Glutamic acid–capped iron oxide quantum dots as fluorescent nanoprobe for tetracycline in urine

Abstract

The fabrication of iron oxide quantum dots (IO-QDs) modified with glutamic acid (Glu) under controllable conditions is reported. The IO-QDs have been characterized by transmission electron microscopy, spectrofluorometry, powder X-ray diffraction, vibrating sample magnetometry, UV–Vis spectroscopy, X-ray photoelectron spectroscopy, and Fourier-transform infrared spectroscopy. The IO-QDs exhibited good stability towards irradiation, temperature elevations, and ionic strength, and the quantum yield (QY) of IO-QDs was calculated to be 11.91 ± 0.09%. The IO-QDs were furtherly measured at an excitation wavelength of 330 nm with emission maxima at 402 nm, which were employed to detect tetracycline (TCy) antibiotics, including tetracycline (TCy), chlortetracycline (CTCy), demeclocycline (DmCy), and oxytetracycline (OTCy) in biological samples. The results indicated that TCy, CTCy, DmCy, and OTCy in urine samples show a dynamic working range between 0.01 and 80.0 μM; 0.01 and 1.0 μM; 0.01 and 10 μM; and 0.04 and 1.0 μM, respectively, with detection limits of 7.69 nM, 120.23 nM, 18.20 nM, and 67.74 nM, respectively. The detection was not interfered with by the auto-fluorescence from the matrices. In addition, the obtained recovery in real urine samples suggested that the developed method could be used in practical applications. Therefore, the current study has prospect to develop an easy, fast, eco-friendly, and efficient new sensing method for detecting tetracycline antibiotics in biological samples.

Graphical abstract

Supplementary Information

The online version contains supplementary material available at 10.1007/s00604-023-05801-3.

Introduction

The use of antibiotics has increased during COVID-19 pandemics [1] due to antibacterial and anti-inflammatory properties. Tetracyclines (TCys) can suppress bacterial growth by preventing protein synthesis [2]. About 30% of a dose of 250 mg of TCy is excreted via urine, and the concentration of TCy can reach 2 mg L⁻¹ in 2 to 4 h [3]. Due to the widespread use, the residues of TCys have been found to have different harmful effects, such as genetic changes, mutations, and other changes in human microbiota, which lead to resistance in bacteria [4]. Thus, developing a quick and cost-effective detection method for TCys in biological samples is crucial.

A variety of analytical methods have been developed for the determination of TCys, including high-performance liquid chromatography [5], ELISA combined with immunochromatographic assay [6], SERS [7], microbial assay [8], colorimetric [9], and amperometric methods [10]. Chromatographic determination methods benefit from reliable results, good repeatability, and sensitivity. However, the chromatographic technique suffered from time-consuming and required extended sample processing stages. ELISA can overcome the drawback of costly chromatography experiments though many factors could still give a false positive result. The other reported methods are limited by the complicated treatments, a relatively long detection time, and the possible interferences that may be considered in practical applications.

Different designs of fluorescence sensors have attracted substantial attention in recent years due to their ability to provide sensitive and rapid ways at a lower cost. Among the reported methods, the fluorescence-based detection method [17, 18] is preferred for detecting TCys with high selectivity. Nanomaterials-based fluorescent probes have been studied for TCys detection, such as water-soluble QDs [11], metal nanoclusters [12], and silver nanoparticles [13]. These sensors usually detect the target analytes by observing the variation of fluorescence response, and the accuracy of these methods is affected by environmental and systematic errors [14]. The main problem of fluorescence sensors to detect TCys in real samples is the emergence of auto-fluorescence from samples themselves as interference [15]. In addition, the self-absorption of fluorescent nanomaterials also reduces the photostability and quantum yields of the prepared fluorescent nanomaterials. Self-absorption and auto-fluorescence interference are widely found in biological systems, making fluorescence probes somewhat unfavorable for biological samples [16].

Herein, an iron oxide quantum dots (IO-QDs) based method to probe tetracycline antibiotics in biological samples without background auto-fluorescence is reported. Our method demonstrated a higher fluorescence signal of the IO-QDs (quantum yield (QY) = 11.19 ± 0.01%) than the reported results [17, 18]. Scheme 1 shows the schematic illustration of this study. The hydrothermal treatment produced the fluorescent IO-QDs using iron oxide nanoparticles (IO-NPs) as the precursor.

