Formation of reactive nitrogen species promoted by iron ions through the photochemistry of a neonicotinoid insecticide

Nitrous acid (HONO) and nitrogen oxides (${\mathrm{NO}}_x=\mathrm{NO}+{\mathrm{NO}}_{\mathrm{2}}$) are important atmospheric pollutants and key intermediates in the global nitrogen cycle, but their sources and formation mechanisms are still poorly understood. Here, we investigated the effect of soluble iron (Fe³⁺) on the photochemical behavior of a widely used neonicotinoid (NN) insecticide, nitenpyram (NPM), in the aqueous phase. The yields of HONO and NO_x increased significantly when NPM solution was irradiated in the presence of iron ions (Fe³⁺). We propose that the enhanced HONO and NO₂ emissions from the photodegradation of NPM in the presence of iron ions result from the redox cycle between Fe³⁺ and Fe²⁺ and the generated reactive oxygen species (ROS) by electron transfer between the excited triplet state of NPM and molecular oxygen (O₂). Using the laboratory-derived parameterization based on kinetic data and gridded downward solar radiation, we estimate that the photochemistry of NPM induced by Fe³⁺ releases 0.50 and 0.77 Tg N yr⁻¹ of NO_x and HONO, respectively, into the atmosphere.

This study suggests a novel source of HONO and NO_x during daytime and potentially helps to narrow the gap between field observations and model outcomes of HONO in the atmosphere. The suggested photochemistry of NPM can be an important contribution to the global nitrogen cycle affecting the atmospheric oxidizing capacity and climate change.

1 Introduction

Neonicotinoids (NNs) are a class of systemic insecticides that have been widely used in agriculture and horticulture since the 1990s (Bass et al., 2015), accounting for one-third of the total world insecticide market (Simon-Delso et al., 2015) and experiencing growing use in recent decades (Botías et al., 2015; Morrissey et al., 2015). They are highly water-soluble and persistent in the environment and can be transported to surface waters via runoff, leaching, or spray drift. NNs have been detected in various aquatic ecosystems, such as rivers, lakes, wetlands, and coastal waters, at concentrations ranging from 12.45 ng L⁻¹ to 225 µg L⁻¹ (Pan et al., 2020; Anderson et al., 2013). Increasing public perception of NN insecticide pollution has led to significant research efforts devoted to revealing the effect of insecticide application on humans (Cimino et al., 2017; Han et al., 2017), birds (Hallmann et al., 2014; Millot et al., 2017), animals (Morrissey et al., 2015; Gibbons et al., 2015), and pollinators (especially bees) (Kessler et al., 2015; Raine and Gill, 2015; Goulson et al., 2015). In the environment, NN insecticides can undergo various chemical processes, photolysis being one of their major fates (Lu et al., 2015; González-Mariño et al., 2018). Recent studies have focused mainly on the photochemistry of NN insecticides and their related atmospheric lifetimes and quantum yields (Lu et al., 2015; González-Mariño et al., 2018; Aregahegn et al., 2017, 2018). It has been shown that the ozonolysis of NN insecticides on various surfaces could contribute to the formation of gaseous nitrous acid (HONO) (W. Wang et al., 2020). Gaseous nitrous oxide (N₂O), which is a potent greenhouse gas, was previously identified as the gas-phase product in the photolysis of solid thin films of NNs (nitenpyram, acetamiprid, thiamethoxam, thiacloprid, clothianidin, and dinotefuran), with yields of $\mathrm{\Delta }{\mathrm{N}}_{\mathrm{2}}\mathrm{O}/\mathrm{\Delta }\mathrm{NN}>\mathrm{0.5}$ in air at both 313 and 254 nm (Wang et al., 2019; Aregahegn et al., 2017, 2018). Palma et al. (2020) used a gas-flow reactor connected to a NO_x analyzer, and the production of gaseous $\mathrm{NO}/{\mathrm{NO}}_{\mathrm{2}}$ began during irradiation (300–450 nm) of imidacloprid. However, the crucial role of NN insecticides in the global nitrogen cycle at the air–water interface is largely unknown.

Nitenpyram (NPM) is one of the most commonly used NN insecticides. It represents a systemic NN insecticide which is widely distributed among soils, dust particles, and the aqueous environment (Botías et al., 2015; Ezell et al., 2019). Once released into the environment, NPM will be transformed into other products by absorbing sunlight (λ>290 nm) and/or reacting with atmospheric oxidants such as the hydroxyl radical (OH) and ozone (O₃) (W. Wang et al., 2020). NPM is a nitroalkene, which is structurally similar to nitroaromatic compounds (Ar-NO₂). Previous studies have indicated that photolysis of Ar-NO₂ can be a source of HONO and NO_x in the atmosphere (Fukuhara et al., 2001; Yang et al., 2021; Bejan et al., 2006). HONO represents one of the main sources of OH radicals in the urban atmosphere, contributing up to 80 % of the total OH production (Alicke et al., 2003; Young et al., 2012; Zheng et al., 2020). The main identified HONO sources in urban air are photolysis of nitrates (Ye et al., 2017; Gen et al., 2021) and light-induced heterogeneous reaction of NO₂ with environmental surfaces (Liu et al., 2019, 2020, 2023; Monge et al., 2010; Han et al., 2016). However, there is a discrepancy between the modeled HONO values and field observations of HONO during the daytime, suggesting that there are missing HONO sources in the atmosphere. Meanwhile, the quantification of NO_x is also of great significance for the atmospheric cycle of nitrogen species, as NO_x plays a crucial role in photochemical smog and acid rain formation. Therefore, it is worthwhile exploring the contribution of NPM photolysis to HONO and NO_x, which in turn can offer guidance for the development of more sustainable next-generation insecticide products.

