Design and Feasibility Evaluation of a Prototype Setup for Contemporary Easy Nitrates and Nitrites UV Detection in Water for Agriculture

TL;DR

This paper introduces a prototype device for quickly and easily detecting nitrates and nitrites in water used for agriculture and other purposes.

Contribution

A new prototype setup for real-time, chemical-free detection of nitrates and nitrites in water is proposed and tested.

Findings

The device achieved detection limits of 2.2 mg/L for nitrates and 0.5 mg/L for nitrites.

Common substances in water did not significantly interfere with the detection process.

The setup uses UV absorption at 302 nm and 355 nm for simultaneous detection of nitrates and nitrites.

Abstract

Nitrates and nitrites are inorganic anions which, beyond specific concentration threshold, are classified as water pollutants. Nitrate compounds are commonly used as fertilizers; however, their high concentration in soil and in wastewater, as well as their reduction to nitrites, pose serious environmental and human health risks. Therefore, detecting these ions in water intended for human consumption, zootechnical use, and agricultural applications is essential. This work presents a proof of concept for a spectroscopic prototype setup enabling simple, direct, and simultaneous detection of nitrates and nitrites in water. The device employs solid-state sensor technology and requires no sample pretreatment or chemicals. Ultimately, this apparatus will allow real-time, in-line process analysis. UV absorption bands centered at approximately 302 nm and 355 nm were selected for detecting nitrates and nitrites, respectively. Because nitrite exhibits a slight absorption at 302 nm as well, a straightforward method for simultaneous nitrate and nitrite detection is proposed. The proposed system incorporates a UV deuterium lamp, a 10 cm path length optical cuvette, and a custom home-built silicon carbide detector. This configuration enables testing various concentrations, achieving detection limits of 2.2 mg/L for nitrates and 0.5 mg/L for nitrites. Potential interferences from substances commonly found in drinking and treated agricultural wastewaters, including sodium bicarbonate, sodium sulfate, ammonium chloride, hydrogen peroxide, and sodium hypochlorite, were also investigated. Finally, a compact on-site and online monitoring future device is illustrated.

Figures

Full text

1. Introduction

Primary sources of nitrogen in the environment are urban wastewater, industrial discharges, and synthetic fertilizers. Once in the environment, nitrogen often exists as nitrogen oxides, which can interact with water to produce acid rain [1]. Anthropogenic activities including agriculture, combustion processes, transportation, have significantly amplified the input of nitrogen compounds into natural biogeochemical cycles [2]. Excesses of these compounds can lead to eutrophication, threatening ecosystems. For this reason, monitoring nitrogen levels in water is crucial to ensure compliance with environmental standards. Among nitrogen compounds, nitrates and nitrites are especially important, as their presence in natural waters serves as a key indicator of water quality and they play essential roles in the nitrogen cycle within soil systems.

Nitrate, widely recognized as a pollutant in both water and soil, is commonly used as a fertilizer [3]. However, its use extends beyond agriculture; it also finds applications as an oxidizing and explosive agent in numerous industrial applications. For instance, nitrate is functional to the production of various types of glass and is included in the composition of fiberglass and charcoal briquettes. Additionally, it plays a role in rubber vulcanization, petrochemical processes, and in the production of enamels, porcelain, and heat-transfer salts used in metal treatment. Nitrate-based compounds are also employed in water treatment, industrial cleaning, pharmaceutical manufacturing, and even in the extraction of metals like lead, uranium, and copper [4].

Nitrites, on the other hand, act as transitional products in the nitrogen cycle. They are formed during the degradation of nitrogen-rich organic substances, such as ammoniacal nitrogen and nitrates. Due to their chemical instability, they may be further reduced to ammonia or oxidized back to nitrates in the presence of dissolved oxygen. In some cases, their presence in water may result from industrial water treatment where nitrites act as corrosion inhibitors. Like nitrates, they enter aquatic environments through urban runoff, industrial effluents, and agricultural drainage from fertilized soils [5].