Scheme 1.

Schematic illustration of this study

The addition of glutamic acid as a dopant in the hydrothermal treatment made the IO-QDs’ surfaces abundant with carboxylic groups and amino groups, which provide specific interactions between IO-QDs and TCy antibiotics. With the addition of TCy antibiotics, the fluorescence of IO-QDs would be quenched due to the static quenching. Furthermore, the IO-QDs were successfully applied to detect tetracycline in urine samples. The method offers convenient, low-cost, stable, good accuracy and precision.

Experimental

Reagents and instruments

FeCl₃ was obtained from High Standard Enterprise. Quinine, glutamic acid, ascorbic acid, and oxytetracycline hydrochloride were purchased from Alfa Aesar. Trimethoprim, sulfamethoxazole, demeclocycline, and hippuric acid were obtained from Sigma-Aldrich. Chlortetracycline was purchased from Thermo Scientific. Tetracycline and chloramphenicol were purchased from EMD Millipore. Ampicillin, erythromycin, and kanamycin sulfate were purchased from Bio Basic. Vancomycin was purchased from Calbiochem. CaCl₂, ZnCl₂, KCl, NaCl, and NH₄Cl were purchased from Showa Chemicals. Urea was obtained from Shimakyu’s pure chemicals. All the reagents were of reagent grade and utilized without further purification. A Milli-Q water system (Simplicity, Millipore) was used to prepare all the aqueous solutions.

The fluorescence spectra were recorded using a Varian Cary eclipse fluorescence spectrophotometer. The absorption spectra were measured using a Spectra Academy SV-2100 UV–Visible spectrometer. The pH of the solutions was measured using a Hotec PH-10C pH meter. FTIR spectroscopy (ALPHA FTIR spectrometer, Bruker) was performed to identify the functional groups. A Nanoplus HD-zeta/Nano Particle Analyzer was run to obtain the zeta potential information. A Hitachi HT-7700 transmission electron microscope (TEM) with an accelerating voltage of 100 kV was used to explore the morphologies of the IO-QDs. A Bruker 2^nd Gen XRD 000,705 was employed to obtain the crystalline structure data of IO-NPs and IO-QDs. A SQUID0000600 MPMS-XL7 was implemented for measuring the magnetic properties of the IO-NPs and IO-QDs. A JEOL, JAMP-9500F Auger electron spectroscopy was used to confirm the composition of IO-QDs.

Preparation of IO-NPS and IO-QDs

To prepare IO-QDs, 1.0 g of glutamic acid was mixed with 20 mL of Fe₃O₄ NPs solutions (1 mg/mL) and poured into the hydrothermal reactor tube where Fe₃O₄ NPs precursors were synthesized following the established work [18]. The details are given in the electronic supporting material (ESM) supplementary information. The sealed hydrothermal reactor was heated at 200 °C for 14 h in the heating oven. Subsequently, the reactor was cooled to the ambient temperature and filtered with a Whatman filter paper. The filtered IO-QDs were stored at 4 °C before further application.

Fluorescence stability of IO-QDs

Different concentrations of NaCl aqueous solutions were added into IO-QDs solutions to observe the ionic strength effect on IO-QDs. pH effects of IO-QDs were carried out by adjusting pH between 2 and 11 using 0.1 M of HCl and NaOH aqueous solutions. The exposure of IO-QDs determined the photostability of IO-QDs under visible light (LED) and UV light (λ_{265 nm} and λ_{365 nm}) for 30 min. Thermal stability of IO-QDs was conducted by heating the IO-QDs solutions from 26 to 100 °C. The long-term stability of IO-QDs was monitored by storing them at 4 °C and room temperature for 30 days.