Iron species are ubiquitous on Earth surfaces, including water, soil, and the air–water interface (Gen et al., 2021). A recent study (Kebede et al., 2016) showed that one of the less explored HONO sources could be highly dependent on the photochemical reaction of iron. The photosensitivity, oxidation state, and catalytic properties of iron could enable it to possibly react with NN insecticide compounds that are enriched at the air–water interface. Previous studies on the mechanism of NN oxidation in the ferric aqueous phase have focused on the photo-Fenton reaction (Malato et al., 2001; Lacson et al., 2018; Wang et al., 2022; Nguyen et al., 2020; Sedaghat et al., 2016) and heterogeneous-phase photocatalysis (Rózsa et al., 2019; Sun and Liu, 2019; Hayat et al., 2019; Soltani-nezhad et al., 2019). As reported recently, the photolysis of iron can generate several reactive oxygen species (ROS), e.g., ${\mathrm{O}}_{\mathrm{2}}\mathrm{•}-/{\mathrm{HO}}_{\mathrm{2}}\mathrm{•}$, which can trigger the redox cycle between Fe³⁺ and Fe²⁺ (Gen et al., 2021) and promote NN insecticide oxidation. Meanwhile, iron ions inhibit the degradation of organic matter through the formation of complexes, which is mainly due to fluorescence bursting. This complexation may cause inhibition of the excited singlet state and thus photoformation of the triplet excited state (Wan et al., 2019). In addition to the NN insecticides and iron photosensitizers, nitrate (${{\mathrm{NO}}_{\mathrm{3}}}^{-}$) and nitrite (${{\mathrm{NO}}_{\mathrm{2}}}^{-}$) can absorb sunlight in the actinic region and initiate production of ROS (Vione and Scozzaro, 2019). Moreover, reaction between Fe²⁺ and ${{\mathrm{NO}}_{\mathrm{3}}}^{-}$ may be a potentially important source of HONO (Gen et al., 2021). To this end, we suggest that photolysis of NPM in the presence of iron may contribute to a missing atmospheric HONO source.

To the best of our knowledge, this is the first investigation to measure the photochemical production of HONO and NO_x from NPM photolysis in the absence and presence of soluble iron. The photolysis frequency of HONO (J_NPM→HONO), NO₂ ($J_{\text{NPM}\to {\mathrm{NO}}_{\mathrm{2}}}$), and NO (J_NPM→NO) during the NPM reaction at the air–water interface was investigated. The kinetics and mechanism of HONO and NO_x formation in the presence of soluble iron were evaluated. This study highlights an overlooked source of HONO and NO_x from NN-covered water surfaces that may play a critical role in the atmospheric nitrogen cycle and the evaluation of the atmospheric oxidation capacity.

2 Experiments

2.1 Material and sample preparation

Solid NPM (Aladdin, China) was dissolved in ultra-pure water to prepare an aqueous NPM solution (0.5 mg mL⁻¹) before each experiment. FeCl₃ (98 %; Aladdin, China) was used as the source of different concentrations of aqueous Fe³⁺ (0.1–0.8 mg mL⁻¹), and their solutions were prepared by dissolving the corresponding mass of FeCl₃ in ultra-pure water.

2.2 Experimental setup

The circular reactor consisted of a double layer of quartz glass (3.4 cm height, 7.5 cm inner diameter) connected to a thermostatic bath (XOSC-20, China), which allowed operation at a constant temperature of 298 K (Fig. S1 in the Supplement). The previously prepared sample solution was placed in the circular reactor and exposed to a xenon lamp (Perfect Light, PLS-SXE 300, China) vertically above the reactor. The xenon lamp was 12 cm away from the liquid level of NPM. The spectral irradiance of the xenon lamp was measured using a calibrated spectroradiometer (HP 350 UVP, China) (Fig. S1). Dry air collected from an air generator (HY-3, China) was used for the experiment. During the whole experiment, a constant flow of 800 mL min⁻¹ of dry air was controlled using an electronic soap film flowmeter (SCal Plus, China). The UV absorption spectra of the NPM aqueous solutions in the absence or presence of iron ions were measured using the UV–Vis double-beam spectrophotometer (Shimadzu 2600, Japan) (Fig. S2 and Test S1).

