Glyphosate Detection Using a Fast‐Response Naphthalimide‐Al³⁺ Fluorescent Probe With Visual Monitoring and Smartphone‐Assisted Readout

ABSTRACT

The widespread use of glyphosate, a broad‐spectrum organophosphorus herbicide, has raised increasing concern due to its environmental persistence and potential health hazards. However, its high polarity and lack of chromophores make accurate detection challenging. Here, we report a naphthalimide‐based fluorescent probe (NAPA‐Al³⁺) for the rapid and sensitive detection of glyphosate. The probe is constructed by coordination between a pyridinamine‐modified naphthalimide fluorophore and Al³⁺ ions, resulting in quenched fluorescence. Upon the introduction of glyphosate, Al³⁺ preferentially coordinates with glyphosate, releasing the fluorophore and restoring fluorescence. The NAPA‐Al³⁺ probe detects glyphosate rapidly within 5 s and achieves a low detection limit of 57.5 nM, outperforming most reported glyphosate probes. Furthermore, portable test strips incorporating NAPA‐Al³⁺ were fabricated, which display a distinct blue‐to‐green color change with increasing glyphosate concentration. Smartphone‐assisted colorimetric analysis was further applied to achieve real‐time, visual, and quantitative detection. This work demonstrates a simple, low‐cost, and highly effective strategy for glyphosate monitoring, providing a promising platform for food safety regulation and environmental surveillance.

A naphthalimide‐based probe (NAPA‐Al³⁺) has been prepared, whose fluorescence is quenched. Glyphosate preferentially binds Al³⁺, thus releasing the fluorophore and restoring fluorescence. This strategy was integrated into portable test strips, enabling on‐site, visual glyphosate detection via smartphone analysis.

1. Introduction

As global dietary preferences shift from “having enough to eat” toward “eating healthily” and “eating safely,” the concept of green and pollution‐free food has gained increasing public attention. Among food safety concerns, pesticide residues in agricultural products have become a major focus worldwide. Glyphosate, a broad‐spectrum organophosphorus herbicide, has been extensively applied in agriculture due to its high efficiency, cost‐effectiveness, and nonselective action [1, 2]. It is widely present across the production chain of grains, vegetables, and fruits [3, 4, 5, 6, 7]. However, glyphosate exhibits a relatively long half‐life, leading to accumulation in crops, soil, and water, as well as bioaccumulation in organisms. This persistence raises serious health and ecological concerns [8, 9, 10, 11].

Excessive glyphosate exposure has been linked to teratogenic effects and damage to genetic, reproductive, and nervous systems [12, 13, 14, 15]. It can inhibit acetylcholinesterase (AChE) activity, impairing the central nervous system [16, 17, 18], and reduce steroid hormone production in males, thereby decreasing fertility [19, 20, 21]. Recognizing these risks, the World Health Organization (WHO) limits the residual amount of glyphosate in drinking water to 5.33 μM, and the regulatory authorities of the Chinese government have also set the maximum residue limit (MRL) of glyphosate in fruits at 0.5 mg/kg (2.96 μM) [22, 23, 24]. Consequently, developing cost‐effective, accurate, and convenient methods for glyphosate detection is critical for the protection of public health, the assurance of food safety, and the preservation of the environment.

Glyphosate detection is challenging due to its high polarity, low volatility, and lack of chromophores [25, 26, 27, 28]. Although methods such as high‐performance liquid chromatography (HPLC), atomic absorption spectroscopy (AAS), gas chromatography (GC), inductively coupled plasma mass spectrometry (ICP‐MS), enzyme‐linked immunosorbent assay (ELISA), and surface‐enhanced Raman scattering (SERS) have been established [29, 30, 31, 32], they often require sophisticated instrumentation, lengthy sample preparation, and high operational costs.

To overcome these limitations, research has increasingly focused on fluorescent probes, which offer advantages such as high sensitivity, low cost, rapid response, and visual readout [33, 34, 35]. In particular, probes based on competitive complexation have drawn considerable interest. This strategy exploits the strong binding affinity between glyphosate and metal ions, facilitated by the molecule's phosphate, carboxyl, and amino groups [36]. When a probe–metal complex is exposed to glyphosate, the metal ion preferentially coordinates with glyphosate, displacing the fluorophore and triggering a detectable fluorescence change [37, 38, 39, 40, 41]. However, many existing probes still suffer from limitations such as complex synthesis, slow response, poor sensitivity, or indistinct colorimetric changes, restricting their practical use in food safety monitoring.

