Introduction

Nitrite ions (NO2) are among the major contaminating pollutants in the environment and wastewater [1]. They are commonly used in the food industry, fertilizer, and corrosion inhibition [2, 3]. The high level of NO2 ions causes many serious health problems (e.g. blue baby syndrome, gastrointestinal tumors, and stomach cancer) [4]. At a high concentration of NO2 ions in the bloodstream, they can react with hemoglobin leading to the formation of methemoglobin in the body [5]. They can also form secondary amine when reaching the acidic medium in the stomach and thus induce the formation of nitrosamine compounds which are carcinogens [5]. On another hand, the high concentration of ammonia causes eutrophication and the depletion of oxygen, and thus it is harmful to aquatic life [1, 6, 7]. Ammonia might attack the breathing system, skin, and eyes, and at a concentration higher than 300 mg L−1 it might lead to death [8]. In an aqueous solution, ammonia exists in equilibrium with the ammonium ions (NH4+). Owing to the hazardous nature of NO2 ions and ammonia the World Health Organization (WHO) has defined guidance levels for their concentration in drinking water to protect humans from health risks (i.e. 3.0 and 0.2 mg L−1 for short-term and long-term exposure to NO2 ions, respectively, and 1.24 mg L−1 for exposure to ammonia) [9]. According to the standards for drinking water quality in China, the upper limit concentration for NH4+ based on N is 0.5 mg L−1 [10]. The environmental protection agency (EPA) in the united states has regulated 1.0 mg L−1 as the maximum acceptable contaminated level of NO2 ions in drinking water [11], whereas the European Drinking Water Directive established this number at 0.02 mg L−1 [12]. According to these regulations, the concentrations of both nitrite and ammonia must be controlled and kept within the permitted levels.

Great efforts have been devoted to mitigating the risks associated with the dissolved inorganic nitrogen (DIN) in the forms of NO2 and ammonia by biological removal either through nitrification or denitrification processes [13, 14]. Nevertheless, this biological method has some drawbacks, e.g. complicated, sludge production, time-consuming, and in which dissolved oxygen is controlled painstakingly [15, 16]. Therefore, alternative economical methods were introduced such as air stripping [17], ion exchange [18], and chemical precipitation [19]. Recently, the use of photo-induced processes (i.e. photolysis and photocatalysis) for the degradation of DIN has attracted great attention [13, 20, 21]. In these methods, the photo-induced degradation involves three routes: (1) the oxidation of ammonia to nitrite and/or nitrate [22], (2) the reduction of nitrite to nitrogen gas as a mimicking process for nitrification or denitrification process [12, 23], and (3) the simultaneous removal of ammonia and nitrite ions [13].

In the photolysis process, when an aqueous solution of nitrite ions is irradiated with UV light (200–400 nm), a chain of reactions occurs as summarized in Scheme 1 [24] and reactive nitrogen species (RNS) and OH· radicals are generated. These radical species are widely used in the field of advanced oxidation processes mediated by UV irradiation (UV-AOPs) [25]. Generally, the OH· radicals induce the hydroxylation reactions with organic pollutants, whereas the RNS specially NO2· initiate two types of photoreaction with organic contaminate (i.e. nitration or nitrosation). Thus, using nitrite ions in the photolysis tests leads to the formation of nitrated or nitrosated products that are typically more cytotoxic, carcinogenic, and genotoxic than carbonaceous sources.

Scheme 1
scheme 1

Primary photo-processes and subsequent reactions during NO2 photolysis. Reprinted from Ref. [24] with permission from Elsevier

