The fabricated GBCQDs were characterized by applying a variety of techniques including EDX, FT-IR, and TEM, etc. for descriptive elucidation. EDX and SEM pattern of the synthesized GBCQDs (Fig. 1a,1b) confirmed that C, N, O, P and S elements are present. The results also verify that the GBCQDs possess the required composition and are free from unwanted impurities.

Characterization of the as-prepared GBCQDs, where (a) EDX, (b) SEM-elemental mapping (c) FT-IR spectrum, (d) zeta potential, (e) HRTEM image, (f) HRTEM lattice fringes of GBCQDs, (g)size distribution, and (h) XRD.
FT-IR analysis was conducted to gain further insight on the surface functional groups of the GBCQDs as shown in Fig. 1c. A strong and large stretching band at 3358 cm−1 is attributed to (–OH) and (–N–H–). Furthermore, two bands at 1407 and 1055 and cm−1 are assigned to the –OH bending and hydroxyl group’s C–OH stretching. The presence of the C = N, P = O, and P–O–C is indicated by stretching and vibrational bands at 1641 and 1226 cm− 1 and 889 cm−1, respectively. The C-H, C = C, C-N, CH2, and C-S groups are represented by the bands at 2853, 1588, 1260, 866, and 677 cm−131. The surface –OH groups of GBCQDs have been found to enhance their stability, water solubility, and hydrophilicity.
A Zeta potential analyzer was used to investigate the surface properties and electron density of the GBCQDs. Due to doping with electron-rich groups (a significant number of oxygen-containing groups as well as N, S and P-rich groups), the synthesized GBCQDs had negatively charged surfaces, as evidenced by their zeta potential of -27.16 mV, which also confirmed their stability (Fig. 1d).
TEM imaging (Fig. 1e) revealed small spherical GBCQDs that were evenly distributed exhibiting a small particle size dispersion and no discernible agglomeration. Moreover, the HRTEM image revealed distinct lattice fringes with an interplanar spacing of approximately 0.20 nm (Fig. 1f), which can be assigned to the (100) plane of graphitic carbon. This observation provides direct evidence for the presence of locally ordered graphitic/sp2-hybridized carbon domains within the WSCQDs, supporting the formation of a graphitic core structure32. As shown in Fig. 1g, the majority of the particles are distributed between 1.37 and 8.71 nm, demonstrating the successful production of nanoscale particles.
The XRD pattern (Fig. 1h) of the synthesized WSCQDs displayed a broad diffraction peak centered at 2θ = 25.36°, which was assigned to the (002) plane of graphitic carbon. Compared with crystalline graphite (2θ ≈ 26.3°), the slight shift toward lower diffraction angles and the enlarged interlayer spacing (d = 3.51 Å) indicate the presence of disordered graphitic sp2 domains with abundant surface functional groups and low crystallinity, which is characteristic of biomass-derived carbon quantum dots33.
To gain further insight into the optical properties of GBCQDs, UV–VIS absorbance, excitation, and emission spectra were utilized. An absorbance peak observed at 230 nm is attributed to the π-π* transition of N = N/C = C groups. A secondary absorbance peak at approximately 300 nm is attributed to the n–π* transition of C–NH₂/C = O moieties on the GBCQDs surface, which allows emission at longer wavelengths (Fig. 2a). As shown in Fig. 2a, the GBCQDs exhibit peak fluorescence at 420 nm when excitation occurs at 340 nm, giving a Stokes shift of 80 nm that exceeds values documented in the literature. Larger Stokes shifts help prevent re-absorption and enhance fluorophore emission, providing a notable advantage for chemical sensing.

(a) UV–visible spectra along with excitation and emission of GBCQDs (water as a blank) [inset: images of GBCQDs at visible light and under 365 nm UV-light], (b) excitation-independent emission of GBCQDs at λex range of 300–400 nm, (c–g) factors affecting the fluorescence of GBCQDs at λex/em of 340/420 nm [(c) pH of 0.04 M BRB, (d) concentration of NaCl, (e) temperature, (f) photostability, and (g) storage time.
