working distance objective. Spectra were obtained using 0.05 mW power and 10 s integration times. For each sample, five spectra were acquired at different spots and averaged. The standard deviation between spectra for each sample was less than 5%. The absorption spectra of the photocatalysts were determined utilizing diffuse refectance ultraviolet-visible (DR-UV) spectroscopy. To facilitate measurements, the powders were blended with 60 wt.% BaSO₄, subsequently compacted into wafers, and analyzed with a Shimadzu UV-2450 spectrometer (Kyoto, Japan). diffuse reflectance spectroscopy (DRS) profiles were then applied to determine the optical bandgap energy of the samples using Tauc Plot analysis ((F(R)hv)^{0 5} vs. hv); where h is the Planck’s constant, v is the frequency of incident photons, and F(R) is the Kubelka-Munk function.³²

Selectivity and performance comparison

The percent NH₄⁺ conversion $\left({\text{X}}_{{\text{NH}}_4^{+}}\right)$ and the selectivity to ${\text{NO}}_2^{-}\left({\text{S}}_{{\text{NO}}_2^{-}}\right)$ and ${\text{NO}}_3^{-}\left({\text{S}}_{{\text{NO}}_3^{-}}\right)$ were calculated using equations 1, 2 and 3:

(1) ${\text{X}}_{{\text{NH}}_4^{+}}=\frac{{\left[{\text{NH}}_4^{+}\right]}_0-{\left[{\text{NH}}_4^{+}\right]}_{\text{t}}}{{\left[{\text{NH}}_4^{+}\right]}_0}\times 100$

(2) ${\text{S}}_{{\text{NO}}_2^{-}}=\frac{{\left[{\text{NO}}_2^{-}\right]}_{\text{t}}}{{\left[{\text{NH}}_4^{+}\right]}_0-{\left[{\text{NH}}_4^{+}\right]}_{\text{t}}}\times 100$

(3) ${\text{S}}_{{\text{NO}}_3^{-}}=\frac{{\left[{\text{NO}}_3^{-}\right]}_{\text{t}}}{{\left[{\text{NH}}_4^{+}\right]}_0-{\left[{\text{NH}}_4^{+}\right]}_{\text{t}}}\times 100$

where ${\left[{\text{NH}}_4^{+}\right]}_0,{\left[{\text{NH}}_4^{+}\right]}_{\text{t}},{\left[{\text{NO}}_2^{-}\right]}_{\text{t}},\text{and}{\left[{\text{NO}}_3^{-}\right]}_{\text{t}}$, and [NO^-₃]_t are species molar concentrations and where the subscript 0 refers to the initial time and t refers to a given time. The selectivity to molecular N₂ ${\text{N}}_2\left({\text{S}}_{{\text{N}}_2}\right)$ (was calculated from the total nitrogen mass balance (equation 4).

(4) ${\text{S}}_{{\text{N}}_2}=100-\left({\text{S}}_{{\text{NO}}_2^{-}}+{\text{S}}_{{\text{NO}}_3^{-}}\right)$

The potential formation of gaseous N₂O and NO₂ byproducts was verified to not occur in our reaction system via GC-MS measurements.

Performance comparison of various published catalysts for aqueous ammonia photo-oxidation was assessed using quantum yield (QY) and space-time yield (STY)

as proposed and discussed by Vikrant et al.¹⁴ (equations 5 and 6).