For humans, food and drinking water are the primary sources of inorganic nitrogen intake. Nitrate consumption ranges from 75 to 100 mg per day, with vegetables contributing about 80–90% and water around 5–10% of this intake [6,7]. Despite being less toxic than nitrites, nitrates can pose health risks when consumed in large quantities or when converted into nitrites in the body. Nitrites themselves are more harmful: they can react with secondary or tertiary amines to form N-nitrosamines, compounds known to be carcinogenic, mutagenic, and also teratogenic [8]. Moreover, nitrites can oxidize hemoglobin into methemoglobin, causing methemoglobinemia. This disease impairs the blood’s capacity to transport oxygen, a condition especially dangerous for infants, commonly referred to as “blue baby syndrome”.

European and Italian legal limits are established by the “Directive (EU) 2020/2184 of the European Parliament and of the Council of 16 December 2020 on the quality of water intended for human consumption” and by the Legislative Decree 18/2023 respectively. The maximum admitted concentration in drinking water is 50 mg/L for nitrates and 0.5 mg/L for nitrites. For waters whose sources are treatment plants that may generate nitrite ions through the oxidation of ammonium ions, the reduction of nitrate ions, or the hydrolysis of chloramines, a value of 0.10 mg/L is applied to the nitrite parameter. On the other hand, Italian wastewater limits are established by the Legislative Decree 152/2006. It distinguishes between discharge into surface waters, sewer system and soil. For surface waters, the limits are 20 mg/L for nitric nitrogen and 0.6 mg/L for nitrous nitrogen. For the sewer system, up to 30 mg/L of nitric nitrogen is tolerated. In the case of soil discharge, the legal limit is set at 15 mg/L of total nitrogen.

Moreover, nitrates and nitrites can be considered as indicators of potential water pollution; in fact, elevated levels of these inorganic ions suggest the presence of other contaminants, microbial pathogens, or pesticides. So, their detection represents a good starting point to verify the pollution level of the body of water.

Both direct and indirect methods can be adopted to detect nitrate and nitrite ions [9,10]. Traditional analytical techniques include flow-injection methods [11,12,13], which are useful for their high analytical throughput, minimal sample volume requirements, and operational simplicity. However, these methods necessitate the use of chemicals.

Electrochemical approaches, like potentiometric [14,15], amperometric [16,17], and voltammetric [18,19,20] methods, are also proposed. These techniques are relatively straightforward, often requiring little to no sample preparation, though they are limited by issues related to selectivity, stability, and portability.

Chromatographic techniques, particularly ion chromatography [21,22,23], high-performance liquid chromatography (HPLC) [24], and ultra-performance liquid chromatography (UPLC) [25,26], offer high precision and sensitivity. However, they are generally laborious, time-intensive, and demand sophisticated instrumentation and skilled personnel, making them less suitable for on-site or real-time monitoring applications.

In recent years, the demand for low-cost, portable, and real-time monitoring technologies has grown, especially in the context of agricultural water management. This has led to the development of various types of devices, including electrochemical sensors, biosensors and spectroscopic sensors [27]. Electrochemical ones are known for their rapid response and high sensitivity. Moreover, these low-power devices demand minimal sample preparation and offer significant potential for miniaturization, making them well-suited for portable applications. Farina et al. realized an electrocatalytic sensing system for nitrates exploiting copper micro-flowers [28]. Pogăcean et al. developed an electrochemical sensor for nitrites based on graphene/glassy carbon electrode (EGr/GC) [29]. Nanomaterials integrated with sensing elements have recently gained significant attention in sensing applications, due to their nanoscale dimensions, which offer a high surface-to-volume ratio. This enhanced surface ensures a more efficient interaction with analytes, leading to improved sensitivity and the ability to detect trace concentrations [30,31,32,33,34]. In biosensing platforms, biological recognition elements are integrated with a signal detection and processing system to measure the ion concentration in a sample [35,36,37]. These biological components may include DNA, enzymes, antibodies, receptor proteins, or even whole cells. In particular, for nitrate detection, biosensors based on nitrate reductase (NR) are exploited [38,39,40].

