Sensitive and selective determination of tetracycline in milk based on sulfur quantum dot probes

A novel fluorescent probe based on sulfur quantum dots (SQDs) was fabricated for sensitive and selective detection of tetracycline (TC) in milk samples. The blue emitting SQDs were synthesized via a top–down method with assistance of H2O2. The synthesized SQDs showed excellent monodispersity, water solubility and fluorescence stability, with a quantum yield (QY) of 6.30%. Furthermore, the blue fluorescence of the obtained SQDs could be effectively quenched in the presence of TC through the static quenching effect (SQE) and inner filter effect (IFE) between TC and SQDs. Under the optimum conditions, a rapid detection of TC could be accomplished within 1 min and a wide linear range could be obtained from 0.1 to 50.0 μM with a limit of detection (LOD) of 28.0 nM at a signal-to-noise ratio of 3. Finally, the SQD-based fluorescent probe was successfully applied for TC determination in milk samples with satisfactory recovery and good relative standard deviation (RSD). These results indicate that the SQD-based fluorescent probe shows great potential in practical analysis of TC in real samples with high rapidity, selectivity, and sensitivity.


Introduction
Optical probes based on luminescent nanomaterials have drawn considerable attention due to their rapidity, sensitivity, low cost and accuracy for screening of contaminants. [1][2][3] In particular, many uorescent nanomaterials have been applied for fabrication of uorescent probes. Zhang has reported a uorescent probe based on lanthanum loaded graphitic carbon nitride nanosheets for Fe 3+ detection with high selectivity and sensitivity. 4 A uorescent probe based on S-doped carbon dot-embedded covalent-organic frameworks (CDs@COF) was fabricated by Liu et al. for determination of histamine. 5 A sensitive turn-on uorescent probe based on MnO 2 -nanosheet-modied upconversion nanoparticles was developed by Chu et al. for sensitive detection of H 2 O 2 and glucose in blood samples. 6 However, the fabrication of uorescent probes based on nanomaterials with high water solubility, low toxicity, and green synthesis, combined with an obvious and direct uorescence response for target molecules still meets its limitations.
Sulfur quantum dots (SQDs), as a novel class of metal-free quantum dots, have similar advantages including good water solubility, low toxicity, and excellent biocompatibility [7][8][9][10][11] compared to other metal-free quantum dots. Additionally, SQDs have been considered as a promising green uorescent nanomaterial in recent years according to the simple and green synthesis process. 12 SQDs were rst reported in 2014 by Li's group. 11 They applied HNO 3 as an oxidant to slowly oxidize S 2À from CdS QDs to prepare SQDs. However, quantum yield (QY) of their SQDs was as low as 0.549%, and only blue light was observed. In 2018, Shen's group 13 rstly reported a top-down method to convert sublimated sulfur into SQDs through an "assembly-ssion" reaction with the QY of 3.8%. The noticeable disadvantage of Shen's method was the long synthesis time up to 125 h. In the following years, many researchers have been focusing on improving QY and shortening the reaction time including using copper-ion-assisted precipitation etching, 14 oxygen accelerated synthesis, 15 ultrasonication-promoted synthesis, 16 hydrothermal reaction 17 and ultrasonicmicrowave-assisted etching methods. 18 The SQDs reported above demonstrate particular optical properties, superior dispersibility, favorable biocompatibility, and inherent antibacterial properties, which makes them potential candidates for the fabrication of uorescent probes. Nevertheless, the practical applications of SQD-based uorescent probes are still in the primary stage.
Tetracycline (TC), as the most famous member of tetracyclines (TCs), is widely used for treatment of bacterial infections in humans and animals on account of its broad-spectrum antimicrobial activity, low toxicity, low cost, and good oral absorption. 19,20 Unfortunately, the serious abuse of TC by many manufacturers due to its effectiveness and low price has exhibited several potential threats. At present, several analytical methods including microbiological, 21 enzyme-linked aptamer assay, 22 capillary electrophoresis, 23 high-performance liquid chromatography 24 and colorimetric method, 25 have been reported for determination of TC. However, the above methods still meet limitations in long time, high cost and complex operation process, which need to be improved with advanced nanomaterials. Considering the high side effects and urgent requirement of accurate quantitative analysis of TC, it is still necessary to develop a quick and simple method for the determination of TC. [26][27][28] Fabrication of uorescent probes based on SQDs is an ideal candidate for determination of TC with high sensitivity, selectivity and accuracy.
In this paper, a novel uorescent probe based on blue emitting SQDs were fabricated for sensitive and selective detection of TC. The SQDs were synthesized via an H 2 O 2 -assisted top-down approach and several characterizations were conducted to veried the successful synthesis of SQDs. Furthermore, the SQDs exhibit satised uorescence stability. Under this condition, the static quenching effect (SQE) and inner lter effect (IFE) and were applied to determine the content of TC based on SQDs based uorescent probe (Scheme 1) and the linear relationship between the concentration of TC and uorescence intensities was also investigated. Finally, the SQDs based uorescent probe was also used to determine TC in milk samples collected from local market with satised results.
The SQDs based uorescent probe shows huge potentials in rapid, selective and sensitive determination of contaminates in food samples.