(5) $\text{QY}=\frac{{\text{NH}}_3/{\text{NH}}_4^{+}\text{ decay rate}\left({\text{molecules s}}^{-1}\right)}{\text{photon flux}\left({\text{photon s}}^{-1}\right)}$

(6) $\text{STY}=\frac{\text{QY}}{\text{mass of catalyst}\left(\text{mg}\right)}$

QY values were calculated based on the peak wavelength of the light source reported. When the wavelength was not specified, the commonly used wavelength for visible light and UV light was applied to calculate the performance (420 and 365 nm, respectively). The photon flux for our enclosed photoreactor was measured to be 1.2 × 10¹⁷ photons s^–1 via chemical actinometry, as reported previously.²⁹

Results and Discussion

It has been reported that pH plays an important role in the photo-oxidation of ammonia since in acidic conditions fewer OH anions are present, leading to a much lower generation rate of hydroxyl radicals in the reaction media³³ which are proposed to be active for this reaction,¹⁴ and the ideal reaction pH has been reported to be around 9.¹¹,¹⁴ Since initial screening experiments indicated that the as-synthesized Nb₂O₅ was weakly active (Figure 1), we chose to use the modified niobium support (OH-Nb₂O₅) to determine how a neutral or basic pH affects the ammonium photo-oxidation reaction. 20 mg of the modified niobium support (OH-Nb₂O₅) was added to 50 mL of O₂-saturated DI water containing 100 mg L^–1 of NH₄⁺ (Figure S1, Supplementary Information (SI) section). NH₄⁺ degradation did not occur without illumination. Almost no conversion of ammonia to NO₂ and NO₃^– was observed at pH 7, but a 15-fold increase in NH₄⁺/NH₃ conversion and a 5-fold increase in NO₂^– and NO₃^– formation was observed when the pH was raised to 9.

Figure 1
(a) Ammonium concentration-time profiles; (b) NO₂^– concentration-time profiles, and (c) NO₃^– concentration-time profiles for: no catalyst (▬), Nb₂O₅ (▬), and OH-Nb₂O₅ (▬). Reaction conditions: 4 g L^-1 catalyst loading, 600 rpm stirring rate, 1 atm pressure, pH = 9, DI water containing 4.2 mmol N L^–1 NH₄⁺ under O₂-containing headspace.

After confirming the ideal reaction pH to be 9 (Figure S1), we compared the OH-Nb₂O₅ photocatalyst with the as-synthesized Nb₂O₅ and a no-catalyst control experiment (Figure 1). Minimal absorption of NH₄⁺ was observed during the one-hour period of stirring in the dark. In the absence of a catalyst, no degradation of NH₄⁺ was detected, as expected. The as-synthesized Nb₂O₅ material was active, albeit weakly, with an apparent first-order rate constant of 1.6 × 10^–3 min^–1. In contrast, the OH-Nb₂O₅ was ca. 12 times more active, with a rate constant of 2.0 × 1 0 ^–2 min^–1. Undesirable nitrate and nitrite byproducts were detected with both the Nb₂O₅ and OH-Nb₂O₅ catalysts.

Figure 2 shows the selectivity of nitrite, nitrate, and calculated dinitrogen as a function of ammonium conversion for the two catalysts. In the < 1% NH₄⁺ conversion range observed, Nb₂O₅ had selectivities to nitrite and to nitrate in the range of 20 to 30%, and by nitrogen balance, a N₂ selectivity of ca. 50%. In the 2-10% conversion range observed, the OH-Nb₂O₅ photocatalyst had selectivities to nitrite and to nitrate around ca. 15% and a N₂ selectivity of ca. 70%. It is tempting to speculate that OH-Nb₂O₅ is more N₂-selective than Nb₂O₅, but a rigorous comparison of selectivity values cannot be made at different reaction conversions.

Figure 2
Selectivity vs. conversion for (a) Nb₂O₅ and (b) OH-Nb₂O₅. Reaction conditions: 4 g L^-1 catalyst loading, 600 rpm stirring rate, 1 atm pressure, pH = 9, DI water containing 4.2 mmol N L^–1 NH₄⁺ under O₂-containing headspace.