One of the most used techniques is UV spectroscopy. This methodology provides high precision and detection efficiency, nondestructive sampling and low cost [41]. Many methods for detecting nitrates exploit the absorption peak at 220 nm [42,43]. However, this approach is most suitable for unpolluted water samples with low levels of organic matter, since these compounds may absorb at this wavelength, leading to interference. Although some authors have attempted spectral correction techniques, their success is limited. Additionally, due to the spectral overlap between nitrate and nitrite in this range, it is not feasible to determine both simultaneously.

Nevertheless, nitrates and nitrites exhibit distinct spectral features at longer wavelengths, between 250 and 390 nm: nitrates show a peak at 302 nm, whereas nitrites absorb at 355 nm and display a minor shoulder at 290 nm [44,45]. These differences were the basis for the method developed by Wetters and Uglum, which allows simultaneous analysis of both ions [46]. Since the absorbance of nitrite at 355 nm is 2.5 times higher than the absorbance at 302 nm, a correction factor is adopted, enabling the estimation of nitrite concentration. Another solution was proposed by Wang et al. [47] combining UV-Vis spectrophotometry with a second-order differential method to solve the absorbance spectra overlapping problem. Over the past thirty years, spectroscopic sensors have gained prominence for water quality monitoring. Their compact size, sensitivity, and low energy demands make them particularly well-suited for remote sensing. Brandl et al. developed an automatic system for nitrate and nitrite detection in water bodies [48]. It exploits an RGB sensor and the method adopted is based on the Greiss reaction. Ingles et al. described a portable nitrate sensor that uses a smartphone camera for nitrate sensing [49]. As the best of our knowledge, research in nitrates and nitrites monitoring lacks the availability of systems that offer an optimal trade-off between detection limits and cost-effectiveness. In fact, even if electrochemical, biochemical and nanomaterial-based sensors ensure higher sensitivity and detection limits in the order of part of billions, they face several issues like high operational costs and hard application [50,51,52]. In this paper we propose an easy and low-cost approach that exploits UV absorbance for the direct and simultaneous monitoring of nitrates and nitrites. In this context, a spectroscopic prototype setup adopting a bench lamp, a cylindrical cuvette and a Silicon Carbide (SiC) UV photodiode is discussed [53]. SiC technology emerges as the most suitable candidate for the proposed application, as silicon carbide is a wide-bandgap semiconductor used in power microelectronics. Its unique electronic, optical, and chemical properties make it particularly well suited for UV photo-sensing applications. Specifically, its UV sensitivity in the 220–380 nm range and visible-light blindness allow the direct detection of nitrate and nitrite absorption in the UV region, without interference from eventual visible-light fluorescence. Furthermore, its potential for miniaturization, demonstrated by previously developed portable SiC-based devices, supports integration into handheld systems, as shown by Sciuto et al. [54]. Finally, its chemical inertness and thermal stability ensure robustness and long-term durability. The proposed SiC-based UV spectroscopic system also avoids any sample preparation. It was able to achieve 2.2 mg/L and 0.5 mg/L sensitivity to nitrates and nitrites. Thanks to the direct measurements, the sensing system represents a good candidate for nitrogen compound detection, in pre-treated agri-food wastewater set-up, and it could be easily integrated into wireless networks for real-time and widespread monitoring in precision agriculture. In particular, the use of solid-state components and no need for reagents make the system highly suitable for long-term field deployment. According to the available literature, this study represents the first SiC-based technology application to the monitoring of nitrates and nitrites in water.

2. Materials and Methods

2.1. Reagents

Sodium nitrate, sodium nitrite, ammonium chloride, sodium bicarbonate, and sodium sulphate (all with a purity of 99%) were used without further purification. A 30% hydrogen peroxide solution and sodium hypochlorite solution containing 6–14% active chlorine were used as received. All reagents were purchased from Sigma-Aldrich, St. Louis, MO, USA.