The excitation-dependent fluorescence behavior of GBCQDs was examined by scanning excitation wavelengths between 300 and 400 nm in 10 nm interval and recording the emission responses. As shown in Fig. 2b, the emission peak position of GBCQDs was mostly constant as the excitation wavelength increased up to 340 nm, but the fluorescence intensity significantly increased. Further increase in the excitation wavelength was accompanied by a red shift in the emission wavelength and a gradual decrease in fluorescence intensity. The greatest fluorescence emission was recorded at 420 nm upon excitation at 340 nm with a high quantum yield of 27.96% applying the equation in Section “S4”.
The acquired results were typical of CQDs and could be explained by the quantum confinement effect, which is attributed to the varying sizes of the particles, as well as the various surface functional groups on the GBCQDs. FT-IR , EDX and SEM analyses further confirmed the presence of diverse surface functional groups (–OH, –COOH, –NH₂, C–S, …) on the GBCQDs, which contribute to distinct surface state energies and emission traps, thereby causing a red shift in emission wavelength with changing excitation. Previous studies have indicated the incorporation of C = N and –COOH functionalities have been shown to narrow the HOMO–LUMO bandgap, resulting in a modest red shift of the emission wavelength34.
These structural and surface features may also play an important role in the sensing behavior of the GBCQDs toward FXN. The nanoscale particle size provides an accessible surface for analyte interaction, while the oxygen- and nitrogen-containing functionalities, including –OH, –COOH, and –NH₂ groups, can interact with the –COOH, –NH–, and pyridine moieties of FXN through hydrogen bonding and other non-covalent interactions. Moreover, the incorporation of N, O, P, and S heteroatoms can modulate the electronic properties and surface reactivity of the GBCQDs, thereby facilitating fluorescence quenching upon interaction with FXN. The negative zeta potential further characterizes the charged surface of the GBCQDs and may influence their interfacial interaction with FXN. Collectively, these structural features provide a structural basis for the observed quenching efficiency, selectivity, and sensitivity toward FXN.
Stability of the as-prepared GBCQDs
To assess GBCQDs stability, they were subjected to different environmental factors such as pH, temperature, ionic strength, storage duration, and light exposure.
The influence of pH was first investigated within the range of 3.0 to 12.0 using 0.04 M Britton–Robinson buffer (BRB). GBCQDs exhibited constant fluorescence intensity until pH 8.0, after which a significant decline was observed (Fig. 2c). This observed shift is likely related to nitrogen functionalities located on the surface of GBCQDs. Hydrogen ions in acidic solutions may enable the lone electron pairs of these nitrogen atoms to engage in coordination interactions. The reduced fluorescence intensity at alkaline pH may thus result from hydrogen–nitrogen interactions (such as pyrrolic, amino, and pyridinic nitrogen) combined with alterations in the structure, dispersion, and uniformity of GBCQDs35. This rationale clarifies the steep drop in fluorescence beyond pH 8.0; however, the ability to retain steady fluorescence across pH 3.0–8.0 adds to the applicability of our assay.
Furthermore, GBCQDs proved to be highly stable, even when exposed to high ionic strength environments. As presented in Fig. 2d, GBCQDs sustained nearly constant fluorescence intensity, even under high NaCl levels (1.0 M). Such features are especially advantageous for the practical application of GBCQDs in saline environments.
Subsequently, the thermal stability of GBCQDs was evaluated across different temperatures. The fluorescence intensity remained nearly constant, showing only slight variations, upon increasing temperature from 25 to 60 °C. (Fig. 2e).
Furthermore, under continuous 365 nm irradiation for 60 min, GBCQDs maintained stable fluorescence, confirming their strong photostability (Fig. 2f), indicating excellent stability over this timeframe.
Finally, GBCQDs were examined to determine their resistance to photobleaching as depicted in Fig. 2g, where it maintained consistent fluorescence intensity throughout a storage period exceeding three months.