We were also interested in comparing the performance of these catalysts to the most used photocatalyst TiO₂-P25 under the same reaction conditions used here, as were also published for aqueous photo-oxidation of NH₄⁺. Vikrant et al.¹⁴ proposed the quantum yield (QY) and spacetime yield (STY) as a metric to compare photocatalysts that take into account process variables in a unified metric to minimize bias. Based on the QY and STY values for several published photocatalytic materials tested for aqueous photo-oxidation of NH₃/NH₄⁺, OH-Nb₂O₅ was more active than the most active material reported thus far (a Pt-modified TiO₂ catalyst, Table S1, SI section).¹⁴ Moreover, the OH-Nb₂O₅ photocatalyst showed 4 fold-times higher ammonia decay rate per photon flux (QY) and slightly higher STY than the TiO₂-P25 under the same reaction condition and reactor configuration, evidencing the better performance of the OH-Nb₂O₅ photocatalyst.

The XRD pattern of the OH-Nb₂O₅ and Nb₂O₅ structures is very similar to the deformed orthorhombic Nb₂O₅ structure reported by Murayama et al.,³⁴,³⁵ with amorphous arrangement in the a-b plane and crystalline in the c plane, characteristic of a layered structure. We expected the OH-Nb₂O₅ and Nb₂O₅ structures to be different, as the alkaline treatment of niobium oxides is known to cause partial dissolution or the formation of alkaline niobate structures.³⁶,³⁷ No niobate structures were detected via XRD, as shown by the absence of the corresponding peaks between 2θ = 20 to 90° (Figure 3). Two sharp peaks at 2θ = 23.0° and 2θ = 46.7° were present in both samples and attributed to the (001) and (002) planes, respectively, of deformed orthorhombic Nb₂O₅ structures. Additionally, the calculated Nb–O–Nb internal layer distance of 3.86 Å is similar to distances reported for Nb₂O₅ layered structures.³⁴,³⁵,³⁸ Notably, the intensity of the detected peaks is much smaller in the OH-Nb₂O₅ sample and may indicate much smaller crystalline domains, implying disruption of the bulk crystal structure.

Figure 3
X-ray diffraction (XRD) patterns for Nb₂O₅ (▬) and OH-Nb₂O₅ (▬).

SEM micrographs and N₂ physisorption analysis show that the as-synthesized support Nb₂O₅ had a fiber-like structure (Figure 4a) with a measured surface area of 175 m² g^–1, similar to previously reported semicrystalline Nb₂O₅ materials.³⁴,³⁵,³⁸ Upon treatment with base, the Nb₂O₅ structure lost the fiber-like morphology (Figure 4b), and the surface area decreased substantially (6 m² g^–1). With ca. 30 times less surface area, the OH-Nb₂O₅ material showed ca. 11× higher performance than the Nb₂O₅ (STY of 6.29 × 10^–3 vs. 5.98 × 10^–4, respectively) indicating that the surface area was not the most important contributor for the photo-oxidation performance of the Nb₂O₅ based materials. Moreover, the difference in the surface area is a strong sign of different surface sites formed after the basic treatment of the Nb₂O₅ material.

Figure 4
SEM micrographs of (a) Nb₂O₅ and (b) OH-Nb₂O₅.

Raman spectroscopy was used to further investigate the photocatalyst structures, as it is commonly used to identify specific polyhedral structures from niobium and oxygen through their different vibrational modes (Figure 5a).^{39 40 41} For Nb₂O₅, two bands at 210 and 716 cm^–1 were observed. The shoulder around 283 cm^–1 is from Nb–O stretches in NbO₆ octahedra (Figure 5b). The other two arise from vibrations of Nb–O–Nb bonds, where the former is Nb‑O‑Nb bending and the latter Nb–O–Nb stretching of corning sharing NbO₆ octahedra in crystalline Nb₂O₅ (Figure 5c). Moreover, the band at ca. 716 cm^–1 is also indicative of the presence of stretching modes of polyhedral niobium, primarily associated with the orthorhombic phase.⁴²,⁴³

Figure 5
(a) Ex situ Raman spectra for Nb₂O₅ and OH-Nb₂O₅; and (b-g) niobium octahedra with atoms and lines marked in magenta associated with the Raman vibrational modes.