Naphthalimide has emerged as a promising fluorophore due to its high fluorescence quantum yield, visible‐light emission (400–600 nm), photostability, and structural tunability [42, 43, 44, 45]. Herein, we designed and synthesized a naphthalimide‐based fluorophore containing a pyridinamine group as the metal ion binding site. The resulting probe (NAPA‐Al³⁺) is formed by coordination between the fluorophore (NAPA) and Al³⁺ ions, where fluorescence is quenched upon complexation. In the presence of glyphosate, Al³⁺ preferentially binds to glyphosate, releasing the fluorophore and restoring fluorescence (Scheme 1).

SCHEME 1.

Structure of NAPA and NAPA‐Al³⁺ complex and fluorescence response to glyphosate.

The NAPA‐Al³⁺ probe, which is readily obtained with two synthesis steps, enables rapid (5 s) and ultrasensitive (LOD: 57.5 nM) detection of glyphosate, outperforming many reported probes (Table S1). To enhance field applicability, we further fabricated NAPA‐Al³⁺ test strips. These strips exhibit a distinct color change from blue to green with increasing glyphosate concentration, allowing real‐time and visual detection. In combination with smartphone‐assisted colorimetric analysis, the NAPA‐Al³⁺ probe offers a simple, low‐cost, and portable platform for on‐site glyphosate monitoring, holding promise for food safety regulation and environmental surveillance.

2. Experimental Section

2.1. Materials and Instruments

The details regarding the reagents and chemicals as well as the instrumentation used are provided in the Supporting Information.

2.2. DFT Calculation for NAPA and NAPA‐Al³⁺

DFT calculations for NAPA and NAPA‐Al³⁺ were carried out using the Gaussian 09 software package. Geometry optimizations were performed for both NAPA and the NAPA‐Al³⁺ complex, which were subsequently analyzed to investigate their electronic properties and elucidate the fluorescence quenching mechanism.

2.3. Smartphone‐Based Colorimetric Analysis for Visual Detection

Fluorescence images of the test strips under 365 nm UV light were captured in a dark box using a smartphone. The red (R), green (G), and blue (B) values were extracted from these images via a color recognition application. A calibration curve was then generated by plotting the green‐to‐blue intensity ratio (G/B) against the glyphosate concentration. All measurements were conducted in triplicate.

3. Results and Discussion

3.1. Optical Spectral Response of NAPA to Al³⁺ and Complexation Mechanism

The fluorophore (NAPA) was through a two‐step reaction (Scheme 2) and fully characterized by high‐resolution mass and ¹H/¹³C spectra (Figure S1–S4). The probe NAPA‐Al³⁺ was obtained via the coordination between NAPA and aluminum ions.

SCHEME 2.

Synthetic route for NAPA.

We first examined the optical response of NAPA to Al³⁺. As shown in Figure 1A, NAPA displays an absorption maximum at ~426 nm and emits strong fluorescence with a maximum at ~519 nm. Screening experiments with various metal ions reveal that only Al³⁺ induces a pronounced fluorescence quenching effect, demonstrating the high binding specificity of NAPA toward Al³⁺ (Figure 1B). Although Al³⁺ induces the most pronounced fluorescence quenching, it can be observed that Hg²⁺ also causes a certain decrease in fluorescence intensity, though the effect is significantly weaker. Moreover, given the high toxicity and regulated nature of mercury ions, probes with high selectivity for Al³⁺ offer a safer and more practical alternative for routine analytical applications.

FIGURE 1.

(A) Absorption and fluorescence emission spectra of NAPA (20 μM) in ethanol/water solution (9:1, v/v). (B) Fluorescence spectra of NAPA and NAPA (20 μM) with the existence of Al³⁺ and other metal ions in ethanol/water solution (9:1, v/v) (metal ions: 20 μM).