Quantum yield measurement

To evaluate the fluorescence quantum yield of the IO-QDs, quinine (quantum yield, QY = 54%, η = 1.33) as a reference was diluted by using 0.1 M of sulfuric acid while IO-QDs were diluted by ultrapure water (η = 1.33). Fluorescence spectra were recorded with 330 nm of excitation wavelength, and an absorbance value was adjusted to less than 0.1 at 330 nm. To calculate the fluorescence quantum yield of the IO-QDs, the following formula is engaged [19]:

$$Q{Y}_S=Q{Y}_R\left(A_R/A_S\right)\left(I_S/I_R\right)\left({\eta }_S^2/{\eta }_R^2\right)$$

QY stands for quantum yield, A for absorbance, I for integrated fluorescence intensity, and η for solvent refractive index. The S and R subscripts denote IO-QDs and quinine solution, respectively.

IO-QDs-based fluorescence sensor for TCy

To demonstrate the applicability of the fabricated fluorescent sensor for the detection of antibiotics, different antibiotics were mixed separately with the IO-QDs solutions, where the final concentration of each antibiotic is 200 μM. The fluorescence spectra (λ_ex = 330 nm, λ_em = 402 nm) were monitored to record the specific antibiotics that give the quenching on the luminescence response. To assess the sensitivity trials of the IO-QDs sensor, fluorescence responses of IO-QDs in the presence of different concentrations of TCy were recorded and compared.

Selectivity trial

The selectivity of the fabricated IO-QDs-based fluorescence sensor was studied by examining the fluorescence response of the potential interference species, including hippuric acid, ammonium, Ca²⁺, K⁺, Zn²⁺, and Na⁺. To conduct these experiments, the solution of IO-QDs was mixed with 200 μM of TCy in the presence of each possible interference.

Real sample measurements

Fresh urine samples collected from a healthy volunteer who did not consume any single dose of TCy or antibiotic were stored at 4 °C, which were examined to evaluate the feasibility of IO-QDs in practical application and discarded after 24 h. The raw urine sample was filtered through 0.45-μm PES membrane filters and used in the examination without further pretreatment. The required amount of the IO-QDs was added to 0.25 M citrate buffer at a pH of 2. A TCy stock solution was spiked to the undiluted urine solutions at different concentrations, samples were analyzed in triplicate, and the fluorescence spectra of the spiked IO-QDs were recorded upon the excitation wavelength of 330 nm.

Results and discussion

Characterization of the prepared IO-QDs

The morphology and particle size distribution of IO-QDs were confirmed using TEM (Fig. 1a). As can be observed from the images, the IO-QDs have an average size of 7.03 nm with mostly spherical-shaped. The composition of the prepared IO-QDs was evaluated using X-ray photoelectron spectroscopy (XPS). The high-resolution spectra of Fe 2p, O 1 s, C 1 s, and N 1 s were observed (Fig. 1b). The energy band and zeta potential of the QDs were determined to be 4.6 eV and + 1.54 mV, respectively (Fig. S1a–b). It is deducted that the amine group controls a positive surface charge of the IO-QDs. Figure 1c–d further confirmed that the magnetic property of iron oxide nanoparticles vanished as transforming to IO-QDs since the magnetization saturation (Ms) reduced rapidly, which can be rationalized by the oxidation of the IO-QDs surface [20]. Powder X-ray diffraction (XRD) patterns (Fig. 1e) were applied to determine the crystalline structure of the nanoparticles and quantum dots of iron oxide, respectively. The observed relatively strong reflection peaks at 2 theta of 30.1°, 35.5°, 43,1°, 53,4°, 62.6°, and 74.3° were ascribed to the (220), (311), (400), (422), (511), (440), and (553) planes of Fe₃O₄ which match well with the database of magnetite on The Joint Committee on power diffraction standard (JCPDS card: 19–629) file. These findings indicated that the iron oxide nanoparticles have Fe₃O₄ crystalline cubic spinel structure. After transforming into quantum dots, the IO-QDs consist of Fe₃O₄ and Fe₂O₃ in their crystalline structure [21]. Notably, carbon dots were not produced during the preparation of IO-QDs since the single diffraction peak of conventional carbon dots was not found (19.1°, JCPDS card no. 26–1076). Moreover, as can be found in Fig. 1f, the IR absorption spectra of the prepared IO-QDs have found several absorption bands. Further detailed elaboration about IR spectra was available in electronic supporting material (ESM).