The band at 716 cm^–1 was mostly absent in the spectrum for OH-Nb₂O₅ while a small band appeared at 648 cm^–1, which is associated with amorphous cornersharing octahedral NbO₆, implying the destruction of crystalline corner-sharing Nb–O species after treatment with NaOH.³⁶,^{39 40 41},^{44 45 46} The Raman spectrum for OH-Nb₂O₅ also showed a small but sharp band at 457 cm^–1, consistent with internal Nb–O stretches of amorphous edge-sharing octahedral NbO₆ (Figure 5d), and an additional band at 533 cm^–1, consistent with stretching modes of distorted NbO₆ octahedra (Figure 5e). Finally, the appearance of the bands located at 840 and 903 cm^–1 (consistent with the existence of terminal Nb=O bonds, Figures 5f and 5g),³⁶,⁴⁷ further corroborates the notion that the crystal structure of Nb₂O₅ was disrupted by the base treatment.

DR-UV spectroscopy was employed to investigate the potential effects of structural modifications on the semiconductor band gap. Bandgap (Eg) energies were determined utilizing Tauc analysis ((F(R) hν)^0.5 vs. hν), considering the indirect bandgap characteristics inherent to both Nb₂O₅ and OH-Nb₂O₅.²⁸ The calculated bandgap energies for Nb₂O₅ and OH-Nb₂O₅ were 3.8 and 4.0 eV, respectively. Remarkably, although the band gaps of the two materials are nearly identical, they both fall below the corresponding UV-C photon energies of 4.9 e V. This intriguing observation suggests that the augmented performance of the OH-Nb₂O₅ semiconductor is most likely attributed to surface modifications rather than variances in band gap energy.

We hypothesize that the photocatalytic activity of the modified niobium structure is due to the presence of the terminal Nb=O surface sites. At the reaction pH of 9.0 (lower than the pK_a of NH₄⁺), the majority (ca. 64%) of the NH₃ is expected to be protonated in solution at equilibrium. Both Nb₂O₅ materials (with a point of zero charge, pzc, around 4.3)^{48 46 47 48 49 50} would be expected to electrostatically adsorb NH₄⁺ species since the surface is negatively charged (due to the suspension pH exceeding the pzc). Even though it had a higher surface area, the negligible activity of Nb₂O₅ highlights the necessity of having Nb=O adsorption sites to promote the photocatalytic reaction (Scheme 1).

Scheme 1
Proposed initial steps of NH₄⁺ oxidation catalyzed by OH-Nb₂O₅.

Moreover, while higher surface area is usually associated with better adsorption performance, published studies for ammonium/ammonia adsorption using different natural clays reported that a clay with a surface area of 8.45 m² g^–1 can adsorb 3 times more NH₄⁺ than a clay with 208.93 m² g^–1 (ca. 30 fold-times difference), indicating that NH₃/NH₄⁺ adsorption and interaction with a material is not solely dependent on its surface area.⁵¹ Additionally, it has been reported that surface sites similar to the terminal Nb=O (e.g. O–C=O and C=O) enhance the adsorption of ammonium/ammonia.³¹,⁵²

We speculate that the ammonia photocatalytic oxidation mechanism for OH-Nb₂O₅ is similar to the mechanism reported for other materials, in which NH₃ (not NH₄⁺) is the species that reacts with hydroxyl radicals (generated through the reaction of holes with water). The resultant partially dehydrogenated NH_m and NH_n species (where n and m = 0, 1, or 2) further react with ^•OH or the photogenerated h⁺ at the photocatalyst surface for further dehydrogenation and formation of N₂ (equations 7,8,9,10). Ammonia may react with hydroxyl radicals to generate hydroxylamine (HONH₂), which can further react with ^•OH to form NO₂^– (equation 11). Further oxidation of NO₂^– to NO₃^– is proposed to happen in the presence of O₂ (equation 12).¹⁴,³¹,⁵³ Photogenerated electrons are expected to react with dissolved O₂ to form superoxide species (O₂^–) and hydroperoxyl radicals (HO₂^•). Perhaps the Nb=O surface sites can adsorb NH_m and NH_n species as surface intermediates, and can also facilitate the N–N coupling needed to form the desired N₂.