Stock solutions of nitrates and nitrites were prepared by dissolving solid sodium nitrate (MW = 85.0 g/mol) and sodium nitrite (MW = 69.0 g/mol) in deionized water. Solutions containing ammonium, bicarbonate, and sulphate ions were prepared using ammonium chloride (MW = 53.45 g/mol), sodium bicarbonate (MW = 84.0 g/mol), and sodium sulphate (MW = 142.0 g/mol), respectively. Sodium hypochlorite and hydrogen peroxide solutions were obtained by diluting their respective concentrated aqueous stock solutions. Working standards solutions were freshly prepared as needed through appropriate stepwise dilutions.

2.2. Instruments and Experimental Setup

UV-Vis absorption spectra of sodium nitrate and nitrite solutions were acquired by a bench-top spectrophotometer model SR-4UV240-25 (Ocean Optics, Orlando, FL, USA) to pick their relative characteristic signals.

The proposed spectroscopic setup consists of three main components: a bench-top deuterium UV light source (SP-ASBN-D130, Spectral Products, Putnam, CT, USA), a cylindrical cuvette (Sigma-Aldrich^®^, Hellma^®^, Müllheim, Germany, PN: Z803634) with a 10 cm optical path, and a custom-made, low-noise, high-responsivity, visible-blind SiC UV detector (CNR-IMM HQ facilities, Catania, Italy). The light source is equipped with hard-coated bandpass filters (FBH300-10 and FLH355-10, Thorlabs Newton, NJ, USA), centered at 300 nm and 355 nm with a full width at half maximum (FWHM) of 10 nm. Light passes through the cuvette (Figure 1b), and transmitted beam is collected by the SiC detector. This latter is biased at a fixed voltage of −10 V using a Keithley 2636B Source Measure Unit (SMU, Keithley Instruments, Cleveland, OH, USA), which records the photocurrent in the nA range as we have already described in previous studies (Di Bari et al. [55] and Sciuto et al. [56]). The working polarization of −10 V was identified as the best compromise in terms of detector dark current and optical response. The SiC photodiode, at this voltage, exhibits a dark current value of about 0.1 nA and responsivity values of about 0.12 A/W at 300 nm and 0.04 A/W at 355 nm respectively [53,54]. All detector performances are widely reported in SI. A picture of the entire experimental setup is reported in Figure 1. In (a) the SMU, the home-made cuvette holder and the lamp are depicted. The detector is hosted in a 3D-printed PLA head and connected to the SMU with BNC cables; a second head, on the opposite side with respect to the photodetector, hosts the optical filter, which is aligned with the light beam. In (b) a picture of the commercial 10 cm cylindrical cuvette hosted in the opportunely 3D-printed holder is shown (without the shadowing cover).

Lamp emission spectra were measured using a multimodal UV-Vis fiber coupled to a bench-top spectrometer (AvaSpec-HS2048XL-EVO, Avantes B.V., Apeldoorn, The Netherlands), both in the presence and absence of optical filters. Spectra were acquired with integration times of 13 ms (without filters) and 16 ms (with filters), and they are shown overlapped in Figure 2. As expected, filtered spectra reveal distinct peaks centered at 300 nm and 355 nm, each with a FWHM of approximately 10 nm.

2.3. Procedure and Photocurrent Measurement Conditions

Measured photocurrent is proportional to the photon flux. It decreases when chromophoric species, present along the optical path, absorb UV light emitted by the source. More specifically, nitrate and nitrite ions in solutions absorb incident UV light at 300 nm and 355 nm. Consequently, SiC detector reveals a lower number of photons leading to an intensity current decrease in correspondence of increasing concentrations of absorbing molecules. Both dark and photocurrents were monitored to verify the system stability.

Six standard solutions of nitrates and ten standard solutions of nitrites were prepared and analyzed in the ranges between 10 ÷ 80 mg/L and 2 ÷ 75 mg/L respectively. Each standard was analyzed in triplicate and before testing the next solution, the cuvette was washed with deionized water.

3. Results and Discussion

Nitrate and nitrite ions absorb electromagnetic radiation in the UV region because of their molecular structure. Main absorption peaks can be found in two specific regions: one between 190 and 240 nm and the other one between 250 and 400 nm [57].