Photocatalysis is thus particularly suitable for removing nitrite ions under irradiation from liquid samples [26, 27]. Nitrite ions can be either photocatalytically oxidized to nitrate or reduced to nitrogen gas. TiO2 nanoparticles, metal-loaded TiO2, and synergetic heterojunction with TiO2 are widely used in the field of hazard pollutants degradation which nitrite considers one of them. Kominami et al. examined the photocatalytic disproportionation of nitrite ions in aqueous suspensions of bare and metal-loaded TiO2 particles without electron and hole scavengers under irradiation of UV light using a 400 W high-pressure mercury arc lamp [28]. The results of this study showed that in the case of bare TiO2, 16% of nitrite ions are converted to NO3, N2, and N2O after 3.0 h of irradiation, and 70% of nitrite ions are photo-transformed after 24 h but when Pd–TiO2 was used, nitrite was completely converted to NO3, N2, and N2O after photoirradiation for 3.0 h. The largest reaction rate was obtained in the reaction without pH control, indicating that an almost neutral condition was appropriate for photocatalytic disproportionation of NO2. Zafra et al. have investigated the heterogeneous photo-oxidation of nitrite to nitrate over TiO2 P25 using a 250 W xenon lamp at different experimental conditions [29]. Results of this study indicated that as the pH of the TiO2 suspension increased, the photo-oxidation of nitrite to nitrate was decreased. The study of the photocatalytic reduction of nitrite and nitrate over M-TiO2 using a 450 W xenon arc lamp (where M is Pt, Pd, Ru, Rh) has also been reported [30]. RE-doped titania prepared by the sol–gel method (where RE are La3+, Ce3+, Er3+, Pr3+, Gd3+, Nd3+, Sm3+) was also tested for the photocatalytic oxidation of nitrite ion using a 300-W high-pressure mercury lamp [31]. The photocatalytic degradation of nitrite to ammonia over solid-state photocatalyst films (e.g. methyl functionalized silicate sol–gel/TiO2, methyl functionalized silicate sol–gel/(TiO2–Au), methyl functionalized silicate sol–gel/(TiO2–Au)/Ni(II), nafion/TiO2, nafion /(TiO2–Au), and nafion/(TiO2–Au)nps/Ni(II)) was also tested under irradiation using 450 W xenon lamp [32].

The Ag–TiO2 photocatalyst showed 100% selectivity toward the reduction of nitrite to nitrogen in the presence of oxalic acid as a hole scavenger at pH 2.34–3.41using 125-W high-pressure Hg lamp as a light source [23]. Three-dimensional GO/TiO2 hydrogels were used to remove nitrite from water and seawater using LED light sources (430–460 nm) in the presence of formic acid as a hole scavenger [33]. TiO2–GO composite prepared by the sol–gel method showed also high selectivity towards the photocatalytic reduction of nitrite to nitrogen gas using a low-pressure mercury lamp in the presence of formic acid as a hole scavenger. A complete reduction of nitrite to nitrogen in the presence of TiO2–GO was achieved after exposure to UV light for 120 min [34].

Surprisingly, however, oxalic acid and formic acid are commonly used as electron donors in many photolytic and photocatalytic removal of nitrite ions at low pH values (e.g. pH 3 [35], 2.34–3.41 [23], 1.70 [36]), the possible disproportionation of nitrite ions at low pH is less considered. Also, less attention is given to the photolysis and photocatalytic removal of nitrite ions and ammonia under UV-A light. The light sources that have been used in most of the studies are either low, medium, or high-pressure mercury lamps which emit photons in the UV-B region and generate a lot of heat. Thus, in this work, we have carefully used 365 nm LED to avoid the heat and UV-B and UV-C effects. Then, we have revised the influence of pH on the photolytic and photocatalytic oxidation of nitrite ions under aerobic atmosphere. The simultaneous removal of nitrite ions and ammonia has been studied under inert atmosphere. The plausible mechanisms for the photolytic and photocatalytic oxidation of nitrite ions as well as for the simultaneous removal of nitrite ions and ammonia have been discussed.

Experimental section

Photolytic oxidation of nitrite ions

The photolytic experiment of nitrite ions removal was conducted in a batch photoreactor in which 100 mL of nitrite aqueous solution containing 108.5 µmol of NO2 was added (1.085 mmol L−1). Then, the pH was adjusted to the desired value using NaOH or H2SO4. The photoreactor was irradiated from the top using a 365 nm light emitting diode (LED) at a light intensity of 20 mW cm−2. Oxygen was purged during the experiment using an aquarium air pump. At an interval time of 1.0-h, a 3.0 mL solution was sampled for analysis. The concentration of nitrite ions was determined spectrophotometrically using α-naphthyleamine and sulfanilic acid as color developers (Supporting information, Fig. S1) [37].