2.3 NO_x, HONO, NPM, and ROS measurements

NO, NO₂, and HONO were detected using a chemiluminescence NO_x analyzer (42i, THERMO) with a molybdenum converter. NO was measured by reacting NO with O₃ to produce characteristic luminescence, and the intensity of the luminescence was proportional to the concentration of NO. In the detection of NO₂, a molybdenum catalyst was used to convert NO₂ to NO. A quartz tube (25 cm length, 2.9 cm inner diameter) filled with a certain amount of crystalline (Na₂CO₃) was introduced to capture HONO between the circular reactor outlet and the NO_x analyzer. It is well known that almost all HONO molecules can contact Na₂CO₃ when using molybdenum converters, achieving a high capture efficiency of HONO. Therefore, HONO can be quantified indirectly by the difference between the NO₂ signal and the Na₂CO₃ tube (Monge et al., 2010; Cazoir et al., 2014; Brigante et al., 2008; Zhou et al., 2018). The quantification of NPM before and after the reaction was determined using high-performance liquid chromatography (HPLC). The mobile phase was a mixture of water and acetonitrile with a flow rate of 0.5 mL min⁻¹ at 80:20 ($v/v$). The column temperature was kept at 30 °C, the injection volume was 20 µL, and the detection wavelength was set to 270 nm. The external standard method was used for the quantitative determination of NPM. Photoproductions of O₂-•, ¹O₂, and OH were quantified using 5,5-Dimethyl-1-Pyrroline-N-oxide (DMPO), 2,2,6,6-Tetramethylpiperidine (TEMP) and 5,5-Dimethyl-1-Pyrroline-N-oxide (DMPO), respectively, as the chemical probe molecules.

2.4 Kinetic analysis

The NPM photolysis kinetics were described using a first-order reaction (Eq. 1), and the half-life ($t_{\mathrm{1}/\mathrm{2}}$) was calculated using Eq. (2):

$$\begin{array}{}\text{(1)}& {\displaystyle }{C}_t=C_{\mathrm{0}}\times e^{-kt},\text{(2)}& {\displaystyle }{t}_{\mathrm{1}/\mathrm{2}}=\ln \left(\mathrm{2}\right)/k,\end{array}$$

where C₀ (mg mL⁻¹) is the initial concentration of NPM, C_t (mg mL⁻¹) is the NPM concentration at time t, and k is the first-order rate constant.

2.5 The photolysis frequency

The photolysis frequencies of NPM to HONO and NO_x were calculated using Eqs. (3) and (4), respectively:

$$\begin{array}{}\text{(3)}& {\displaystyle }{J}_{\text{NPM}\to \mathrm{HONO}}={\displaystyle \frac{Q{M}_{\text{NPM}}{\int }_{\mathrm{0}}^t{C}_t^{\mathrm{HONO}}\mathrm{d}t}{\mathrm{60}\times {\mathrm{10}}^{-\mathrm{3}}{N}_{\mathrm{A}}\times t\times (m_{\mathrm{0}}+m_t)/\mathrm{2}}},\text{(4)}& {\displaystyle }{J}_{\text{NPM}\to {\mathrm{NO}}_x}={\displaystyle \frac{Q{M}_{\text{NPM}}{\int }_{\mathrm{0}}^t{C}_t^{{\mathrm{NO}}_x}\mathrm{d}t}{\mathrm{60}\times {\mathrm{10}}^{-\mathrm{3}}{N}_{\mathrm{A}}\times t\times (m_{\mathrm{0}}+m_t)/\mathrm{2}}},\end{array}$$

where Q (mL min⁻¹) and M_NPM (g mol⁻¹) are the total gas flow rates in the reactor and the molar mass of NPM, respectively; t (min) is the irradiation time; $C_t^{{\mathrm{NO}}_x}$ ($\mathrm{molec}.{\mathrm{cm}}^{-\mathrm{3}}$) is the concentration of gaseous HONO or NO_x formed by photolysis of NPM during the irradiation period; N_A is the Avogadro number; and M₀ (mg) and M_t (mg) are the masses at the beginning and end of the NPM photolysis experiments.

2.6 Flux densities of HONO and NO_x

The flux densities of HONO and NO_x were estimated using the following equations:

$$\begin{array}{}\text{(5)}& {\displaystyle }{\mathrm{HONO}}_{\text{flux}}={\displaystyle \frac{\left[\mathrm{HONO}\right]\cdot V}{s\cdot t}},\text{(6)}& {\displaystyle }{{\mathrm{NO}}_x}_{\text{flux}}={\displaystyle \frac{\left[{\mathrm{NO}}_x\right]\cdot V}{s\cdot t}},\end{array}$$

where HONO flux is expressed ($\mathrm{molec}.{\mathrm{cm}}^{-\mathrm{2}}{\mathrm{s}}^{-\mathrm{1}}$), [HONO] is the concentration of HONO ($\mathrm{molec}.{\mathrm{cm}}^{-\mathrm{3}}$), V (cm³) is the volume of the reactor, S (cm²) is the surface of the reactor, and t (s) is the residence time of HONO in the circular reactor.

2.7 Global simulation of NO_x and HONO fluxes

We estimated the global inventory of the NO_x and HONO fluxes produced by NPM photochemistry using the observation-constrained parameterization scheme and hourly solar radiation data. Gridded and hourly downward solar radiation data are obtained from Modern-Era Retrospective analysis for Research and Application Version 2 (MERRA-2)-assimilated meteorological fields. We calculated the flux of NO_x and HONO for each model grid at a horizontal resolution of 0.5°×0.625° (consistent with the MERRA-2 radiation dataset) following Eqs. (S1)–(S3) in the Supplement but assuming that the environmental NPM concentration is 3 orders of magnitude smaller than the experimental condition of 0.5 mg L⁻¹. The parameterization of HONO and NO_x productions from NPM photolysis at the Fe³⁺ concentration of 0.025 mg L⁻¹ used in our estimation is based on Eqs. (S1)–(S3), and more details can be seen in Test S2.