Fig. 1.

a TEM image of IO-QDs synthesized in this study. Inset picture: particle size distribution histogram of IO-QDs. b XPS spectra of IO-QDs. Magnetic hysteresis loop of c Fe₃O₄ NPs and d IO-QDs. e Powder X-ray diffraction (P-XRD) patterns of Fe₃O₄ NPs and IO-QDs. f FTIR spectra of different IO-QD systems

Optical properties of IO-QDs

The fluorescence and UV–Vis spectra of IO-QDs were examined to evaluate their optical properties. The IO-QDs displayed intense and intrinsic fluorescence in the aqueous solution. The fluorescence spectra showed the typical excitation-dependent emission wavelength properties of IO-QDs, as shown in Fig. 2a. The radiative recombination induces excitation wavelength-dependent fluorescence through a series of emissive traps [22]. The fluorescence intensity steeply increased as the excitation wavelength increased from 300 to 330 nm and achieved a maximum fluorescence at 330 nm of excitation wavelength. Subsequently, the fluorescence emission intensity decreased gradually when changing the excitation wavelength from 330 to 400 nm. The 3D excitation-emission contour plot, as depicted in Fig. 2b, shows excitation-dependent emission, consistent with Fig. 2a. Figure 2c shows the normalized fluorescence spectra of IO-QDs, which present the red-shift properties of fluorescence peaks with the increasing excitation wavelength. The surface defects created via the oxidation process give rise to lower energy levels of electrons, then lead to a red shift [23].

Fig. 2.

a Fluorescence of IO-QDs. b 3D excitation-emission contour plot of IO-QDs. c normalized FL spectra at λ_ex = 300–400 nm. d Fluorescence excitation and emission of IO-QDs. e UV–Vis spectrum of IO-QDs

The IO-QDs had the most intense fluorescence with an excitation wavelength of 330 nm and emitted at 402 nm, as shown in Fig. 2d. The UV/Vis spectrum of IO-QDs disclosed a wide absorbance between 200 and 400 nm. As shown in Fig. 2e, IO-QDs had a distinct UV absorption peak at approximately 312 nm, which might be ascribed to the n-π* transition of the C = O bond [22]. The other absorption peak located at about 270 nm may result from π-π* transition of the C–C bond [24].

The fluorescence quantum yield (QY) of IO-QDs was determined using the equation mentioned in the previous section [25]. The calculated QY of the IO-QDs is ca. 11.91 ± 0.09%. The as-prepared IO-QDs solution under optimal conditions, which the details are given in electronic supporting material (ESM) supplementary information, was observed to have a solid turquoise color irradiated with UV light of 365 nm and a colorless clear solution under daylight. The hydrothermal treatment produced the IO-QDs to determine TCy through fluorescence quenching, as displayed in Scheme 1.

To scale up the preparation IO-QDs with better efficiency, the effect of the total volume in the hydrothermal reactor was evaluated where the amount of added glutamic acid was fixed. The intense FL signal of QDs (Fig. 3) corresponded to the total volume of 10 mL and 20 mL. It is believed that the pressure inside the hydrothermal reactor may take responsibility for the IO-QDs’ fluorescence features [26]. In order to gain the best fluorescence emission signal, we have also investigated the influence of Fe₃O₄ NP volume by fixing the glutamic acid weight of 1.0 g. It also can be observed in Fig. 3 that 10 mL of Fe₃O₄ NP volume exhibited the best fluorescence signal. It is noted that the suspension of IO-QDs was only found when the total volume of IO-NPs used for preparing IO-QDs was 10 mL.

Fig. 3.

a Fluorescence spectra of IO-QDs synthesized under different glutamic acid weigh to Fe₃O₄ NP volume ratio (T: 200 °C, t: 14 h, Fe₃O₄ NPs concentration: 1.0 mg/mL). b The suspended IO-QDs and the IO-QDs appear under daylight and UV light, respectively

It is also found that such a suspended form has an even higher fluorescence intensity. It is noted that the QY of the suspended form is 13.80 ± 0.28% which existed the highest QY then other IO-QDs reported ever [17, 18]. The suspended form will be used for different application which is under study.