Photocatalytic oxidation of nitrite ions

The photocatalytic removal of nitrite ions was carried out in the same photoreactor. Typically, 50 mg of TiO2 P25 photocatalyst was dispersed in 100 mL of nitrite aqueous solution (108.5 µmol NO2) and the pH of the TiO2 suspension was adjusted to the desired value using NaOH or H2SO4. Before irradiation, the suspension was stirred in the dark under O2 purge for half an hour to ensure that the whole system reached the adsorption–desorption equilibrium. After that, the photoreactor was illuminated with a 365 nm LED with a light intensity of 20 mW cm−2 at the suspension’s surface. At a time interval of 0.5 h, a sample of 3.0 mL was withdrawn by a syringe and the TiO2 particles were removed using a 0.46 μm PVDF syringe filter. The concentration of nitrite ions was determined spectrophotometrically (Supporting information, Fig. S1).

Photolytic and photocatalytic removal of nitrite ions in presence of ammonia

The simultaneous photolytic removal of nitrite and ammonia was performed under Ar inert atmosphere. Typically, 5.0 mL of nitrite ions and ammonia solutions (containing 50 µmol nitrite ions and 100 µmol ammonia) were inserted into a photoreactor at the desired pH value. The reactor was then sealed with a rubber septum and purged with Ar for 15 min. After that, the photoreactor was illuminated with a 365 nm LED lamp with a light intensity of 20 mW cm−2. At an interval time of 1.0-h, the head-space gases were collected with a gas-tight syringe and the concentration of the evolved N2 gas was quantized by gas chromatography (Agilent 7890A GC SYSTEM with TCD detector). After 6 h of the UV-A irradiation, the remaining amount of nitrite ions was determined using the spectrophotometric method described in the supporting information. For the determination of ammonia, the Nessler reagent was used as a color developer and the absorbance was measured at λ = 375 nm (Supporting information, Fig. S2) [13].

The removal efficiencies of nitrogen of nitrite ions (NN), ammonia (NA), and total nitrogen (TN) are given by Eqs. 13, respectively, whereas the selectivity of N2 evolution (SN2) and the nitrogen balance (NB) were calculated by Eqs. 4 and 5, respectively.

$$\mathrm{NN\; removal}\; (\mathrm{\%}) =\frac{( {C}_{0, N }-{C}_{t, N })}{{C}_{0, N }} \times 100$$
(1)
$$\mathrm{NA\; removal}\; (\mathrm{\%})=\frac{( {C}_{0, A }-{C}_{t, A })}{{C}_{0, A }} \times 100$$
(2)
$$\mathrm{TN\; removal }\; (\mathrm{\%}) =\frac{\left( {C}_{0, \mathrm{N} }-{C}_{t, \mathrm{N} }\right)+( {C}_{0, \mathrm{A} }-{C}_{t, \mathrm{A} })}{{C}_{0, \mathrm{A} }+ {C}_{0, \mathrm{N} }} \times 100$$
(3)
$${\mathrm{SN}}_{2}\; (\mathrm{\%})=\frac{{2C}_{{\mathrm{N}}_{2}}}{\left( {C}_{0, \mathrm{N} }-{C}_{t, \mathrm{N} }\right)+( {C}_{0, \mathrm{A} }-{C}_{t, \mathrm{A} })} \times 100$$
(4)
$$\mathrm{NB }(\mathrm{\%})=\frac{{{C}_{t, \mathrm{N} }+{C}_{t, \mathrm{A} }+2C}_{{\mathrm{N}}_{2}}}{{C}_{0, \mathrm{N} }+{C}_{0, \mathrm{A} }} \times 100,$$
(5)

where C0,N and Ct,N are the initial concentration of nitrite ions and their concentrations at time t, respectively, whereas C0,A and Ct,A are the initial concentration of ammonia and its concentration at time t, respectively. CN2 represents the amount of the evolved N2 gas. All concentrations are expressed in µmol.

The simultaneous photocatalytic test for the removal of nitrite and ammonia was carried out using a similar procedure for the photolytic removal of nitrite ions in presence of ammonia except that 5.0 mg of the TiO2 P25 photocatalyst was added before sealing and purging the photoreactor with Ar. And the resulted suspension was stirred in the dark for half an hour to ensure that the whole system has reached the adsorption–desorption equilibrium.

Results and discussion

Photolytic oxidation of nitrite ions

Nitrite ion exhibit absorption in the UV-A region due to its n → π* transition at λmax = 354 nm as shown in Fig. 1. The molar absorption coefficient (ε) was calculated by applying the Beer–Lambert’s law and it is equal to 23.3 M−1 cm−1 in agreement with the reported values [24, 38].