3 Results and discussion

3.1 Absorbance of NPM in the presence of Fe³⁺

Figure S1 shows the absorbance of NPM (0.05 mg mL⁻¹) in the dilute aqueous phase and at different Fe³⁺ concentrations, adjusted by FeCl₃, along with the emission spectrum of the solar simulator and the sunlight. The presence of Fe³⁺ at various initial concentrations slightly enhanced the absorbance of NPM, especially at a high Fe³⁺ concentration (0.08 mg mL⁻¹), indicating that no Fe³⁺–NPM complexes were generated (Liu et al., 2022). Indeed, pH is a sensitive parameter that can significantly affect the light-absorbing properties and degree of photochemical degradation of organic compounds (Cai et al., 2018; Zhou et al., 2019). The interaction between Fe³⁺ and organics as well as possible aggregation of organics at low pH may also influence the light absorption at low wavelengths (Weishaar et al., 2003). The change in Fe³⁺ concentrations may alter the pH of the system, which in turn may affect the protonation or deprotonation degree of NPM and therefore its absorption spectrum (Zhou et al., 2019). The pH value of the NPM solution in the presence of Fe³⁺ varies between 2.4 and 3.4, and under this pH condition NPM (pKa=3.1) exists in both ionic and neutral forms (Hậu et al., 2021; Bonmatin et al., 2014).

3.2 Kinetic analysis

Iron ions are ubiquitous in natural waters, with concentrations ranging from 10⁻⁷ to 10⁻⁴ M and even higher in contaminated waters (Li et al., 2018; Faust and Hoigné, 1990). Previous studies have shown that iron ions play an important role in the photolysis of pesticides and may affect the photodegradation of organic pollutants (Faust and Hoigné, 1990; Zhao et al., 2014). The photolysis kinetics of NPM were performed to account for the loss of NPM. The photolysis of NPM at different concentrations of Fe³⁺ obeyed pseudo-first-order kinetics (Fig. 1), with half-lives ranging from 135.1 to 223.6 min as the Fe³⁺ concentration increased from 0 to 0.8 mg mL⁻¹ (Table S1 in the Supplement).

Figure 1 (a) The kinetics of NPM (0.5 mg mL⁻¹) in the absence of Fe³⁺ (dark line) and in the presence of different Fe³⁺ concentrations: 0.1 mg mL⁻¹ (red line), 0.25 mg mL⁻¹ (blue line), 0.5 mg mL⁻¹ (green line), and 0.8 mg mL⁻¹ (purple line). (b) The rate constants of NPM light-induced degradation (0.5 mg mL⁻¹) at different Fe³⁺ concentrations.

The light-induced degradation of NPM was significantly inhibited at low Fe³⁺ concentrations ($C_{\left({\mathrm{Fe}}^{\mathrm{3}+}\right)}<\mathrm{0.5}$ mg mL⁻¹; Fig. 1 and Table S1). In contrast, when the concentration of Fe³⁺ reaches 0.8 mg mL⁻¹, the degradation of NPM is promoted (Fig. 1), exhibiting a rate constant of 0.00513 min⁻¹ (Table S1). Previous studies have demonstrated that the degradation of organic compounds in the presence of Fe³⁺ is dose-dependent (Lin et al., 2019; Deguillaume et al., 2005). For instance, Fe³⁺ slightly inhibits the photodegradation of fluazaindolizine at concentrations of 1–5 mg L⁻¹ but promotes its degradation rate at concentrations ranging between 0.1 and 0.5 mg L⁻¹ (Lin et al., 2019). Deguillaume et al. (2005) reported that photodegradation of flupyradifurone, a novel neonicotinoid pesticide, was faster at lower Fe³⁺ concentrations and slowed down with the increase in Fe³⁺ concentrations.

The main reason for the inhibition effect of Fe³⁺ is the attenuation of radiation due to the absorption by Fe³⁺ (light screening), which reduces the light absorbance by NPM and its photodegradation. At the same time, it has been extensively confirmed that $\left[{\mathrm{Fe}}^{\mathrm{3}+}\right(\mathrm{OH}){]}^{\mathrm{2}+}$ is the main form of Fe³⁺ and exhibits great photoactivity in aqueous solution at pH=3 (Bai et al., 2023; Li et al., 2023). In the presence of $\left[{\mathrm{Fe}}^{\mathrm{3}+}\right(\mathrm{OH}){]}^{\mathrm{2}+}$, strong oxidizing ROS are produced, which promote hydroxylation and degradation of NPM (Andrianirinaharivelo et al., 1995; Mazellier et al., 1997). As a result, at pH=3, the photodegradation of NPM is accelerated at high Fe³⁺ concentrations.

In this study, a high Fe³⁺ concentration (0.8 mg mL⁻¹) promoted the photodegradation of NPM and the formation of HONO and NO_x (see the section below). The enhanced formation of HONO and NO_x can be ascribed to ROS, as described in Sect. 3.5.

3.3 HONO and NO_x formations by NPM photolysis

The experiments of NPM photodegradation in the aqueous phase were performed to measure the HONO and NO_x production in the presence of different Fe³⁺ concentrations. The HONO and NO_x production by spontaneous reaction of NPM in the dark was negligible (Fig. S3 in the Supplement). When the NPM samples were exposed to light irradiation, the concentrations of HONO and NO_x quickly increased (Fig. 2a).