(7) ${\text{NH}}_3+{}^{·}\text{O}\text{H}\to {\text{NH}}_2+{\text{H}}_2\text{O}$

(8) ${\text{NH}}_2+{}^{·}\text{O}\text{H}/{\text{h}}^{+}\to \text{NH}+{\text{H}}_2\text{O}+{\text{H}}^{+}$

(9) ${\text{NH}}_{\text{m}}+{\text{NH}}_{\text{n}}\to {\text{N}}_2{\text{H}}_{\text{m}+\text{n}}\left(\text{m},\text{n}=0,1,2\right)$

(10) ${\text{N}}_2{\text{H}}_{\text{m}+\text{n}}+\left(\text{m}+\text{n}\right){\text{h}}^{+}\to {\text{N}}_2+\left(\text{m}+\text{n}\right){\text{H}}^{+}$

(11) ${\text{NH}}_3+{}^{·}\text{O}\text{H}\to {\text{HONH}}_2+{\text{H}}_{\text{ads}}+5{}^{·}\text{O}\text{H}\overrightarrow{\to }4{\text{H}}_2\text{O}+{\text{NO}}_2^{-}+{\text{H}}^{+}$

(12) $2{\text{NO}}_2^{-}+{\text{O}}_2\to 2{\text{NO}}_3^{-}$

Figure 6.
(a) DR-UV spectra, and (b) Tauc plots for Nb₂O₅ and OH-Nb₂O₅.

Conclusions

We demonstrate that Nb₂O₅, once treated in a basic solution, is active for the photocatalytic oxidation of ammonium under ambient conditions and UV-C irradiation. The modified Nb₂O₅ photocatalyst was more than ca. 7× active than the most active reported catalysts, which we correlate to the spectroscopically detected Nb=O surface species and not to surface area. These findings suggest that synthesizing Nb₂O₅ with these surface species and with higher surface areas may lead to more rapid photocatalytic degradation of ammonium and higher selectivity to N₂.

Supplementary Information

Supplementary information including results of the pH optimization experiments (Figure S1) 0103-5053-jbchs-35-04-e-20230163-suppl01 and a tabulation of prior photocatalytic studies (Table S1) is available free of http://jbcs.sbq.org.br as a PDF file..

Acknowledgments

The research was sponsored by the Army Research Office (Cooperative Agreement Number W911NF-19-2-0269) and CCDC-ARL (Cooperative Agreement Number W911NF-21-2-0118). The views and conclusions contained in this document are those of the authors and should not be interpreted as representing the official policies, either expressed or implied, of the ARL or the U.S. Government. The U.S. Government is authorized to reproduce and distribute reprints for Government purposes notwithstanding any copyright notation herein. We also acknowledge additional support from the NSF Nanosystems Engineering Research Center for Nanotechnology-Enabled Water Treatment (EEC-1449500). We thank Prof A. Martí and Ms Cecilia Martínez for conducting the DR-UV measurements.