Fig. 1
figure 1

UV–Vis absorption spectra of nitrite ions in aqueous solution at different concentrations. The inset shows the Beer–Lambert’s plot at 354 nm

However, nitrite ions absorb UV-A portion of the solar spectrum, UV-C radiation is commonly employed for their photolysis. In principle, the interaction of nitrite ions with UV light (200–400 nm) induces the formation of reactive nitrogen species (RNSs) and OH· radicals, and ultimately nitrate ions as shown in Scheme 1. The pH value of the aqueous solution is highly expected to influence the rate of photolytic degradation of nitrite ions as they exist in the HONO form at low pH value. The latter form has a different photo-reactivity than NO2 ions [39]. Figure 2a shows the time course of nitrite ions removal at different pH values (i.e. from 1.0 to 11.0). It is obvious from Fig. 2a that there is no significant photolytic degradation of nitrite ions in basic, neutral, and slightly acidified mediums. The photolytic decomposition significantly starts at pH 3.0. By decreasing the pH value of nitrite aqueous solution to one, the augmentation depletion of nitrite ions was observed and reaches 100% after 3 h. By measuring the dark stability of nitrite ions in aqueous solutions at pH 1.0 and 3.0 (Fig. 2b), we found that the rate of nitrite disproportionation is high. This fact is less considered in literature when oxalic and formic acid is used as sacrificial reagents [23, 35, 36].

Fig. 2
figure 2

Photolytic decomposition of NO2 ions at different pH values under UV-ALED365nm irradiation (a), and in the dark (b). For ease of comparison, the curves at pH 1.0 and 3.0 measured under light are added to part (b). Light intensity (20 mW cm−2, 365 nm); initial concentration of NO2 (1.085 mmol L−1)

By investigating the products after one hour of irradiation, an absorption band centered at a wavelength of 300–305 nm is observed (Fig. 3a). The intensity of this absorption band is increased with the prolongation of the photoirradiation time. This absorption band might belong to either peroxynitrite (ONO2) [40] or nitrate (NO3) [25]. Its stability for a long period when kept in the dark (Fig. 3b) suggests that it more likely belongs to the nitrate ions, rather than to the peroxynitrite. The latter is not stable and should be decomposed into nitrate ions in a short time. The same peak has previously been accepted as an indication of the formation of nitrate ions [24, 25].

Fig. 3
figure 3

a UV–Vis absorption spectra of the remaining nitrite ions in presence of the color developer (scans: 800–400 nm) and of the formed products in the absence of the color developer (scans: 400–200 nm) at different time courses for the photolytic oxidation test; b stability of the photolytic decomposition products in the dark after the photolytic test. Conditions: concentration of nitrite (1.085 mmol L−1), pH 1.0, UV-A 365 nm irradiation, light intensity (20 mW cm−2)

By analyzing the products of the dark decomposition of nitrite ions at pH 1.0, similar absorption bands to that presented in Fig. 3a have been observed (Fig. 4). This confirms that nitrite ions are not stable at low pH even in the dark and thus using low pH to evaluate the photolytic and/or photocatalytic oxidation of nitrite ions is not reasonable. The photolytic decomposition of nitrite ions can be summarized as follows [24, 40]:

Fig. 4
figure 4

UV–Vis absorption spectra of the remaining nitrite ions in presence of the color developer (scans: 800–400 nm) and of the formed products in the absence of the color developer (scans: 400–200 nm) at different time courses for the dark decomposition at pH 1.0