Figure 2 (a) Concentrations of NO, NO₂, and HONO produced by NPM photolysis at different concentrations of Fe³⁺. (b) J_NPM→HONO and $J_{\text{NPM}\to {\mathrm{NO}}_x}$ from NPM at different concentrations of Fe³⁺ and an irradiation intensity of 169.4 W m⁻² at $\mathrm{300}<\mathit{\lambda }<\mathrm{400}$ nm, T=298 K.

Only the concentration of NO formed upon irradiation of NPM is almost the same in the absence of Fe³⁺ and in the presence of 0.25 mg mL⁻¹ of Fe³⁺ (Fig. 2a). In the meantime, the NO₂ formation increased significantly with the increase in Fe³⁺ concentrations and remained almost steady during the whole light exposure time (Fig. S3). Moreover, when the experiments were shifted to high concentrations of soluble iron (0.25–0.8 mg mL⁻¹), significantly enhanced NO₂ and NO formation was observed that then slowly decreased with the light exposure time. In order to better understand the effect of iron on HONO and NO_x production, the quantities of HONO and NO_x were compared when the NPM photolysis reached a relatively stable stage (120 min). It is important to note that the formed HONO (341 ppb) was significantly higher at an iron concentration of 0.8 mg mL⁻¹ compared to the HONO (37 ppb) that formed in the absence of iron. Similarly, the quantity of the formed NO₂ increased from 17 ppb in the absence of iron to 96 ppb in the presence of 0.5 mg mL⁻¹ of Fe³⁺. However, further increases in the iron concentration to 0.8 mg mL⁻¹ tended to decrease the production of NO₂. Figure 2 shows that the NO concentrations remained almost unchanged with the increase in the iron concentration. To quantify the photolysis quantum yields of HONO, NO₂, and NO formation from NPM photolysis, we estimated the photolysis frequency of HONO (J_NPM→HONO), NO₂ ($J_{\text{NPM}\to {\mathrm{NO}}_{\mathrm{2}}}$), and NO (J_NPM→NO) formation, respectively (Fig. 2b). J_NPM→HONO varied from (2.99 ± 0.46) $\times {\mathrm{10}}^{-\mathrm{7}}$ s⁻¹ in the absence of Fe³⁺ to (2.79 ± 0.10) $\times {\mathrm{10}}^{-\mathrm{6}}$ s⁻¹ in the presence of 0.8 mg mL⁻¹ Fe³⁺. Simultaneously, $J_{\text{NPM}\to {\mathrm{NO}}_{\mathrm{2}}}$ increased ca. 5-fold from (1.25 ± 0.06) $\times {\mathrm{10}}^{-\mathrm{7}}$ s⁻¹ in the absence of Fe³⁺ to (6.77 ± 0.44) $\times {\mathrm{10}}^{-\mathrm{7}}$ s⁻¹ at 0.8 mg mL⁻¹ of Fe³⁺. Regarding J_NPM→NO, there were nearly no discernible changes observed, with values ranging from (2.38 ± 0.27) $\times {\mathrm{10}}^{-\mathrm{7}}$ to (2.92 ± 0.15) $\times {\mathrm{10}}^{-\mathrm{7}}$ s⁻¹. A previous study (Yang et al., 2021) found that the photolysis frequency of HONO and NO in nitrophenol solid-phase films (4-nitrophenol, 4-nitrocatechol, 3,5-dinitrosalicylic acid, 3-nitrosalicylic acid, and 5-nitrosalicylic acid) varied in the ranges (0.34–4.16) $\times {\mathrm{10}}^{-\mathrm{7}}$ and (0.38–3.21) $\times {\mathrm{10}}^{-\mathrm{7}}$ s⁻¹, respectively, when irradiated by xenon lamps. NPM liquid-phase photolysis produced HONO and NO_x at a photolysis frequency of 10⁻⁷, but the addition of iron resulted in a photolysis frequency of 10⁻⁶ for HONO, suggesting that iron significantly facilitated the release of HONO. In order to compare the efficiency of NPM at different Fe³⁺ concentrations in producing HONO and NO, Φ_HONO and ${\mathrm{\Phi }}_{{\mathrm{NO}}_x}$ were displayed (Table S2 in the Supplement). It can be concluded that NPM with high Fe³⁺ concentrations had more important HONO formations as compared to pure NPM.

Figure 3 Flux densities of HONO (a), NO₂ (b), and NO (c) determined as a function of the photolysis time of NPM in the presence of different concentrations of Fe³⁺.

3.4 HONO and NO_x surface flux densities

Figure 3 summarizes the results obtained in terms of HONO formation rates per unit of exposed surface area, flux densities of HONO, NO₂, and NO.

The flux density values of HONO and NO_x indicate that direct photolysis dominated the transformation process of the NPM samples in the absence of Fe³⁺. However, the introduction of soluble iron leads to significantly increased HONO and NO₂ yields during the first 10 min of the reaction time. Further progress of the reaction up to 2 h leads to slightly increased flux densities of NO₂ and HONO. In contrast, the NO formation showed a slow decrease after the addition of Fe³⁺. A recent study (Aregahegn et al., 2017) demonstrated that photolysis of a solid film consisting of imidacloprid (IMD) did not lead to HONO and NO_x formation and that N₂O was rather the main gas-phase product. However, it is important to note that the introduction of Fe³⁺ promotes the photodegradation of NPM to produce more HONO and NO_x. In the section below we suggest a tentative reaction mechanism to describe the formation of HONO and NO₂ upon irradiation of NPM at the water surface in the presence of soluble iron.