The study here proposed targeted the absorption peaks at 302 and 355 nm, relatively to n–π* transitions, for nitrates and nitrites respectively, clearly visible in Figure 3 where the overlapping of normalized absorption spectra of a sodium nitrate solution (with a [NO_3_^−^] = 840 mg/L) and a sodium nitrite solution (with a [NO_2_^−^] = 860 mg/L) is reported. By the spectra, is also evident that nitrites present a slight absorption in the same region of nitrates.

For this reason, a method for detecting both nitrates and nitrites is proposed. Experimental measurements were conducted with the aim of testing the suitability of the spectroscopic system. Complete apparatus, already described in previous lines and depicted in Figure 1, includes 300 nm and 355 nm optical filters, for selective detecting of nitrates and nitrites. These wavelengths lie within the SiC photodetector spectral responsivity as widely discussed in previous author works and also in Supplementary Materials [53,54,55,56].

In conducted experiments, six standard solutions of known concentration of nitrates and ten standard solutions of nitrites were prepared and analyzed using a 10 cm cylindrical cuvette, as listed in Table 1.

On each solution measurements were performed three times to test consistency of the data collected for any concentration. Reference photocurrent under full illumination I0 was acquired. Intensity photocurrent value is about 285.0 nA when 300 nm optical filter is used, while it is about 76.5 nA using the 355 nm one.

Considerable current reduction ΔI = I0 − I was observed for both anions. ΔI of 6, 15, 20, 24, 31 and 45 nA with increasing concentration in the case of nitrates moving from the lower concentration to the higher one. As for nitrites, ΔI equal to 1, 2, 3, 8, 15, 21, 26, 31 and 41 nA were obtained. Therefore, it is possible to state that there is a correlation between current decreasing and analyte concentration. Considering optical signal attenuation, A = I/I0, as the ratio between the intensity current in the presence of chromophore species I and the reference blank current (nitrate or nitrite-free baseline) I0 is more convenient. The A was calculated in triplicate for each concentration, and the corresponding mean values and relative standards deviations are reported in Table 2 and Table 3, for nitrates and nitrites respectively.

Lambert–Beer law was exploited to build calibration curves. The goal is to find a correlation between the measured optical attenuation (related to absorbance Abs) and the concentration of nitrates or nitrites expressed in mg/L. As is well known, this law states that the optical absorbance Abs is directly proportional to the concentration c of a chromophoric species and the optical path length d through the sample, by the following relationship:

$$[eqn]$$

where ε is the molar absorption coefficient. Absorbance values were obtained applying the decimal negative logarithm of optical attenuation A, by $[eqn]$ . Two calibration curves were generated by plotting the medium absorbance value, obtained for each solution (Abs) vs. concentration. Obtained results are reported in Figure 4 and Figure 5.

Linear regression of the plotted data in Figure 4 and Figure 5 resulted in R^2^ = 0.995 for nitrates and R^2^ = 0.999 for nitrites. These values confirm the direct proportionality in the concentration ranges here investigated. Molar absorption coefficients at target wavelengths (300 and 355 nm) were estimated by the ratio of the slopes and the optical path length (d = 10 cm). Calculated values are $[eqn]$ = 5.58 L/mol × cm corresponding to 9.0 × 10^−4^ L/mg × dm and $[eqn]$ = 20.7 L/mol × cm corresponding to or 4.5 × 10^−3^ L/mg × dm. Obtained values are in full agreement with the data reported in the literature [44,58,59]. Nitrates and nitrites limits of detection (LoDs) and limits of quantification (LoQs) at target wavelengths were determined. Ten blank intensity current measurements (I0), for each wavelength, were registered obtaining standard deviation values of 0.4 at 300 nm and 0.1 at 355 nm, and consequently absorbance standard deviations of 6.1 × 10^−4^ and 7.4 × 10^−4^ respectively. LoD was calculated as follows:

$$[eqn]$$

where σ_b_ is the absorbance standard deviation and m is the slope of the calibration curve. Calculated values are equal to 2.2 mg/L for nitrates and 0.5 mg/L for nitrites. LoQ, instead, is defined as:

$$[eqn]$$

The equation provides LoQ values of 6.8 mg/L and 1.6 mg/L for nitric and nitrous species respectively. LoD and LoQ values confirm the good sensitivity for both species. As known and shown in Figure 3, nitrites slightly absorb in the region between 250 nm and 310 nm. So, in a solution containing both nitrates and nitrites it is not possible to determine nitrate concentration directly. A simple way to determine both nitrates and nitrites is based on the principle of absorbance additivity [46]. It states that the total absorbance of a mixture of substances is equal to the sum of the absorbances of each individual component at a given wavelength. Therefore, it is possible to set up the following equation:

$$[eqn]$$

where $[eqn]$ is the total absorption due to the presence of both nitrates and nitrites, $[eqn]$ and $[eqn]$ are their molar absorption coefficients at 300 nm, $[eqn]$ and $[eqn]$ are the concentrations values and d is the optical path. By rearranging Equation (4):

$$[eqn]$$

So, once nitrite concentration is determined at 355 nm, a region where nitrates do not absorb UV light, it is possible to apply this equation. $[eqn]$ was determined setting up another calibration curve (Figure 6). Absorbance values of standard solutions of nitrite were plotted as a function of concentration (mg/L), operating at 300 nm wavelength. Optical attenuation values are reported in Table 4.

Plotted data in Figure 6 provides R^2^ = 0.998. Relative molar extinction coefficient is equal to $[eqn]$ = 7.82 L/mol × cm corresponding to 1.7 × 10^−3^ L/mg × dm. This parameter allows nitrite LoD and LoQ determination at this wavelength. Obtained results are 1.2 mg/L and 3.6 mg/L respectively. This implies that nitrite ion contribution to the total absorbance measured at 300 nm is reliable above a concentration value of 3.6 mg/L.

The effectiveness of Equation (5) was tested by preparing four solutions containing a mixture of nitrates and nitrites. Their relative concentrations are reported in Table 5.

All solutions were first analyzed using 355 nm wavelength to extrapolate nitrite concentration values from the calibration curve and compare them with the real ones (Table 6). Absorbances of the same solutions were then registered at 300 nm. Equation (5) application yielded the results reported in Table 7.

Calculated nitrate concentrations are in very good agreement with real ones, although a slight overestimation can be observed. In particular, data relative to solution 4 show a higher deviation. This occurs because the solution has a nitrite concentration of 3.1 mg/L, slightly below its LoQ at 300 nm (3.6 mg/L). Figure 7 shows a graphical comparison among real nitrate concentrations, values calculated using nitrite concentrations obtained via Lambert–Beer law and those derived by extrapolating nitrite concentrations from the calibration curve.

Another series of measurements was conducted on aqueous solutions containing substances potentially present in treated wastewater that might interfere with the detection of target ions. Two types of substances were analyzed: inorganic ions, commonly present in natural water, and additives used as disinfectants. In particular, ammonium (NH_4_^+^), bicarbonate (HCO_3_^−^) and sulphate ions (SO_4_^2−^) for the first class, and sodium hypochlorite (NaClO) and hydrogen peroxide (H_2_O_2_) for the second one, were tested. Photo-intensity current values measured for standard solutions of these substances are reported in Table 8 and Table 9 adopting irradiation through optical filters at 300 nm and 355 nm respectively.

Photocurrent intensity values suggest that ammonium, sulphate and bicarbonate ions and hydrogen peroxide, for concentration values below the maximum one tested, do not absorb UV light at target wavelengths. In fact, these current intensity values are not significantly different from distilled water ones (284.8 ± 0.4 nA at 300 nm and 76.4 ± 0.1 nA at 355 nm). On the other hand, sodium hypochlorite causes a high current intensity decrease. Notably, a concentration value of 1.0 mg/L, at 300 nm, causes a 99.9% reduction in the photocurrent signal intensity. This is because sodium hypochlorite is characterized by a strong absorption peak at 292 nm; at this wavelength, its molar absorption coefficient is equal to 350 L/mol × cm [60]. This implies that the system would be able to detect concentration values in the order of micrograms per liter. So, if an agri-food processing plant makes use of sodium hypochlorite as disinfectant, the approach proposed here could be not directly adopted. Verifying the absence of sodium hypochlorite traces in treated water samples becomes essential. Various commercial kits for active chlorine monitoring can be exploited, in portable versions too. Despite this, in agri-food industry, sodium hypochlorite or hypochlorous acid are not always used; for example, sodium hydroxide is sometimes adopted [61].