Stability of the prepared IO-QDs

The influence of ionic strength on the fluorescence features of the IO-QDs was investigated. As shown in Fig. S4a, the IO-QDs showed a gradual decrease in fluorescence signal, and no aggregation was observed upon salt addition. Therefore, the IO-QDs were electrostatically stable under high ion concentrations. Meanwhile, the pH effect on the IO-QDs fluorescence emission was also evaluated. As shown in Fig. S4b–c, the fluorescence signal of the IO-QDs was amplified at pH = 3. The carboxylic groups on the surface of IO-QDs may cause the amplified fluorescence emission signal. It is confirmed that the protonated state of the functional groups on the surface IO-QDs depends on the pH solution during functionalization [27] and may remarkably influence the radiative combination which allows the increase of the fluorescence intensity [28]. Furthermore, the additional OH⁻ in the solution during pH adjustment may cause oxidation of the IO-QD surface [29]. The fluorescence intensity decreased at pH 4 to 11 because the positive charge of IO-QDs has lost and the repulsion process between them also decreased, leading to IO-QDs aggregation. These results confirmed the pH-dependent aggregation and fluorescence behaviors of IO-QDs, as presented in Fig. S4c. The fluorescence emission intensity of the IO-QDs practicality remained high (pH 3) and was selected as the controlled pH in further experiments.

Furthermore, the photostability of IO-QDs was evaluated by irradiating with white light (LED) and UV light (λ_{265 nm} and λ_{365 nm}) for 30 min, respectively. As shown in Fig. S4d, when IO-QDs were illuminated with white light, the fluorescence intensity showed only a limited variation. In contrast, the IO-QDs exposed to UV light showed a significant fluorescence decrease in 5 min, but the fluorescence was restored after 10 min of UV irradiation exposure. The stability of IO-QDs under UV light exposure may be due to carboxylic acid on IO-QDs blocking the reactive sites and sterically decreasing chromophore reactivity by introducing bulky groups [30]. Thus, the QDs exhibited good stability and anti-photobleaching properties.

As for the thermal stability of IO-QDs, Fig. S5a shows that the fluorescence intensities of IO-QDs gradually decreased as the temperature increased, which may be caused by the destruction of IO-QDs structures and lead to the dissociation of passivant on the IO-QD surfaces [31]. Interestingly, the IO-QDs prepared in this study could return to the initial fluorescence intensity after cooling to the ambient temperature (25 °C). This behavior is thermally reversible fluorescence, suggesting the surface of IO-QDs would not be oxidized due to covalent bond protection from oxidation [32]. Furthermore, by adjusting the temperature from 25 to 100 °C, the reversible fluorescence quenching and recovery were observed, confirming the thermally reversible behavior through four cycles, as shown in Fig. S5b. The IO-QDs remained stable as the temperature altered through several cycles, indicating the thermostability of IO-QDs.

As for the long-term stability of the IO-QDs, it was found that the fluorescence intensities of IO-QDs decreased gradually as the storage time increased under both mild and ambient temperatures (Fig. S5c). To avoid the signal deviation caused by the aging of IO-QDs, the IO-QDs were used as soon as they were prepared.

The feasibility of FRET and static quenching in probing TCy

This study developed a fluorescence sensing strategy for determining the TCy based on the Fӧrster resonance energy transfer (FRET) between QDs and TCy. FRET results from long-range dipole–dipole interaction between the energy donor and acceptor. In the sensing system established in this study, the IO-QDs were energy donors, whereas TCy molecules were energy acceptors. FRET can cause fluorescence quenching without a covalent bond between the absorbent and fluorophore. To affirm the FRET mechanism of the established sensing system, we investigated QDs-TCy behavior through different approaches. The absorption of TCy partially overlapped with the emission of the IO-QDs excited by the wavelength of 330 nm (Fig. S6), which was one of the required characteristics for the FRET. Due to the overlap between the emission donor and excitation acceptor, the distance between them is short enough [33]. The FRET as a spectroscopic ruler occurs at the Fӧrster distance of 1–10 nm [34]. Fӧrster distance calculation was given by following the formula [34]:

$$R_0=0.02108{({\text{k}}^2{\mathrm{\Phi }}_{\text{D}}{\text{n}}^{-4}J)}^{1/6}\text{nm}$$

where R₀ is the Fӧrster distance (in nm), k² denotes dipole orientation factor (2/3), Φ_D denotes fluorescence quantum yield of IO-QDs, η denotes the refractive index of the solvent, and J denotes the overlap integral. The calculated R₀ between IO-QDs and TCy is ca. 1.71 nm (< 2 nm), indicating that the static quenching occurs as reported results [35]. The IO-QDs-TCy distance was calculated to be 2.71 nm engaging the formula [34]:

$$E=\frac{1}{[1+(\frac{r}{R_0})6]}$$

where E is energy transfer efficiency which can be obtained from $\mathrm{E}=1-\frac{{}^{\tau }\mathrm{DA}}{{}^{\tau }D}$, and ${}^{\tau }\mathit{DA}$ and ${}^{\tau }D$ are the fluorescence lifetime of IO-QDs with TCy and without TCy, respectively.

Nonradiative energy transfer between QDs and antibiotics induces luminescence changes as well as new binding events in the antibiotic, which result in a drop in the fluorescence intensity of the IO-QDs that can be used as a sensing device. The lifetime of IO-QDs and IO-QDs/TCy were 4.38 ns and 4.12 ns, respectively (Fig. 4a), As can be seen in the figure, there is no remarkable difference between the fluorescence lifetime of IO-QDs and those in the presence of tetracycline. This finding suggests that FRET was not responsible for the fluorescence quenching in this sensing system. The quenching efficiency between IO-QDs and TCy can be calculated by the Stern–Volmer (S-V) equation: (I₀/I) = 1 + K_SV [Q], where I₀ and I is the emission intensity of IO-QDs before and after the addition of TCy, respectively. K_SV is Stern–Volmer quenching constant, and [Q] is the concentration of TCy. The K_SV is calculated to be 5.6 × 10³ M⁻¹. The bimolecular quenching constant (K_q) is calculated to be 1.28 × 10¹² M⁻¹ s⁻¹ which is higher than the maximum collision quenching constant 2 × 10¹⁰ M⁻¹ using the equation (K_SV = K_qτ₀), suggests that the static quenching is presented [36]. The binding constant and binding site was determined from the modified Stern–Volmer plot, as displayed in Fig. 4b, by using the formula [36]: log (F₀ − F)/F = log K_b + n log [Q], where F₀ and F are the steady-state fluorescence intensities in the absence and presence of quencher, respectively. K_b is the binding constant, n is the number of binding sites, and Q is the concentration of the quencher. K_b and n were calculated to be 1.66 × 10³ M⁻¹ and 0.87, respectively, indicating the strong binding between IO-QDs and TCy.

Fig. 4.

a Lifetime fluorescence of IO-QDs and IO-QDs/TCy system. b modified Stern–Volmer plot of IO-QDs with TCy

In addition, the absorption peak of IO-QDs was shifted from 345 to 357 nm (Fig. S2), which implied the presence of static quenching. In the static quenching, the fluorophores form non-fluorescent complexes with analytes in the ground state [37]. The fluorescence of IO-QDs was turned off upon adding TCy because of the electrostatic interaction between the cationic (NH₃⁺) parts of TCy and negatively charged parts (e.g. -COO⁻) on the surface of IO-QDs, leading to a substantial quenching of the fluorescence of IO-QDs.