3.5 Mechanism describing the formation of HONO and NO_x

We speculate that, in the presence of Fe³⁺, the decrease in dissolved nitrogen species that resulted from the photodegradation of NPM is the reason for the formation of HONO and NO_x. Therefore, ROS and dissolved nitrogen-containing ions were measured upon photodegradation of NPM in the presence of Fe³⁺. The generation of superoxide radicals (O₂-•), singlet oxygen (¹O₂), and hydroxyl radicals (OH) was quantified using DMPO, TEMP, and DMPO, respectively, as the chemical probe molecules. Figure 4a shows that, in the absence of Fe³⁺, the photodegradation of NPM induces generation of OH, O₂-•, and ¹O₂, which can be ascribed to the electron transfer between the excited triplet state of NPM and the molecular oxygen (O₂) (Segura et al., 2008; Mostafa and Rosario-Ortiz, 2013; Marin et al., 2012; Wang et al., 2021).

Figure 4 (a) EPR spectra obtained upon photodegradation of NPM (0.5 mg mL⁻¹) in the absence of Fe³⁺ as a function of the reaction time. (b) EPR spectra obtained upon 45 min photodegradation of NPM (0.5 mg mL⁻¹) in the absence of Fe³⁺ (dark line) and in the presence of 0.25 mg mL⁻¹ of Fe³⁺ (red line) and 0.8 mg mL⁻¹ of Fe³⁺ (blue line).

It has been reported that, under UV light irradiation, Fe³⁺ photoreduction regenerates Fe²⁺, accelerating the process due to the formation of new OH radicals (Segura et al., 2008). The electron paramagnetic resonance (EPR) measurements revealed an interesting phenomenon where the increase in the Fe³⁺ concentration promotes the consumption rate of ROS (Fig. 4b) rather than the production rate. The generated ROS would react with lower-valence nitrogen-containing species to form HONO and NO_x. Based on this finding, we suggest a tentative reaction mechanism which could explain the formation of large quantities of HONO and NO_x during the photochemical degradation of NPM. The photochemical generation of ROS could be driven by two pathways. Pathway I is the excited triplet state of NPM (³NPM^∗) that can be formed under light irradiation (Reaction R1) (Mora Garcia et al., 2021), and then, by reacting with water molecules (Reaction R2), it can trigger the formation of ROS such as OH radicals, accompanied by the generation of O₂-• through the transformation between radical anions of NPM (NPM-•) and dissolved oxygen (Reaction R3) (Wang et al., 2021). Furthermore, with the progress of the photodegradation of NPM, an increase in O₂-• and OH formation was observed (Fig. 4a), favoring HONO and NO₂ formation (Reactions R6–R8). In the presence of Fe³⁺, formation of OH radicals by Reaction (R4) occurs as well (Mazellier et al., 1997). In addition, nitrate ions (${{\mathrm{NO}}_{\mathrm{3}}}^{-}$) and nitrite ions (${{\mathrm{NO}}_{\mathrm{2}}}^{-}$) in the aqueous phase are formed by Reactions (R5)–(R7). Peroxynitrate (${{\mathrm{OONO}}_{\mathrm{2}}}^{-}$) is formed by reaction of O₂-• with NO₂, which thermally decomposes to form ${{\mathrm{NO}}_{\mathrm{2}}}^{-}$ and O₂, which further leads to HONO formation (Reaction R6) (X. Wang et al., 2020; Lammel et al., 2002; Goldstein et al., 1998). The reaction between O₂-• and NO can lead to the formation of ${{\mathrm{NO}}_{\mathrm{2}}}^{-}$ and ${{\mathrm{NO}}_{\mathrm{3}}}^{-}$, with a relatively fast rate constant of 4.3×10⁹ ${\mathrm{M}}^{-\mathrm{1}}{\mathrm{s}}^{-\mathrm{1}}$ (Goldstein and Czapski, 1995) producing a peroxynitrite (OONO⁻) which then yields ${{\mathrm{NO}}_{\mathrm{3}}}^{-}$ through internal rearrangement (Reaction R7) (Loegager and Sehested, 1993). At neutral pH (pKa=6.5), the OONOH product can also be formed by protonation, which can coexist with OONO⁻ to form ${{\mathrm{NO}}_{\mathrm{2}}}^{-}$ (Reaction R7) (Guptaet et al., 2009). Previous studies have shown that the reaction between OH and ${{\mathrm{NO}}_{\mathrm{2}}}^{-}$ will generate NO₂ (Reaction R8) (Loegager and Sehested,, 1993), and a sharp increase in the HONO concentration occurs immediately from reaction between ${{\mathrm{NO}}_{\mathrm{2}}}^{-}$ and H⁺ (Reaction R9), which is expected to be an important pathway of HONO formation.

Figure 5 Global emissions of HONO (a) and NO_x (b), produced by photochemistry of NPM in the presence of iron ions for the year 2017.