References

• ¹
  Randall, D. J.; Tsui, T. K. N.; Mar. Pollut. Bull 2002, 45, 17. [Crossref]
  » Crossref
• ²
  Canadian Environmental Protection Act, 1999 and Bills and Related Documents, https://www.canada.ca/en/environment-climate-change/services/canadian-environmental-protection-act-registry/related-documents.html, accessed in October 2023.
  » https://www.canada.ca/en/environment-climate-change/services/canadian-environmental-protection-act-registry/related-documents.html,
• ³
  Bernardi, M.; Le Du, M.; Dodouche, I.; Descorme, C.; Deleris, S.; Blanchet, E.; Besson, M.; Appl. Catal., B 2012, 128, 64. [Crossref]
  » Crossref
• ⁴
  Sommer, S. G.; Christensen, M. L.; Schmidt, T.; Jensen, L. S.; Animal Manure Recycling: Treatment and Management; John Wiley & Sons, Ltd: Chichester, UK, 2013.
• ⁵
  Zhang, H.; Liu, H.; Zhao, X.; Qu, J.; Fan, M.; J. Hazard. Mater 2011, 190, 645. [Crossref]
  » Crossref
• ⁶
  van Dongen, U.; Jetten, M. S. M.; van Loosdrecht, M. C. M.; Water Sci. Technol 2001, 44, 153. [Crossref]
  » Crossref
• ⁷
  The Water Environment Federation (WEF); Design of Water Resource Recovery Facilities, 6^th ed.; McGraw-Hill Education: New York, 2018.
• ⁸
  Zhong, C.; Hu, W. B.; Cheng, Y. F.; J. Mater. Chem. A 2013, 1, 3216. [Crossref]
  » Crossref
• ⁹
  Chmielarz, L.; Jabłońska, M.; RSC Adv 2015, 5, 43408. [Crossref]
  » Crossref
• ¹⁰
  Zhu, X.; Castleberry, S. R.; Nanny, M. A.; Butler, E. C.; Environ. Sci. Technol 2005, 39, 3784. [Crossref]
  » Crossref
• ¹¹
  Stathatos, E.; Papoulis, D.; Aggelopoulos, C. A.; Panagiotaras, D.; Nikolopoulou, A.; J. Hazard. Mater 2012, 211-212, 68. [Crossref]
  » Crossref
• ¹²
  Shavisi, Y. ; Sharifnia, S.; Hosseini, S. N.; Khadivi, M. A.; J. Ind. Eng. Chem 2014, 20, 278. [Crossref]
  » Crossref
• ¹³
  Bonsen, E.-M.; Schroeter, S.; Jacobs, H.; Broekaert, J. A. C.; Chemosphere 1997, 35, 1431. [Crossref]
  » Crossref
• ¹⁴
  Vikrant, K.; Kim, K.-H.; Dong, F.; Giannakoudakis, D. A.; ACS Catal 2020, 10, 8683. [Crossref]
  » Crossref
• ¹⁵
  Hayati, F.; Moradi, S.; Saei, S. F.; Madani, Z.; Giannakis, S.; Isari, A. A.; Kakavandi, B.; J. Environ. Manage 2022, 321, 115851. [Crossref]
  » Crossref
• ¹⁶
  Hasanvandian, F.; Moradi, M.; Samani, S. A.; Kakavandi, B.; Setayesh, S. R.; Noorisepehr, M.; Chemosphere 2022, 287, 132273. [Crossref]
  » Crossref
• ¹⁷
  Mehregan, S.; Hayati, F.; Mehregan, M.; Isari, A. A.; Jonidi Jafari, A.; Giannakis, S.; Kakavandi, B.; Environ. Sci. Pollut. Res 2022, 29, 74951. [Crossref]
  » Crossref
• ¹⁸