Cytotoxicity Study
Although the developed nanoprobe is mainly intended for environmental water and food analysis rather than biomedical applications, the cytotoxicity evaluation was included to provide preliminary insight into its environmental and human safety profile. Nanomaterials released during analytical processes may eventually reach aquatic ecosystems or enter the food chain, resulting in potential indirect biological exposure. Therefore, assessing cytocompatibility of GBCQDs is crucial to ensure their safety and the possible environmental impact and biosafety of the synthesized nanoprobe, particularly considering its prospective large-scale analytical use. Accordingly, cytotoxicity was evaluated using WI-38 cell line via MTT assay as described in Section S5. As depicted in Fig. 2h, WI-38 cells viability exhibited good viability after 48h incubation with GBCQDs. The MTT assay results indicated that GBCQDs exhibit high cytocompatibility and minimal toxicity with an IC50 of 79.15 μg/mL, in comparison to 6.72 μg/mL for doxorubicin, the control. This finding implies that the GBCQDs were relatively non-toxic to WI-38 cells, making them potentially useful for cell imaging or other medicinal applications. Consequently, the as- prepared nanoprobe can be safely applied in systems containing living organisms and can be utilized as a promising cellular imaging agent.
Optimization of parameters
Several experimental parameters influencing FXN estimation such as pH, CQDs volume, and incubation time were systematically investigated and optimized to establish the best sensing conditions and maximize sensitivity. The detailed explanation of OFAT (one factor at time) strategy for optimization is provided in Section “S.7”. The best experimental conditions were using 400.0 μL of GBCQDs without using any buffer solution and carrying out the measurements instantaneously. The findings obtained are presented in Fig. 3.

Effect of (a) pH, (b) volume of GBCQDs, (c) incubation time on the fluorescence quenching of GBCQDs by FXN (d) Fluorescence emission spectra of the GBCQDs under the optimum conditions upon adding increasing concentrations of FXN (0.0–30.0 μg/mL) at λex/em of 340/420 nm [inset; linear correlation between F0-F and FXN concentration (0.0–30.0 μg/mL) under the optimized conditions.
FXN sensing mechanism
There are several possible mechanisms that may underline the quenching effect of FXN (Fig. 3d) on the fluorescence of GBCQDs including static, dynamic, and inner filter effect (IFE). Static quenching refers to a nonradiative process of energy loss that occurs via the creation of a non-fluorescent ground state complex between the fluorophore and quencher. Dynamic or collisional quenching is another radiation-free pathway of energy loss in which an excited fluorophore returns to its ground state as a result of interactions or collisions with a quencher. In dynamic quenching, raising the temperature enhances molecular collisions, resulting in an increased Stern–Volmer constant (Ksv). In contrast, for static quenching, higher temperatures cause the dissociation of the ground-state complex, thereby decreasing the Ksv value37,38.
The inner filter effect (IFE) arises when the fluorophore’s excitation spectrum and the absorption spectra of the quencher overlap. It is unaffected by temperature and does not share the characteristics of either static or dynamic quenching38,39.
To determine the mechanism behind FXN-induced quenching, the excitation and emission spectra of GBCQDs were recorded along with the absorption spectrum of FXN, and all were superimposed on a single graph to evaluate the extent of overlap. It was observed that the UV–Vis absorption spectrum of FXN overlaps with the excitation spectrum of the GBCQDs, confirming that IFE significantly contributes to the quenching mechanism (Fig. 4a). To further assess the extent of IFE contribution, the corrected fluorescence was calculated as follows:
$${F}_{Corrected}= {F}_{Observed}\times Antilog (\frac{{A}_{ex}+{A}_{em}}{2})$$
(1)
Where, Aex and Aem correspond to the absorbance of FXN at the excitation and emission wavelengths of the GBCQDs, respectively, while, Fobserved denotes the measured fluorescence intensity, while Fcorrected refers to the fluorescence intensity after deducting the IFE. The percentage suppressed efficiency (% E) was then determined using the following formula, based on both the corrected and observed FI:
$$\% E=\left[1-(\frac{F}{{F}^{0}})\right] \times 100$$
(2)
As illustrated in Fig. 4b, the correlation between %E and FXN concentration (μg/mL) confirms that the fluorescence quenching mechanism is significantly influenced by IFE. This was demonstrated by the significant 69.35% decrease in suppression efficiency that happens when IFE is excluded. However, 30.65% of the fluorescence quenching can still be attributed to an alternative mechanism. To further elucidate the possible quenching pathway, the Stern–Volmer equation was applied as follows:
$${F}^{0}/F=1+ {K}_{sv}\left[FXN\right]$$
(3)
Where, the molar concentration of FXN is denoted by [FXN], the fluorescence intensities in the presence and absence of FXN are denoted by F0 and F, and the Stern–Volmer quenching constant, Ksv, is determined by calculating the slope of (F0/F) vs [FXN]. Therefore, (F0/F) versus FXN concentration at four different temperatures (303, 313, 323, and 333 K) were plotted to obtain Stern–Volmer plots. The findings in Table S2 and Fig. 4c show that Ksv values rise with temperature, pointing to a dynamic quenching mechanism. As the temperature rises, the molecule collision is intensified, which results in increased dynamic quenching. Furthermore, the absorption spectra of GBCQDs were recorded both in the absence and presence of FXN and no new peaks were observed which further confirms dynamic quenching (Fig. 4d). Based on these results, it was concluded that FXN’s quenching effect of the GBCQDs fluorescence is attributed to both IFE and dynamic quenching.

(a) A co-plot of the UV–visible spectrum of GBCQDs and FXN in water and the excitation and emission spectra of GBCQDs, (b) suppressed efficiency of the observed and corrected fluorescence measurements of GBCQDs after adding different concentrations of FXN in the range of 0.05–30.0 μg/mL, and (c) Stern–Volmer plots of GBCQDs-FXN at different temperatures (303, 313, 323, and 333 K) over a FXN concentration range of 0.05–30.0 μg/mL, and (d) absorption spectra of FXN, GBCQDs before and after addition of FXN.
Validating the evolved approach
The developed method was validated in compliance with the criteria outlined by ICH40 to evaluate its accuracy and verify its reliability and precision for the assay of FXN. Various validation criteria were examined including linearity, range, accuracy, precision, detection limit (LOD), quantitation limit (LOQ) and selectivity.
Linearity and range
To generate the calibration plot, (F0-F) was plotted against FXN concentration in (μg/mL) (Fig. 3d). The method’s linear dynamic range was identified as the concentration interval across which the method achieves reliable measurements. The approach showed excellent linear correlation (R2 > 0.999) under optimal conditions over a wide concentration range (0.05 – 30 μg/mL).
LOD and LOQ
The standard deviation of the intercept was used to calculate LOQ and LOD according to ICH guidelines applying the following equations:
$${\text{LOD }} = { 3}.{3}Sa/b$$
(4)
$${\text{LOQ }} = { 1}0Sa/b$$
(5)
Where, Sa is the standard deviation of the regression line’s intercept and b is its slope. The low values of LOD and LOQ confirmed the high sensitivity of our approach for the studied drug revealing that the developed approach could be applied for quantifying FXN in different complex matrices such as food and environmental water samples. The analytical performance of the proposed method was shown in Table 1.
Accuracy and precision
Recovery studies were carried out to evaluate the method’s accuracy. The developed approach was employed for determining FXN in a variety of real samples including pharmaceuticals, food and environmental water samples. High recovery percentages and low relative error percentages confirmed the approach’s accuracy.
Conversely, both inter- and intra-day precision were assessed. To examine inter-day precision, different concentrations of FXN were repeatedly analyzed multiple times on the same day, while the measurements were conducted on three consecutive days for evaluating the inter-day precision. The evolved approach’s precision is demonstrated by its low %RSD values (< 2.0%) as presented Table S3.