To further evaluate this interaction, Al³⁺ titration experiments were performed. The fluorescence intensity of NAPA decreases progressively with increasing Al³⁺ concentration (Figure 2A), and visual inspection confirms a gradual fluorescence fading upon stepwise addition of Al³⁺ (20 μM; Figure S5). Notably, the fluorescence response occurs rapidly within 6 s and then stabilizes, suggesting fast and efficient complexation (Figure 2B). The binding constant was calculated to be 7.59 × 10⁴ M⁻¹ using a Benesi–Hildebrand plot (Figure 2C). Moreover, Job's plot analysis reveals a stoichiometric inflection at a 0.5‐M fraction of Al³⁺, consistent with a 1:1 complexation ratio between NAPA and Al³⁺ (Figure 2D). The complexation ratio of NAPA and Al³⁺ was verified by HRMS (Figure S2B). The HRMS result shows a characteristic peak at m/z 198.5347, corresponding to the [NAPA + Al³⁺ + Cl⁻ + H₂O]²⁺ adduct, which confirms the 1:1 coordination ratio between NAPA and Al³⁺—consistent with the result from Job's plot.

FIGURE 2.

(A) Fluorescence response of NAPA (20 μM) to increasing concentrations of Al³⁺ (0–20 μM) in ethanol/water solution (9:1, v/v). (B) Time‐dependent fluorescent response of NAPA to Al³⁺ (20 μM). (C) The Benesi–Hildebrand plot of NAPA‐Al³⁺ complex (Y = (−7.68) × 10⁻⁹ X − 5.83 × 10⁻⁴, R ² = 0.99). (D) Job's plot of NAPA's complexation with Al³⁺ (the total concentration of NAPA and Al³⁺ is 20 μM).

To gain mechanistic insights, frontier molecular orbital (HOMO–LUMO) calculations were conducted for NAPA and the NAPA‐Al³⁺ complex (Figure 3A). In free NAPA, the LUMO electron density is localized on the fluorophore, favoring radiative relaxation and strong fluorescence. Upon coordination with Al³⁺, however, the LUMO electron density shifts toward the binding site, which promotes non‐radiative decay pathways and thus accounts for the observed fluorescence quenching.

FIGURE 3.

(A) Calculated HOMO and LUMO energy diagrams for NAPA and NAPA‐Al³⁺ complex. (B) The FTIR spectra of the NAPA and NAPA‐Al³⁺ in the state of solid.

The binding mode was further validated by FTIR spectroscopy (Figure 3B). After Al³⁺ coordination, the bending vibration bands of the amino N–H group (1350–1200 cm⁻¹) and the weak stretching vibration of the C–N bond on the pyridine ring (1500–1430 cm⁻¹) both decrease in intensity. These spectral changes confirm that Al³⁺ coordinates with the two nitrogen atoms of the pyridinamine moiety, forming the NAPA‐Al³⁺ complex.

3.2. Detection of Glyphosate Using the NAPA‐Al³⁺ Complex

Glyphosate contains amino, phosphate, and carboxyl groups that exhibit strong affinity toward metal ions, enabling it to displace Al³⁺ from the NAPA‐Al³⁺ complex and restore the probe's fluorescence. As shown in Figure 4A, fluorescence titration reveals a concentration‐dependent enhancement of emission intensity upon incremental addition of glyphosate (corresponding fluorescence images in Figure S6). The response is remarkably rapid, reaching equilibrium within 5 s (Figure 4B).

FIGURE 4.

(A) Fluorescence response of NAPA‐Al³⁺ (20 μM) to increasing concentrations of glyphosate (0–40 μM) in ethanol/water solution (9:1, v/v). (B) Time‐dependent fluorescent response of NAPA‐Al³⁺ to glyphosate (40 μM). (C) The Benesi–Hildebrand plot of glyphosate‐Al³⁺ complex (Y = 1.22 × 10⁻⁸ X + 1.24 × 10⁻³, R ² = 0.98). (D) Job's plot for the NAPA‐Al³⁺ complex with glyphosate (20 μM is the total concentration of glyphosate and NAPA‐Al³⁺).

Binding analysis further confirms this strong interaction. The Benesi–Hildebrand plot (Figure 4C) yields a binding constant of 1.02 × 10⁵ M⁻¹, which exceeds that of NAPA with Al³⁺, indicating stronger complexation between glyphosate and Al³⁺. Job's plot (Figure 4D) shows an inflection at 0.5 M fraction, consistent with a 1:1 binding stoichiometry.

Fluorescence titration reveals a strong linear correlation between glyphosate concentration (0–20 μM) and emission intensity (Y = 28.73 X + 365.37, R ² = 0.97; Figure 5A), yielding a detection limit (LOD = 3σ/k) of 57.5 nM. Visual inspection under a 365‐nm UV lamp clearly demonstrates a distinct fluorescence transition: the bright emission (NAPA) → the quenched fluorescence (NAPA‐Al³⁺) → the restored green fluorescence (NAPA‐Al³⁺ + glyphosate) (Figure 5B).