Quantitative aspects and the interference study

The variation in fluorescence intensity in Fig. 5a indicated that the IO-QDs prepared in this study had shown a specific quenching effect towards TCy. An obvious red-shifted of IO-QDs from 402 to 408 nm was only observed in the presence of TCy, while other antibiotics showed no shifts. Figure 5b displays the fluorescence of the IO-QDs in the presence of different TCy concentrations at 402 nm under the optimum conditions. Upon the addition of TCy in the range concentration of 0 to 200 μM, it is noted that the emission of IO-QDs red-shifted from 402 to 409 nm (Fig. 5a–b). As discussed in the previous sections, the interaction between IO-QDs and TCy leads to a drop in the fluorescence signal due to the nonradiative recombination of the excitons. The calibration curve of TCy is plotted in Fig. 5c. A highly linear relationship between the fluorescence response (F₀ − (F/F₀)) of IO-QDs and the concentrations of TCy was observed in the range of 0.1–80 μM. The detection limit of 7.69 nM was calculated based on a signal-to-noise ratio of 3. The fluorescence spectra of IO-QDs in the presence of chlortetracycline (CTCy) and a highly linear relationship between the fluorescence response (F₀ − (F/F₀)) of IO-QDs and the concentrations of CTCy was observed in the range of 0.01–1.0 μM (Fig. S8a–b). Meanwhile, Fig. S8 c–d shows the fluorescence spectra of IO-QDs upon the addition of oxytetracycline (OTCy), and a linear relationship was observed in the range of 0.04–1.0 μM. The fluorescence spectra of IO-QDs in the presence of demeclocycline (DmCy) and a fitting curve in the range of 0.01–10.0 μM can be seen in Fig. S8e–f. The detection limit of CTCy, DmCy, and OTCy was calculated to be 120.23 nM, 18.20 nM, and 67.74 nM, respectively. To further evaluate the feasibility of utilizing IO-QDs for detecting TCy in the biological samples, we assessed the selectivity of QDs towards potentially competing compounds presented in urine samples. In typical urine compositions, hippuric acid and NH4 + can act as sensitizers [15] upon interacting with IO-QDs. Different cations and antibiotics, including other TCy antibiotics (CTCy, DmCy, and OTCy), were also checked as possible interferences. Gratifyingly, the fluorescence of IO-QDs can only be quenched by TCy antibiotics but not other compounds (Fig. 5d). To distinguish TCy, CTCy, DmCy, and OTCy, IO-QDs absorbance in the presence TCys was measured. The spectral difference between IO-QDs in the presence of TCy, CTCy, DmCy, and OTCy was observed in Figs. S8e and S9. Absorption spectra of TCy, CTCy, DmCy, and OTCy can be differentiated at the peak 357 nm, 372 nm, 367 nm, and 352 nm, respectively. Meanwhile, emission spectra of TCys show different quenching at the peak 406 nm (TCy), 414 nm (CTCy), 421 nm (DmCy), and 408 nm (OTCy).

Fig. 5.

a Fluorescence emission spectra of the IO-QD solutions in the presence of various antibiotics at 200 μM. b fluorescence emission spectra of the IO-QDs upon adding TCy (0–200 μM). c The relation between the fluorescence intensity of IO-QDs ((F₀ − F)/F₀) and the TCy concentration range of 0–200 μM. Inset shows their linear relationship in the range of 0.1–80 μM. d the relative fluorescence intensity (F/F₀) of IO-QDs in the presence of TCy antibiotics and other interfering agents where the concentration of all examined species was 200 μM

Comparison of detection probe for TCy

Table 1 presents a comparison of TCy detection results with the reported works. The comparison was further constructed between some reported methods for determining tetracyclines. Amperometric magneto-immunosensor have been synthesized to detect tetracycline in milk with a reaction time of 30 min and LOD of 20 nM. Shen et al. (2014) synthesized colorimetric-based Au NPs to quantify tetracycline in urine samples with a 10-min reaction time, and the LOD was 1.1 μM. The colorimetric method of Au nanocluster was used to determine tetracycline in drug and milk with a shorter reaction time (5 min) and lower LOD (0.5 μM). To improve the selectivity of the developed method, graphene QDs and Eu³⁺ and integrated nano-clay and CDs can tackle the method’s limitation, resulting in lower LOD and short-time reaction. However, these methods showed complicated in preparation. Furthermore, compared with the mentioned methods, glutamic acid–capped IO-QDs offer a wider linear range, a lower LOD (7.7 nM), higher sensitivity, short reaction time, and better selectivity.

Table 1.

Comparison of the proposed method with other reported methods for TCy detection in different samples

Method	Sample matrix	Detection time	Linear range	LOD	Ref
Colorimetric using AuNPs	Urine	10 min	10–330 μM	1.12 μM	[38]
Colorimetric aptasensor based on Au nanocluster	Drugs and milk	5 min	1–16 μM	0.50 μM	[9]
Graphene QDs and Eu3+	River water and milk	1 min	0–20 μM	8.20 nM	[11]
Integrated nano-clay and CDs	Milk, honey, tap water, and lake water	5 min	0.1–20 μM	37.50 nM	[39]
Amperometric magneto-immunosensor	Milk	30 min	0.04–0.42 μM	0.02 μM	[10]
Glutamic acid–capped IO-QDs	Urine	2 min	0.1–80 μM	7.69 nM	This work