At low Fe³⁺ concentrations (0.25–0.5 mg mL⁻¹), the degradation rate of NPM was completely inhibited, which was not the case for higher Fe³⁺ concentrations (0.5–0.8 mg mL⁻¹) (Fig. 1). Notably, Fe³⁺ plays an important role in providing an acidic environment (pH=2.4–3.4) in the reaction system, which is followed by the redox reaction between Fe²⁺ and ${{\mathrm{NO}}_{\mathrm{3}}}^{-}$ to produce NO₂ and consequently increase the amount of NO₂ (Reaction R10) (Fig. S3). It has been shown that ${{\mathrm{NO}}_{\mathrm{3}}}^{-}$ undergoes a photochemical process and thus produces HONO (Reaction R11) and NO₂ (Reaction R12) (Ye et al., 2016; Zhou et al., 2011).

$$\begin{array}{}\text{(R1)}& {\displaystyle }\mathrm{NPM}\stackrel{hv}{⟶}{}^{\mathrm{3}}{\mathrm{NPM}}^{\ast }\text{(R2)}& {\displaystyle }{}^{\mathrm{3}}{\mathrm{NPM}}^{\ast }+{\mathrm{H}}_{\mathrm{2}}\mathrm{O}\to {\mathrm{H}}^{+}+\mathrm{•}\mathrm{OH}+{\mathrm{NPM}}^{-\mathrm{•}}\text{(R3)}& {\displaystyle }{\mathrm{NPM}}^{-\mathrm{•}}+{\mathrm{O}}_{\mathrm{2}}\to \mathrm{NPM}+{{\mathrm{O}}_{\mathrm{2}}}^{-\mathrm{•}}\text{(R4)}& {\displaystyle }{\mathrm{Fe}}^{\mathrm{3}+}+{\mathrm{H}}_{\mathrm{2}}\mathrm{O}\stackrel{hv}{⟶}{\mathrm{Fe}}^{\mathrm{2}+}+\mathrm{•}\mathrm{OH}+{\mathrm{H}}^{+}\end{array}$$

$$\begin{array}{}\text{(R5)}& {\displaystyle }\mathrm{2}{\mathrm{NO}}_{\mathrm{2}}+{\mathrm{H}}_{\mathrm{2}}\mathrm{O}\to {\mathrm{HNO}}_{\mathrm{2}}+{\mathrm{HNO}}_{\mathrm{3}}\text{(R6)}& {\displaystyle }\begin{array}{rl}{\displaystyle }{{\mathrm{O}}_{\mathrm{2}}}^{-\mathrm{•}}& {\displaystyle }+{\mathrm{NO}}_{\mathrm{2}}\to {{\mathrm{OONO}}_{\mathrm{2}}}^{-}\to {\mathrm{O}}_{\mathrm{2}}+{{\mathrm{NO}}_{\mathrm{2}}}^{-}\\ {\displaystyle }& {\displaystyle }\stackrel{+{\mathrm{H}}^{+}(\text{pKa}=\mathrm{3.2})}{⟶}\mathrm{HONO}\end{array}\text{(R7)}& {\displaystyle }\begin{array}{rl}{\displaystyle }{{\mathrm{O}}_{\mathrm{2}}}^{-\mathrm{•}}& {\displaystyle }+\mathrm{NO}\to {\mathrm{OONO}}^{-}\left({{\mathrm{NO}}_{\mathrm{3}}}^{-}\right)\leftrightharpoons \mathrm{OONOH}\\ {\displaystyle }& {\displaystyle }\stackrel{+{\mathrm{OONO}}^{-}}{⟶}{\mathrm{O}}_{\mathrm{2}}+\mathrm{2}{{\mathrm{NO}}_{\mathrm{2}}}^{-}+{\mathrm{H}}^{+}\stackrel{\text{pKa}=\mathrm{3.2}}{⟶}\mathrm{HONO}\end{array}\text{(R8)}& {\displaystyle }{{\mathrm{NO}}_{\mathrm{2}}}^{-}+\mathrm{•}\mathrm{OH}\to {\mathrm{NO}}_{\mathrm{2}}+{\mathrm{OH}}^{-}\text{(R9)}& {\displaystyle }{{\mathrm{NO}}_{\mathrm{2}}}^{-}+{\mathrm{H}}^{+}\stackrel{\text{pKa}=\mathrm{3.2}}{⟶}\mathrm{HONO}\text{(R10)}& {\displaystyle }{\mathrm{Fe}}^{\mathrm{2}+}+{{\mathrm{NO}}_{\mathrm{3}}}^{-}+\mathrm{2}{\mathrm{H}}^{+}\to {\mathrm{Fe}}^{\mathrm{3}+}+{\mathrm{NO}}_{\mathrm{2}}\left(\mathrm{g}\right)+{\mathrm{H}}_{\mathrm{2}}\mathrm{O}\text{(R11)}& {\displaystyle }{{\mathrm{NO}}_{\mathrm{3}}}^{-}\stackrel{hv}{⟶}[{{\mathrm{NO}}_{\mathrm{3}}}^{-}{]}^{\ast }\to \mathrm{O}{(}^{\mathrm{3}}\mathrm{P})+{{\mathrm{NO}}_{\mathrm{2}}}^{-}\stackrel{{\mathrm{H}}^{+}}{⟶}\mathrm{HONO}\text{(R12)}& {\displaystyle }{{\mathrm{NO}}_{\mathrm{3}}}^{-}\stackrel{hv}{⟶}[{{\mathrm{NO}}_{\mathrm{3}}}^{-}{]}^{\ast }\stackrel{{\mathrm{H}}^{+}}{⟶}{\mathrm{NO}}_{\mathrm{2}}+\mathrm{•}\mathrm{OH}\end{array}$$