Selectivity of the developed nanoprobe
Selectivity is one of the key characteristics of a sensor, referring to its ability to respond specifically to the target analyte in the presence of other potentially interfering ions or molecules. The selectivity of the developed platform was evaluated by examining the influence of potential interfering excipients to determine whether they had a significant impact on the assay of FXN. Different potentially interfering substances were examined including (glucose, sucrose, citric acid, lactose, ascorbic acid, starch, acetate, thiosulfate, sulfite, (Fe2+, Co2+, Cu2+, Ni2+, Na+, K+, Ca2+, Mg2+, Ba2+, Pb2+, Cd2+, NH4+, Zn2+, Cl-, SO42- and CO32-) were investigated. Additionally, the effect of other veterinary drugs which may be either co-administered or co-formulated with FXN including levamisole, tylvalosin, clanobutin, imidocarb dipropionate, toltrazuril, diclazuril, phenylbutazone, rafoxanide, tolfenamic acid, oxytetracycline, and florfenicol were investigated. These potentially interfering substances were added to the GBCQDs system and the percentage change in FI was calculated using the following equation:
$${\%} F={(F}_{2}-{F}_{1})/{F}_{1} \times 100$$
(6)
The FI of GBCQDs solutions with and without an interfering component are denoted by F1 and F2, respectively. Results demonstrated in Table S4 showed that interfering substances had a minimal effect on the probe’s fluorescence, which did not affect its selectivity. Therefore, the probe can be effectively utilized for the analysis of FXN in various matrices without notable interference from interfering substances such as its co-administered or co-formulated drugs. The selectivity of the fabricated platform can be rationalized by the contribution of the inner filter effect (IFE), as significant spectral overlap with the excitation wavelength is not expected for all tested drugs. Moreover, as all measurements were carried out near 400 nm, the likelihood of interference from other drugs or foreign substances was minimal. Nevertheless, the observed selectivity may also involve differences in surface interactions and chemical structures among the tested species, rather than being exclusively governed by the IFE.
Assay of FXN in various real samples
The feasibility, practicability, reliability and applicability of our suggested approach was assessed by its application for the quantification of FXN in various real samples such as pharmaceutical preparations, food, and environmental food samples. These different samples represent several matrices with likely variations in components and complexity, so, sample preparation strategies, possible interferences, and the accuracy and precision of the approach should all be critically examined.
Assay of FXN in dosage forms
The developed sensing platform was applied for the quantification of FXN in Maroflam®, Megluflor® and Floxon® veterinary injections. As depicted in (Table 2), the method achieved high recovery rates exceeding 97% with low %RSD values, demonstrating its reliability and efficiency in determining the studied analyte in veterinary dosage forms in presence of its co-formulated drugs and excipients.
Assay of FXN in food samples
Literature surveys indicate that FXN has been employed as a NSAID in veterinary medicine. European Commission Regulation No. 2377/90 establishes the maximum residue limits (MRL) of 20, 300, 40 μg/Kg for FXN in muscle, liver and milk, respectively27. FXN residues in food products are quite serious since they might trigger allergic reactions in susceptible people or contribute to adverse side effects ranging from mild gastrointestinal irritation to aplastic anemia26,41. Hence, to guarantee food safety, monitoring the levels of FXN in milk, egg, and other edible tissues from treated livestock or chickens is crucial. The obtained results (Table 2) demonstrated that the fabricated sensing platform was effective in detecting FXN in various food samples including milk, meat, liver, egg and chicken muscles samples. Recovery testes of spiked FXN showed high percentages of recovery exceeding 97%, demonstrating its applicability for FXN assay in food samples.
Assay of FXN in environmental water samples
Animal feces and urine in pastures can directly introduce FXN into the environment. After being released, these contaminants can spread and disperse into water bodies. Therefore, FXN detection in environmental water samples is crucial. The applicability of the sensing platform for detecting FXN in river water was examined. The spiking recovery approach was used to determine the added FXN in river water samples. As shown in Table 2 the percentage of recoveries exceeds 97% with low %RSD, confirming the effectiveness of the developed method for analyzing FXN in water samples.
Additionally, the obtained results were statistically compared to that obtained by a reported spectrophotometric method42 using the variance ratio F-test and the student t-test. The F-test was used to compare the variances and assess precision, while the student t-test was used to compare the average percentage found. No significant difference was found between the performance of the developed approach and the comparison method (Table 2). The findings clearly demonstrate that the probe is effective for determining FXN in food products and yields results comparable to the reference method42.