Detection of TCys in real samples

The spiking experiments for the TCy sensing to examine their potential in the real application were performed. Urine samples were collected from a healthy volunteer person. Raw and undiluted urine was used, and fluorometric analysis was conducted directly on urine without any other pretreatment. Generally, auto-fluorescence emerges in the biological samples is commonly observed due to their biomolecules present in raw urine [40]. Here, the influence of auto-fluorescence from urine samples is negligible as less peak is observed in bare urine, as plotted in Fig. 6. Our method can detect TCy without any disturbance from auto-fluorescence originating from urine. Different concentrations of TCy were spiked in urine samples and examined using the prepared IO-QDs. Recoveries were expressed as the ratio found to add, which correlates closely to sensor accuracy. The relative standard deviation (RSD) of all spiked concentrations are all less than 6%, and the recoveries were between 86.9% and 114.9%, as displayed in Table 2. These results have demonstrated the acceptable and great potential of IO-QDs for detecting TCy antibiotics in real samples.

Fig. 6.

Fluorescence spectra of urine, urine treated with the IO-QDs, and urine treated IO-QDs spiking with different concentrations of TCy. The raw urine was diluted tenfold with DI water. The TCy was added at different concentrations ranging from 1 to 100 μM with respect to the actual raw urine sample. The spectra were recorded at λ_ex = 330 nm

Table 2.

Determination of TCy in the real sample using IO-QDs

TCy antibiotic	Analytical parameter	The amount found (μM)	Added (μM)	Total found (μM)	Recovery (%)	RSD (%, n = 3)
TCy	Intra-day	N/A	1	0.86 ± 0.04	86.89 ± 4.41	5.08
			5	5.50 ± 0.21	110.05 ± 4.22	3.83
	Inter-day	N/A	1	1.07 ± 0.04	107.68 ± 4.73	5.36
			5	5.74 ± 0.09	114.96 ± 1.88	1.64
CTCy	Intra-day	N/A	1	1.05 ± 0.02	105.22 ± 2.14	2.03
			5	5.02 ± 0.15	100.44 ± 3.06	3.04
	Inter-day	N/A	1	1.04 ± 0.01	104.07 ± 1.43	1.38
			5	4.40 ± 0.24	88.13 ± 4.95	5.61
OTCy	Intra-day	N/A	0.5	0.52 ± 0.01	104.09 ± 2.02	1.94
			1	1.03 ± 0.02	103.81 ± 2.64	2.54
	Inter-day	N/A	0.5	0.46 ± 0.01	93.08 ± 3.04	3.27
			1	1.07 ± 0.02	107.94 ± 2.06	1.91
DmCy	Intra-day	N/A	1	1.05 ± 0.71	105.81 ± 5.04	4.76
			5	5.15 ± 0.09	103.10 ± 5.15	4.99
	Inter-day	N/A	1	1.02 ± 1.09	102.13 ± 5.47	5.36
			5	5.66 ± 1.16	113.29 ± 1.16	1.02

N/A, not available

Conclusions

This study developed a strategy for probing tetracycline in raw urine samples using glutamic acid–capped IO-QDs as the sensing material, providing a quick and sensitive approach. However, multi-step preparation is required to acquire the fluorescent properties. This method has a broader detection range compared to other methods to detect TCy. 1.71 nm of Fӧrster distance indicated that static quenching existed. The IO-QDs-based nanoprobe was used to analyze TCy, CTCy, DmCy, and OTCy in real urine samples, and no interference from auto-fluorescence. It was applied to human urine samples with acceptable recovery. In brief, the present study has demonstrated a rapid, sensitive, eco-friendly, and biocompatible nanosensor for analyzing tetracycline in practical applications and can be used in environmental safety monitoring.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contribution

S. S.: conceptualization, method, investigation, initial manuscript. M. Z.: investigation, formal analysis. S. D.: visualization and software. S. C. N. H.: investigation, formal analysis. G. G. H.: project administration, manuscript preparation, supervision, writing review, and editing.

Data Availability

The authors confirm that the data supporting the findings of this study are available within the manuscript and its supplementary information. Raw data supporting this study's findings are available from the corresponding author upon reasonable request.

Declarations

Conflict of interest

The authors declare no competing interests.