A simplified illustration of the reaction mechanism is shown in Fig. S4 in the Supplement. As shown in Fig. S3, the HONO and NO₂ production during the photodegradation of NPM in the presence of Fe³⁺ is significantly enhanced relative to that in the absence of iron ions. High Fe³⁺ concentrations (0.5–0.8 mg mL⁻¹) promote HONO and NO₂ formation compared to low Fe³⁺ concentrations (0.25–0.5 mg mL⁻¹). The formed ${{\mathrm{NO}}_{\mathrm{3}}}^{-}$ and ${{\mathrm{NO}}_{\mathrm{2}}}^{-}$ were also measured by ion chromatography analysis to evaluate the effect of Fe³⁺ (see the details in Test S1 and Fig. S5 in the Supplement). As shown in Fig. S5, the concentrations of ${{\mathrm{NO}}_{\mathrm{3}}}^{-}$ and ${{\mathrm{NO}}_{\mathrm{2}}}^{-}$ decreased sharply in the presence of Fe³⁺ compared to those in the absence of Fe³⁺. These results suggest that HONO and NO₂ enhancement during the irradiation of NPM solutions containing Fe³⁺ can be ascribed to the transformation of the product distribution from ${{\mathrm{NO}}_{\mathrm{3}}}^{-}$ and ${{\mathrm{NO}}_{\mathrm{2}}}^{-}$ rather than a change in the product formation from the photodegradation of NPM.

4 Conclusions and outlook

Laboratory study revealed the formation of a greenhouse gas (N₂O) through photolysis of NPM (Aregahegn et al., 2018), but previously the theoretical calculation had predicted that photolysis of NNs would generate NO₂ (Palma et al., 2020).

The current study reveals that the light-induced degradation of NPM leads to enhanced production of HONO and NO_x driven by secondary photochemistry between redox reaction of ${\mathrm{Fe}}^{\mathrm{3}+}/{\mathrm{Fe}}^{\mathrm{2}+}$ and photoproduced ROS. We quantified the photochemical HONO and NO_x formation through NPM photodegradation, and we suggest that this chemistry may represent a significant source of HONO and NO_x in regions where surface waters are polluted with NN insecticides.

In order to estimate the relative importance of the NPM photolysis to global HONO and NO_x emissions in the atmosphere, we parameterized the global HONO and NO_x production related to NPM photochemistry, based on the NPM photolysis kinetic data and gridded downward solar radiation. The parameterization of HONO and NO_x production from NPM photolysis at the Fe³⁺ concentration of 0.025 mg L⁻¹ used in our estimation is based on Eqs. (S1)–(S3). The concentrations of NNs vary from several nanogram per liter to hundreds of microgram per liter (Anderson et al., 2013). In view of the high concentration of NPM (50 000 µg L⁻¹) used in our experiments, we selected a rationalization parameter scheme related to the environmental concentration of NPM (50 µg L⁻¹). The kinetic data have shown that the rate constant (k) is faster at low NPM concentrations compared to that of high NPM concentrations (Fig. S6 in the Supplement). Current chemical models do not explicitly consider this source of reactive nitrogen species. In this way, we are able to generate an hourly dataset of the NO_x and HONO fluxes released from NPM chemistry, and we analyze the amounts and spatial patterns of the fluxes in Fig. 5. We note that, although such estimation is rather simplified and can be biased in terms of the spatial heterogeneity as we do not consider the spatial variation of environmental NPM concentrations, our study represents a pioneering attempt to quantify the global source of HONO and NO_x from the NPM photochemistry, as current chemical models do not explicitly consider this source of reactive nitrogen species. This inventory can then be applied in chemical models to quantify the environmental impact of HONO and NO_x fluxes emerging from NPM photochemistry. Details about the parameterization of HONO and NO_x production that emerged from NPM photochemistry are given in Text S2 in the Supplement. Figure 5 shows the spatial distributions of HONO and NO_x fluxes produced by NPM photochemistry in the tested year of 2017. The results indicate that the globally produced HONO and NO_x fluxes based on NPM photochemistry are 0.77 and 0.5 Tg N yr⁻¹, respectively, making a total of 1.27 Tg N yr⁻¹.

The total production of HONO and NO_x emissions due to NPM photochemistry (1.27 Tg N yr⁻¹) represents 3.5 % of the anthropogenic emissions of NO_x related to fossil fuel in the year 2017 (36.2 Tg N yr⁻¹, from the Community Emissions Data System (CEDS) inventory) and about 14.8 % of the soil emissions (8.6 Tg N yr⁻¹, Lu et al., 2021). The highest HONO and NO_x fluxes (74 %) are produced by the photochemistry of NPM at the ocean surface in the presence of iron ions, especially tropical oceans. The latter can be ascribed to the higher solar radiation in the tropical regions. As displayed in Fig. S7 in the Supplement, it is obvious to see that the spatial distribution of solar radiation is particularly strong in tropical oceanic regions, which can further confirm the higher HONO and NO_x fluxes at the ocean surface. The high reactive nitrogen emissions could also appear over other water surfaces like inland waters and lakes worldwide through similar mechanisms induced by NPM photochemistry. Further studies are needed to quantify the relative importance of the recognized HONO and NO_x sources from NPM photochemistry on a global scale as well as the impact on tropospheric ozone and OH in the marine boundary layer.