$${\mathrm{NO}}_{2}^{-} + h\nu \to {[{\mathrm{NO}}_{2}^{-}]}^{*}$$
(6)
$${[{\mathrm{NO}}_{2}^{-}]}^{*} \to {\mathrm{NO}}^{\cdot } + {\mathrm{O}}^{\cdot -}$$
(7)
$${\mathrm{O}}^{\cdot -} + {\mathrm{H}}_{2}\mathrm{O }\leftrightharpoons {}^{\cdot }\mathrm{OH }+ {\mathrm{OH}}^{-}$$
(8)
$${K}_{1} = 1.7\times 10^{6} {\mathrm{M}}^{-1} {\mathrm{s}}^{-1},{\mathrm{ K}}_{-1} = 1.2\times {10}^{10} {\mathrm{M}}^{-1} {\mathrm{s}}^{-1}$$
$${[{\mathrm{NO}}_{2}^{-}]}^{*} \to {\mathrm{NO}}_{2}^{\cdot } + {\mathrm{e}}_{\mathrm{aq}}^{-}$$
(9)
$${\mathrm{NO }}_{2}^{-}+ {\mathrm{H}}^{+} \leftrightharpoons \mathrm{ HONO}$$
(10)
$$\mathrm{p}K\mathrm{a }=3.39$$
$$\mathrm{HONO }+\mathrm{ h}\nu \to {}^{\cdot }\mathrm{OH }+ {\mathrm{NO}}^{\cdot }$$
(11)
$$2{\mathrm{NO}}^{\cdot } + {\mathrm{O}}_{2} \to 2{\mathrm{NO}}_{2}^{\cdot }$$
(12)
$${}^{\cdot }\mathrm{OH }+ {\mathrm{NO}}_{2}^{-} \to {\mathrm{NO}}_{2}^{\cdot } + {\mathrm{OH}}^{-}$$
(13)
$${K}_{2} = 1.0\times {10}^{10} {\mathrm{M}}^{-1} {\mathrm{s}}^{-1}$$
$${\mathrm{e}}_{\mathrm{aq}}^{-} + {\mathrm{O}}_{2} \to {\mathrm{ O}}_{2}^{\cdot -}$$
(14)
$${K}_{3}= 2\times {10}^{10} {\mathrm{M}}^{-1} {\mathrm{s}}^{-1}$$
$${\mathrm{NO}}^{\cdot } + {\mathrm{O}}_{2}^{\cdot -} \to {\mathrm{O}}_{2}{\mathrm{NO}}^{-}$$
(15)
$${K}_{4}= 3.7\times {10}^{7} {\mathrm{M}}^{-1}{\mathrm{s}}^{-1}$$
$${\mathrm{NO}}_{2}^{\cdot } +{}^{\cdot }\mathrm{OH }\to \mathrm{ O}{\mathrm{NO}}_{2}\mathrm{H}$$
(16)
$${K}_{5} =1.3\times {10}^{9} {\mathrm{M}}^{-1} {\mathrm{s}}^{-1}$$
$${\mathrm{O}}_{2}{\mathrm{NO}}^{-} + {\mathrm{H}}^{+} \leftrightharpoons \mathrm{ O}{\mathrm{NO}}_{2}\mathrm{H}$$
(17)
$$(\mathrm{Acid}-\mathrm{ base\; equilibrium }) (\mathrm{p}K\mathrm{a}=6.8)$$
$${\mathrm{ONO}}_{2}\mathrm{H }\to {\mathrm{NO}}_{3}^{-} + {\mathrm{H}}^{+}$$
(18)
$${K}_{6} =2\times {10}^{3} {\mathrm{M}}^{-1} {\mathrm{s}}^{-1}$$

The photolysis of NO2 under irradiation with UV-ALED365nm might result in the formation of NO·, O·−, and NO2· radicals that react with water to form ·OH radicals [24]. These radical species can also contribute to forming peroxynitrite O2NO. The peroxynitrite is protonated to peroxynitous acid O2NOH according to the acid–base equilibrium (pKa = 6.8 equilibrium reaction (Eq. 17), then O2NOH isomerizes to nitrate according to Eq. 18 [40]. The pH value of nitrite aqueous solution after 6 h of irradiation decreased from 1.0 to 0.88. This decrease in the pH value agrees with Eq. 18. In principal the nitrite ions can be completely decomposed at lower pH than 3.0 even in absence of light. However, adjusting the pH to be below 3 seems unpractical and thus the oxidation of nitrite ions should be studied at the near-neutral pH. In the coming section, we have investigated the effect of pH on the photocatalytic oxidation of nitrite ions over different TiO2 photocatalysts, namely, TiO2 P25, UV100, and home-made anatase/brookite mixture.

Photocatalytic oxidation of nitrite ions

Figure S3 shows the X-ray diffraction (XRD) of TiO2 P25, UV100, and the home-made anatase/brookite mixture prepared according to Ref. [41]. The obtained XRD patterns indicated that TiO2 P25 consists of a mixture of anatase and rutile, whereas TiO2 UV100 consists of pure anatase. The existence of the brookite phase in the anatase/brookite sample is discernible from the (121) diffraction located at 2θ = 30.81° where no overlap with the anatase and rutile diffraction peaks exists. The surface areas of the investigated TiO2 photocatalysts have been calculated from the N2 adsorption isotherms and presented in Figure S4. The N2 adsorption/desorption curves of TiO2 P25, TiO2 UV100, TiO2 A/B belong to the type IV isotherm, which resembles an H3 hysteresis loop associated with monolayer formation and capillary condensation behavior for mesoporous materials. The BET surface areas of TiO2 P25, TiO2 UV100, and TiO2 A/B were calculated to be 50.4, 236, and 106.79 m2/g, respectively.