Furthermore, the proposed UV sensing system is conceived as a part of monitoring stage within a broader integrated wastewater reclamation and valorization framework. In this context, the sensor is not intended for direct analysis of untreated raw wastewater, but rather for post-treatment monitoring of residual nitrate and nitrite concentrations after upstream purification processes including advanced membrane bioreactor [62] and innovative adsorption materials [63,64] to significantly reduce suspended solids, dissolved organic matter (DOM), and organic/inorganic contaminants. As result, the residual water matrix entering the proposed spectroscopic device is expected to present low turbidity and limited spectral interferences in the 290 ÷ 400 nm region. Therefore, the role of the solid-state based system is to provide real-time, reagent-free, final compliance monitoring of nitrate and nitrite concentrations in treated effluents, ensuring that regulatory limits are not exceeded before reuse in agricultural applications.

4. Conclusions and Outlook

This study demonstrates the potential of SiC-based spectroscopic system for the optical detection of nitrates and nitrites, paving the way for applications in post wastewater treatment system monitoring, such as ones in the field of precision agriculture.

Stable performance and high sensitivity have been confirmed, as the system detected nitrate and nitrite concentrations as low as 2.2 and 0.5 mg/L (corresponding to 3.6 × 10^−5^ mol/L and 1.1 × 10^−5^ mol/L) respectively. Italian law establishes wastewater limits of 20 mg/L and 0.6 mg/L for nitric and nitrous nitrogen. These values correspond to 88.6 mg/L of nitrate and 2.0 mg/L of nitrite. On the other hand, European and Italian drinking water limits are set at 50 mg/L for nitrates and 0.5 mg/L for nitrites. So, our system was able to detect a nitrate concentration twenty times lower than drinking water limit. At the same time, nitrite LoD is four times lower than nitrate one and quite close to its legal limit. The next goal is to reach higher sensitivity towards nitrite ions. Increasing the optical path length is a possible solution to this issue.

Regardless, the obtained results are very promising; an optimal balance between detection limits and cost-effectiveness has been reached. To the best of our knowledge, SiC technology has found this application for the first time. Different to other photodetection devices, such as UV-enhanced silicon-based detectors, even equipped with visible filters, SiC sensors do not suffer from low UV response and UV–visible rejection ratio (see Supplementary Materials), ensuring the absence of collateral fluorescence phenomena.

To contextualize and emphasize the significance of the results obtained in this study, Table 10 provides a comparison between our sensor system and other optical detection platforms reported in the literature. We reach system performances similar to ones in the literature in terms of LoD in detecting nitrate and nitrite by a free-reagent approach. Our system also permits them to discriminate against their contemporary presence in mixture in concentration ranges close to their respective LoQ.

The integration of solid-state components within a monitoring technique reagent-free, unnecessary recalibration, and robustness under harsh environmental conditions ensure long-term field deployment and support the development of a wireless real-time platform for remote control. This capability is particularly advantageous for precision agriculture applications and the optimization of pest management strategies. Additionally, SiC technology offers high compatibility with the future integration of sensing, signal acquisition, and on-board data processing electronics. In Figure 8 we propose a representative scheme of a potential portable sensor system envisioned for nitrate and nitrite detection in agricultural contexts. This apparatus is expected to incorporate UV LED (light emitting diode) array, UV photodetector, fluidic tubing to enable water inflow and outflow maintaining uniform distribution within the chamber and the necessary LED controller and lecture electronic chain (LEC) for powering, data acquisition and remote operation. The cylindrical cuvette (Figure 1b), adopted in the experimental apparatus, mimics the final application, where it will be replaced by a realistic pipeline.