  Naldoni, A.; D’Arienzo, M.; Altomare, M.; Marelli, M.; Scotti, R.; Morazzoni, F.; Selli, E.; Dal Santo, V. ; Appl. Catal., B 2013, 130-131, 239. [Crossref]
  » Crossref
• ¹⁹
  Sola, A. C.; Ramírez de la Piscina, P.; Homs, N.; Catal. Today 2020, 341, 13. [Crossref]
  » Crossref
• ²⁰
  Nobijari, L. A.; Schwarze, M.; Tasbihi, M.; J. Nanosci. Nanotechnol 2020, 20, 1056. [Crossref]
  » Crossref
• ²¹
  Luo, X.; Chen, C.; Yang, J.; Wang, J.; Yan, Q.; Shi, H.; Wang, C.; Int. J. Environ. Res. Public Health 2015, 12, 14626. [Crossref]
  » Crossref
• ²²
  Ou, H.-H.; Hoffmann, M. R.; Liao, C.-H.; Hong, J.-H.; Lo, S.-L.; Appl. Catal., B 2010, 99, 74. [Crossref]
  » Crossref
• ²³
  Nair, R. V; Gummaluri, V. S.; Matham, M. V.; C, V.; J. Phys. D: Appl. Phys 2022, 55, 313003. [Crossref]
  » Crossref
• ²⁴
  Nunes, B. N.; Lopes, O. F.; Patrocinio, A. O. T.; Bahnemann, D. W.; Catalysts 2020, 10, 126. [Crossref]
  » Crossref
• ²⁵
  Marszewski, M.; Cao, S.; Yu, J.; Jaroniec, M.; Mater. Horizons 2015, 2, 261. [Crossref]
  » Crossref
• ²⁶
  Gong, X.; Wang, H.; Yang, C.; Li, Q.; Chen, X.; Hu, J.; Future Cities Environ 2015, 1, 12. [Crossref]
  » Crossref
• ²⁷
  Lin, M.; An, B.; Niimi, N.; Jikihara, Y.; Nakayama, T.; Honma, T.; Takei, T.; Shishido, T.; Ishida, T.; Haruta, M.; Murayama, T.; ACS Catal 2019, 9, 1753. [Crossref]
  » Crossref
• ²⁸
  Duan, L.; Wang, B.; Heck, K. N.; Clark, C. A.; Wei, J.; Wang, M.; Metz, J.; Wu, G.; Tsai, A.-L.; Guo, S.; Arredondo, J.; Mohite, A. D.; Senftle, T. P.; Westerhoff, P.; Alvarez, P.; Wen, X.; Song, Y. ; Wong, M. S.; Chem. Eng. J 2022, 448, 137735. [Crossref]
  » Crossref
• ²⁹
  Duan, L.; Wang, B.; Heck, K.; Guo, S.; Clark, C. A.; Arredondo, J.; Wang, M.; Senftle, T. P.; Westerhoff, P.; Wen, X.; Song, Y.; Wong, M. S.; Environ. Sci. Technol. Lett 2020, 7, 613. [Crossref]
  » Crossref
• ³⁰
  Long, M.; Brame, J.; Qin, F.; Bao, J.; Li, Q.; Alvarez, P. J. J.; Environ. Sci. Technol 2017, 51, 514. [Crossref]
  » Crossref
• ³¹
  Liu, D.; Xiu, Z.; Liu, F.; Wu, G.; Adamson, D.; Newell, C.; Vikesland, P. ; Tsai, A.-L.; Alvarez, P. J.; J. Hazard. Mater 2013, 262, 456. [Crossref]
  » Crossref
• ³²
  Menezes, B. A. T.; Moreira, D. E. B.; Oliveira, H. A.; Marques, L. F.; Lima, J. F.; J. Braz. Chem. Soc 2020, 31, 153. [Crossref]
  » Crossref
• ³³
  Shavisi, Y. ; Sharifnia, S.; Zendehzaban, M.; Mirghavami, M. L.; Kakehazar, S.; J. Ind. Eng. Chem 2014, 20, 2806. [Crossref]
  » Crossref
• ³⁴
  Murayama, T.; Haruta, M.; Chin. J. Catal 2016, 37, 1694. [Crossref]
  » Crossref
• ³⁵