Figure 5 shows the time courses of the photocatalytic oxidation of nitrite ions over TiO2 P25, TiO2 UV100, and TiO2 A/B under UV-A irradiation at near-neutral pH where the photolytic decomposition of nitrite ions is not expected. TiO2 P25 exhibited the highest activity and thus it has been used to illustrate the effect of pH on the rate of the photocatalytic oxidation of nitrite ions.

Fig. 5
figure 5

Photocatalytic oxidation of NO2 ions over different TiO2 photocatalysts under UV-ALED365nm irradiation at pH 5. Light intensity (20 mW cm−2, 365 nm); initial concentration of NO2 (1.085 mmol L−1)

Figure 6 shows the time courses of the photocatalytic oxidation of nitrite ions over TiO2 P25 at different pH values under UV-ALED365nm irradiation. The oxidation of nitrite ions occurs at all pH values (from 5.0 to 11) where the decomposition of nitrite ions is not significant as proven in Sect. 3.1. A 100% removal was possible at pH 5 and 7 after 3.5 and 4 h of UV-A irradiation, respectively. But at a pH greater than 7, the degradation becomes more difficult. A considerable explanation of this behavior can be made by looking at the formal charge of the TiO2 surface at different pH values. TiO2 P25 exhibits a point of zero charge (PZC) in the pH range from 6.0 to 7.0 [42], and thus if the pH of the suspension is below the pHPZC, most of the surface hydroxyl groups will be protonated. Accordingly, the formal charge of the TiO2 surface will be positive and thus increases the adsorption of the negatively charged NO2 ions [43] and hence increases the rate of nitrite oxidation. By increasing the pH to pH = 7, the degradation rate was decreased due to the aggregation and agglomeration processes at the PZC, which would result in decreased surface area due to flocculation and would slow the surface reaction rate [44]. After that by increasing the pH of TiO2 suspension above the pHPZC, the bridging hydroxyl ions will be deprotonated and the formal charge of the TiO2 will be negative and thus repulses the negatively charged NO2 [11]. For this reason, the photocatalytic oxidation rate will decline.

Fig. 6
figure 6

Photocatalytic oxidation of NO2 over TiO2 P25 under UV-ALED365nm irradiation at different pH values. Light intensity (20 mW cm−2, 365 nm); initial concentration of NO2 (1.085 mmol L−1)

It is believed that the photocatalytic oxidation of nitrite occurs on the surface of TiO2 and that O2 and H2O are necessary for photocatalytic oxidation. So, the potential possible steps of the photocatalytic degradation of nitrite ions over TiO2 nanoparticles in the presence of oxygen were listed in Eqs. (1924) [40, 45, 46].

$$ {\mathrm{TiO}}_{2} + h\nu \to {\mathrm{e }}_{\mathrm{CB}}^{-} + {h }_{\mathrm{VB}}^{+}$$
(19)
$${\mathrm{e }}_{\mathrm{CB}}^{-} + {\mathrm{ O}}_{2} \to {\mathrm{O}}_{2}^{\cdot -}$$
(20)
$$h^+_\mathrm{ VB }+ {\mathrm{H}}_{2}\mathrm{O }\to {}^{\cdot}\mathrm{OH}+ H^+$$
(21)
$${\mathrm{NO}}_{2}^{-} + 2{\mathrm{OH}}^{\cdot } \to {\mathrm{NO}}_{3}^{-} + {\mathrm{H}}_{2}\mathrm{O}$$
(22)
$${\mathrm{NO}}_{2}^{-} + {\mathrm{OH}}^{\cdot } \to {}^{\cdot } {\mathrm{HNO}}_{3}^{-}\leftrightharpoons {({}^{\cdot }{\mathrm{NO}}_{3}) }^{2-} + {\mathrm{H}}^{+}$$
(23)
$${({}^{\cdot }{\mathrm{NO}}_{3})}^{2-} + {\mathrm{O}}_{2} \to {\mathrm{NO}}_{3}^{-} + {\mathrm{O}}_{2}^{\cdot -}$$
(24)