FIGURE 5.

(A) Linear correlation between fluorescence intensity at 519 nm and glyphosate concentration (0–20 μM). (B) Photographs of NAPA sequentially being treated with Al³⁺ and glyphosate under ambient light and 365 nm UV light. (C) Fluorescence spectra of NAPA‐Al³⁺ with the existence of each pesticide (40 μM). (D) Fluorescence intensity of NAPA‐Al³⁺ with the existence of glyphosate (40 μM) and together with one of the other pesticides (40 μM). The orange bar represents the fluorescent intensity of NAPA‐Al³⁺ with the existence of each of the pesticides, and the green bar represents the fluorescence intensity of NAPA‐Al³⁺ with the existence of glyphosate and each of other pesticides.

To evaluate selectivity, seven structurally diverse pesticides (Figure S7) were tested. As shown in Figure 5C, only glyphosate triggers fluorescence recovery, whereas the others have negligible influence (images in Figure S8). Moreover, co‐incubation experiments confirm that the presence of other pesticides does not interfere with glyphosate‐induced fluorescence restoration (Figure 5D).

The effect of pH was also examined. Fluorescence responses of NAPA, NAPA‐Al³⁺, and NAPA‐Al³⁺ + glyphosate across different pH values (Figure S9) reveal that glyphosate detection is optimal under neutral to mildly acidic conditions.

Collectively, these results demonstrate that the NAPA‐Al³⁺ probe enables rapid, highly sensitive, and selective detection of glyphosate with strong anti‐interference capability.

3.3. Detection of Glyphosate in Real Samples, by Test Strips and Smartphone‐Based Applications

To validate the practical applicability of the NAPA‐Al³⁺ probe, recovery experiments were performed by spiking soil and vegetable samples (Chinese cabbage and lettuce) with known concentrations of glyphosate (Figure S10). As summarized in Table 1, the method achieves excellent recovery rates (99.05%–102.71%) with low relative standard deviations (RSD: 0.45%–2.15%), confirming the reliability and accuracy of NAPA‐Al³⁺ for glyphosate determination in complex real‐world matrices. These results highlight the potential of this probe for food safety monitoring.

TABLE 1.

Detection results of glyphosate in real samples.

Samples	Spiked	Found	Recovery	RSD
	(μM)	(μM)	(%)	(%, n = 3)
Chinese cabbage	2.00	2.00	100.13	0.91
	4.00	4.03	100.78	1.57
	6.00	5.99	99.70	1.33
Lettuce	2.00	2.05	100.12	0.92
	4.00	4.07	101.74	2.05
	6.00	5.93	99.05	2.15
Soil	2.00	2.05	102.71	1.54
	4.00	4.02	100.45	0.45
	6.00	5.92	99.27	0.81

To further enhance usability, we fabricated glyphosate detection test strips. As glyphosate concentration increases, the strips display distinct and easily distinguishable color changes. A standard color chart was established to correlate glyphosate concentrations with strip color responses (Figure 6A). Building on this, we integrated a smartphone‐based readout to enable quantitative analysis. The RGB values of the strip images captured by a smartphone reveal a clear color transition from blue to green with increasing glyphosate levels. A strong linear correlation (Y = 3.51 × 10⁻² X + 0.74, R ² = 0.98) is observed between the G/B ratio and glyphosate concentration (Figure S11).

FIGURE 6.

(A) The fluorescence's color changes of test strips upon response to different glyphosate levels (0–50 μM). (B) Photographs of test strip responses under 365 nm UV illumination, showing fluorescence recovery upon spiking real samples with glyphosate concentrations at 0, 10, and 20 μM and smartphone‐based on‐site glyphosate detection.

To demonstrate the feasibility of this portable detection platform, roadside weeds, supermarket‐purchased Pokchoi cabbage, and oranges were selected as test samples. After spiking with standard concentrations of glyphosate, the test strips were analyzed using smartphone imaging (Figure 6B). Fluorescence images were processed to obtain G/B ratios, and glyphosate concentrations were calculated using the established calibration curve. Spike‐recovery experiments confirm excellent accuracy, with recovery rates of 100.62%–105.03% and RSDs of 0.48%–2.36% (Table 2).