  Murayama, T.; Chen, J.; Hirata, J.; Matsumoto, K.; Ueda, W.; Catal. Sci. Technol 2014, 4, 4250. [Crossref]
  » Crossref
• ³⁶
  Zhu, H.; Zheng, Z.; Gao, X.; Huang, Y.; Yan, Z.; Zou, J.; Yin, H.; Zou, Q.; Kable, S. H.; Zhao, J.; Xi, Y.; Martens, W. N.; Frost, R. L.; J. Am. Chem. Soc 2006, 128, 2373. [Crossref]
  » Crossref
• ³⁷
  Nico, C.; Monteiro, T.; Graça, M. P. F.; Prog. Mater. Sci 2016, 80, 1. [Crossref]
  » Crossref
• ³⁸
  Lin, M.; An, B.; Takei, T.; Shishido, T.; Ishida, T.; Haruta, M.; Murayama, T.; J. Catal 2020, 389, 366. [Crossref]
  » Crossref
• ³⁹
  Lopes, O. F.; Paris, E. C.; Ribeiro, C.; Appl. Catal., B 2014, 144, 800. [Crossref]
  » Crossref
• ⁴⁰
  Santos, I. C. M. S.; Loureiro, L. H.; Silva, M. F. P.; Cavaleiro, A. M. V. ; Polyhedron 2002, 21, 2009. [Crossref]
  » Crossref
• ⁴¹
  Nowak, I.; Ziolek, M.; Chem. Rev 1999, 99, 3603. [Crossref]
  » Crossref
• ⁴²
  Braga, V. S.; Dias, J. A.; Dias, S. C. L.; de Macedo, J. L.; Chem. Mater 2005, 17, 690. [Crossref]
  » Crossref
• ⁴³
  Tiozzo, C.; Bisio, C.; Carniato, F.; Gallo, A.; Scott, S. L.; Psaro, R.; Guidotti, M.; Phys. Chem. Chem. Phys 2013, 15, 13354. [Crossref]
  » Crossref
• ⁴⁴
  Zhu, K.; Cao, Y.; Wang, X.; Bai, L.; Qiu, J.; Ji, H.; CrystEngComm 2012, 14, 411. [Crossref]
  » Crossref
• ⁴⁵
  Brayner, R.; Bozon-Verduraz, F.; Phys. Chem. Chem. Phys 2003, 5, 1457. [Crossref]
  » Crossref
• ⁴⁶
  Jehng, J. M.; Wachs, I. E.; Chem. Mater 1991, 3, 100. [Crossref]
  » Crossref
• ⁴⁷
  Iliev, M. N.; Phillips, M. L. F.; Meen, J. K.; Nenoff, T. M.; J. Phys. Chem. B 2003, 107, 14261. [ Crossref]
  » Crossref
• ⁴⁸
  dos Santos, A. J.; Batista, L. M. B.; Martínez-Huitle, C. A.; Alves, A. P. M.; Garcia-Segura, S.; Catalysts 2019, 9, 1070. [Crossref]
  » Crossref
• ⁴⁹
  Shan, Y.; Zheng, Z.; Liu, J.; Yang, Y. ; Li, Z.; Huang, Z.; Jiang, D.; npj Comput. Mater 2017, 3, 11. [Crossref]
  » Crossref
• ⁵⁰
  Kosmulski, M.; Langmuir 1997, 13, 6315. [Crossref]
  » Crossref
• ⁵¹
  Alshameri, A.; He, H.; Zhu, J.; Xi, Y.; Zhu, R.; Ma, L.; Tao, Q.; Appl. Clay Sci 2018, 159, 83. [Crossref]
  » Crossref
• ⁵²
  Wang, S.; Ai, S.; Nzediegwu, C.; Kwak, J.-H.; Islam, M. S.; Li, Y.; Chang, S. X.; Bioresour. Technol 2020, 309, 123390. [Crossref]
  » Crossref
• ⁵³
  Lee, J.; Park, H.; Choi, W.; Environ. Sci. Technol 2002, 36, 5462. [Crossref]
  » Crossref

Edited by

Editor handled this article: Célia M. Ronconi (Associate)

Publication Dates

• Publication in this collection
  01 Mar 2024
• Date of issue
  2024

History

• Received
  06 July 2023
• Accepted
  10 Oct 2023