When light with photon energy greater than or equal to the bandgap of TiO2 (n-type semiconductor) interacts with its surface, an electronic transition occurs and the valance band (VB) electron is excited into the conduction band (CB) to form a free-electron/hole pairs (Eq. 19). Oxygen molecules adsorbed onto the TiO2 surface trap the conduction band electrons; and thus, superoxide ions (O2·−) are formed (Eq. 20). On the other side, the photogenerated holes react with H2O or OH to form ·OH radicals (Eq. 21). These hydroxyl radicals are widely accepted as a primary oxidant in heterogeneous photocatalysis. ·OH radicals are strong enough to oxidize nitrite to nitrate (Eqs. 2224). But, the molecular oxygen plays also vital roles. It acts as an electron acceptor to reduce the electron/hole recombination and to facilitate the formation of NO3 ions (Eq. 24). As seen from the results presented in Sects. 3.1 and 3.2, nitrite ions can be converted to nitrate ions in photolytic and photocatalytic processes and the latter is superior in the near-neutral medium. In the coming section, we will present the possible pathways for the simultaneous removal of nitrite ions and ammonia as nitrogen gas under inert atmosphere.

Simultaneous removal of nitrite and ammonium ions

Nitrite ions in a concentrated aqueous solution might react with ammonium ions to form dinitrogen (N2) according to Eq. 25 [28]:

$${\mathrm{NO}}_{2}^{-} + {\mathrm{NH}}_{4}^{+} \to {\mathrm{N}}_{2} + 2{\mathrm{H}}_{2}\mathrm{O}$$
(25)

The oxidation states of nitrogen in NH4+, NO2 and N2 are − 3, + 3, and 0, respectively. Therefore, this reaction includes a change in the oxidation states of nitrogen in NH4+ and NO2 (+ 3 and − 3, respectively) to the same oxidation state (zero). In our study, the occurrence of such reaction was not observed in the absence of UV-A light and no study has shown simultaneous photolysis of NH3/NH4+ and NO2 by UV-ALED365nm before. Thus, the effect of pH on the simultaneous removal of nitrite and ammonium ions under UV-ALED365nm irradiation was investigated here in the pH range from 3 to 11 (Fig. 7). The results are summarized in Table 1. It is found that in an acidic medium (pH 3.0), the major degradation belongs to nitrite ions (32.1% removal) while a small amount of ammonia was removed (i.e. 4.9%). The reason may be attributed to the decomposition of nitrite ions as observed in Fig. 2b. In contrast, the amount of NH4+ ions (17.3 µmol) which undergo photolytic oxidation into nitrogen gas was higher than that of nitrite ions (i.e. 9.46 µmol) at pH 11. This might be attributed to the presence of ammonia as NH3 instead of NH4+ at this pH value (pH 11) which supports the reported finding that NH3 is more reactive than NH4+ [20]. The photolytic reaction between nitrite and ammonia does not occur in near-neutral and neutral mediums (pHs 5 and 7). Interestingly, at pH 9.0, the results showed that 42.69 ± 0.66% of nitrogen of nitrite ions (NN) can be removed after 6 h of irradiation of a solution containing 50 µmol nitrite ions and 100 µmol ammonia under inert conditions. The percentages of nitrogen of ammonia (NA) and total nitrogen (TN) removal were 27.75 ± 1.7% and 32.74 ± 0.59%, respectively. The selectivity of N2 gas evolution (SN2) was found to be 77.6% (19.07 ± 1.6 µmol). A control test displayed that the concentration of nitrite almost does not change in the absence of ammonia and vice versa at this pH (i.e. pH 9).

Fig. 7
figure 7

Time courses of N2 evolution during the photolytic/photocatalytic simultaneous removal of nitrite and ammonium at different pH values from solution/suspension containing 50 µmol nitrite ions and 100 µmol ammonia. Photocatalyst loading (5 mg), and light intensity (20 mW cm−2)

Table 1 Effect of pH on the percentage of nitrogen removal, N2 evolution selectivity, and nitrogen balance during the photolytic and photocatalytic simultaneous removal of nitrite ions and ammonia