TABLE 2.

Smartphone‐based glyphosate detection.

Samples	Spiked	G/B	Found	Recovery	RSD
	(μM)		(μM)	(%)	(%, n = 3)
Weeds	0.00	0.74	0.00	—	—
	10.00	1.10	10.26	103.23	1.55
	20.00	1.45	20.23	100.63	1.88
Pokchoi cabbage	0.00	0.75	0.28	—	—
	10.00	1.11	10.54	102.18	2.36
	20.00	1.46	20.51	100.62	1.38
Oranges	0.00	0.75	0.28	—	—
	10.00	1.12	10.83	105.03	0.86
	20.00	1.47	20.80	102.59	0.48

Together, these findings demonstrate that the NAPA‐Al³⁺ probe, coupled with test strips and smartphone‐based RGB analysis, provides a rapid, user‐friendly, and accurate platform for on‐site glyphosate detection. This approach offers great promise for real‐time environmental monitoring and food safety applications.

4. Conclusions

In this work, we have developed a fluorescent probe (NAPA‐Al³⁺) that enables rapid and sensitive detection of glyphosate via a competitive metal‐coordination mechanism. The amino and pyridine nitrogen atoms in NAPA coordinate with Al³⁺, leading to fluorescence quenching. Upon introduction of glyphosate, its strong metal‐binding groups displace Al³⁺ from the NAPA‐Al³⁺ complex through ligand exchange, effectively restoring fluorescence.

The probe exhibits excellent sensing performance, characterized by a low detection limit (57.5 nM) and a fast response time (5 s), as well as high selectivity with strong anti‐interference capability. Recovery experiments in spiked soil and food samples achieved high recovery rates (99.05%–102.71%) and low RSD values (0.45%–2.15%), confirming its reliability for practical applications.

Furthermore, the fabrication of test strips coupled with smartphone‐based RGB analysis provides a low‐cost, portable, and user‐friendly platform for on‐site glyphosate detection. This integrated strategy combines visual fluorescence changes with quantitative smartphone readout, eliminating the need for expensive instrumentation. Overall, the NAPA‐Al³⁺ system offers a promising tool for real‐time, accessible, and reliable glyphosate monitoring in environmental protection and food safety.

Author Contributions

Yali Zhu: conceptualization, investigation, data curation, formal analysis, methodology, writing – original draft. Chen Xu: conceptualization, investigation, data curation, formal analysis, writing – original draft. Fang Zeng: conceptualization, supervision, formal analysis, funding acquisition, writing – review and editing. Shuizhu Wu: conceptualization, supervision, resources, funding acquisition, writing – review and editing.

Funding

This study was supported by the National Natural Science Foundation of China, 52373209 and 22274057.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

FIGURE S1: ¹H NMR spectrum of NA‐Br in CDCl₃.

FIGURE S2: HRMS spectrum of NAPA and NAPA‐Al3+‐Cl−‐H₂O.

FIGURE S3: 1H NMR spectrum of NAPA in DMSO‐d6.

FIGURE S4: 13C NMR spectrum of NAPA in DMSO‐d6.

FIGURE S5: Photographs showing the fluorescence color changes of NAPA solution upon gradual addition of Al3+ (20 μM), excitation: hand‐held UV lamp (365 nm).

FIGURE S6: Photographs showing the fluorescence color changes of NAPA‐Al3+ solution upon gradual addition of glyphosate (40 μM), excitation: hand‐held UV lamp (365 nm).

FIGURE S7: Chemical structures of pesticides used in selectivity experiment.

FIGURE S8: Fluorescence images of NAPA‐Al3+ upon addition of various pesticides, showing (from left to right): (1) glyphosate, (2) nicosulfuron, (3) imidacloprid, (4) acetamiprid, (5) thiram, (6) mesotrione, and (7) oxyfluorfen.

FIGURE S9: Fluorescence changes of NAPA, NAPA‐Al3+, and NAPA‐Al3+ + glyphosate under different pH conditions.

FIGURE S10: Chinese cabbage, lettuce, and soil used for standard addition recovery tests.

FIGURE S11: Linear relationship between the G/B ratio and glyphosate concentration.

TABLE S1: Comparison of the probe with the literature‐reported representative fluorescent probes for glyphosate detection.

Data Availability Statement

The data that support the findings of this study are available in the Supporting Information of this article.