The proposed mechanism for the photolytic conversion of nitrite and ammonia might involve the formation of NH2NO intermediate which decomposed to N2 and H2O. The primary reaction of ammonia photolysis at λ > 180 nm is accompanied by the formation of amidogen (NH2·, Eq. 26). The photolysis of nitrite ions produces NO· and OH· radicals as discussed in Sect. 3.1 (Eqs. 68). The OH· radicals can then oxidize ammonia according to Eq. 27 leading to the formation of NH2· radicals. The formed NH2· radicals react with NO· to form NH2NO intermediate, which is dissociated to form nitrogen and water as illustrated in Eqs. 2629) [24, 47, 48]. The selectivity toward the N2 gas formation will be thus determined by the formation of NH2NO intermediate which is sensitive to the pH values.

$${\mathrm{NH}}_{3}+ h\nu \to {\mathrm{NH}}_{2}^{\cdot } + {\mathrm{H}}^{\cdot }$$
(26)
$${\mathrm{NH}}_{3} + {\mathrm{OH}}^{\cdot } \to {\mathrm{NH}}_{2}^{\cdot } + {\mathrm{H}}_{2}\mathrm{O}$$
(27)
$${\mathrm{NH}}_{2}^{\cdot } + {\mathrm{NO}}^{\cdot } \to {\mathrm{ NH}}_{2}\mathrm{NO}$$
(28)
$${\mathrm{NH}}_{2}\mathrm{NO}\to {\mathrm{N}}_{2} + {\mathrm{H}}_{2}\mathrm{O}$$
(29)

For comparison between the photolytic and photocatalytic simultaneous removal of nitrite and ammonium ions at the optimal pH (i.e. pH 9), a photocatalyst test has been executed at this pH and the results are presented in Fig. 7. It was observed that the rate of nitrogen evolution is inhibited in comparison with that of the photolytic reaction. The photocatalytic N2 evolution from nitrite/ammonia solution requires the reduction of nitrite via one-electron transfer to NO· [49] and the oxidation of ammonia over TiO2 photocatalyst to form the NH2NO intermediate as schematically illustrated in Scheme 2. Thus, the reason for this inhibition may be attributed to the fact that at pH 9 the formal charge of the TiO2 surface is negative and thus repulses the NO2 ions and reduce the rate of their photocatalytic reduction leading to decreasing the overall rate of N2 evolution. Also, the TiO2 particles might prevent the photolytic reaction of ammonia and the formation of NH2· by shielding the light thus reducing the formation of the NH2NO intermediate required for the N2 evolution.

Scheme 2
scheme 2

Schematic illustration for the simultaneous photocatalytic removal of nitrite ions and ammonia

Conclusions

The photolytic and photocatalytic oxidation/removal of nitrite ions from aqueous solutions under UV-A illumination using 365 nm LED were studied at different pH values in the absence/presence of ammonia. It is found that nitrite ions are not stable at pH less than 3.0. At such low pH, the rate of the photolytic oxidation of nitrite ions is close to that of nitrite ions disproportionation in the dark, particularly, at pH 1.0. At higher pH values, the nitrite ions are stable and the photolytic rate is insignificant. The photocatalytic degradation tests over different TiO2 photocatalysts, namely, TiO2 P25, TiO2 UV100, and TiO2 anatase/brookite mixture, indicated that TiO2 P25 exhibits the highest activity and pH 5 is optimal. This pH is below the zero point of charge of TiO2 P25 and thus the adsorption of nitrite ions on the positively charged surface of TiO2 P25 facilitates the photocatalytic reaction. The photocatalytic oxidation of nitrite ions was more favorable relative to the photolytic oxidation at near-neutral mediums. 100% conversion of nitrite ions to nitrate was achieved at pH 5 and 7 after 3.5 and 4 h of UV-A irradiation, respectively. However, in the presence of ammonia and the absence of molecular oxygen (Ar atmosphere), the rate of photolytic removal of nitrite ions and ammonia, in particular at pH 9.0, is significantly higher than that of the photocatalytic one. The presence of ammonia was essential. No photo-induced reactions (N2 evolution) have been observed in the absence of ammonia. It is proposed that the photogenerated NO· and NH2· radicals recombine to form the NH2NO intermediate which is subsequently decomposed into N2. The percentages of photolytic removal of nitrogen based on nitrite and ammonia are 42.69 ± 0.66 and 27.75 ± 1.7, respectively, and the selectivity of N2 evolution is 77.6%. These values were reduced in the presence of TiO2 P25 photocatalyst. This is attributed to either the repulsion between the negatively charged TiO2 particles and nitrite ions or to the light-shielding effect.