Preparation of SQDs
SQDs were prepared according to previous report with some modications. 29 In general, NaOH (4.0 g) and PEG-400 (3.0 mL) were dissolved in 50.0 mL of redistilled water in a round-bottom ask before gradual addition of sublimed sulfur (1.4 g). Then, the mixture was continuously stirred and heated at 70 C for 72 h. Aer cooling down to room temperature, H 2 O 2 solution (30 wt%, 2.0 mL) was quickly added into the above resulted solution with volume ratio of 2 : 5 under vigorous stirring followed by been stirred for another 30 min. Finally, the obtained yellowish solution was puried through dialysis against distilled water (500 Da, molecular weight cutoff) to obtain SQDs.
The puried SQDs were stored in the refrigerator at 4 C for further use.

Measurement of quantum yield
QY of SQDs was measured according to the method reported previously, 30 using quinine sulfate (QY is 54% in 0.1 M H 2 SO 4 solution) as standard. First, the absorbances (between 0.01 and 0.1) and uorescent spectra of quinine sulfate and SQDs solutions were obtained at a wavelength of 355 nm. Then, the QY of SQDs was calculated by the following equation: where F represents the QY. Subscript x and s represent the SQDs and quinine sulfate, respectively. k is the slope obtained from the plot of the integrated uorescence intensity versus absorbance, and h corresponds to the refractive index of solvent. h x and h s both are 1.33.

Fluorescence detection of TC
For TC determination, 2.0 mL SQDs and 250 mL phosphate buffer saline (PBS, pH 7.0, 0.2 M) were rst mixed together and then added into the TC solutions with different concentrations followed by been diluted to 5.0 mL with redistilled water. The uorescence emission spectra of the solutions were measured at an excitation wavelength of 355 nm.

Determination of TC in milk samples
Milk samples were obtained from local supermarket and pretreated according to previous reported method. 31 Firstly, the proteins in the milk samples were precipitated by adding 1% (w/ v) trichloroacetic acid into the samples and sonicating for 10 min. Secondly, the mixture was centrifuged at 12 000 rpm for 5 min to remove the proteins. Thirdly, the obtained supernatant was ltered through a 0.22 mm membrane to remove lipids. For standard addition recovery experiment, different concentrations including 7.0, 10.0 and 14.0 mM of TC were spiked into milk samples.

Characterization of SQDs
The morphologies and size distribution of SQDs were observed by TEM. As shown in Fig. 1A, the as-prepared SQDs are well monodispersed with a nearly spherical morphology, which may be attributed to the electrostatic repulsion between the anionic groups on the surface of SQDs. HR-TEM image (Fig. 1B)   . 11,15 The binding energies at 166.2 eV, 167.7 eV and 169.4 eV are attributed to SO 3 2À . 29 The peak appeared at 168.4 eV comes from SO 4 2À . 15 In addition, the use of passivation agent PEG-400 during the synthesis of SQDs is crucial for the uorescence activity and stability of obtained SQDs. 13,29 Thus, the FT-IR spectra of pure PEG-400 and the prepared SQDs were also measured. As shown in Fig. 2B, the peaks at 1456 cm À1 , 1352 cm À1 , 1113 cm À1 and 950 cm À1 are attributed to the existence of PEG-400 on the surface of SQDs. The peaks at 1456 cm À1 and 1352 cm À1 are both ascribed to C-H bending vibration. 33,34 The peaks centered at 1113 cm À1 and 950 cm À1 could be attributed to the stretching vibration of C-O-H or C-O-C. 33 The peak at 2873 cm À1 in PEG-400 is split into two relatively sharper peaks at 2912 cm À1 and 2876 cm À1 , both of them are attributed to the stretching vibration of C-H. 35,36 The peaks observed at 3411 cm À1 and 1640 cm À1 are belonged to -OH and C]O, respectively. 37 No other new IR peaks are observed from the synthesized SQDs, which indicates the physical interaction between the SQDs and PEG-400 instead of chemical interaction. The XPS and FT-IR characterization results above show that the surface of the synthesized SQDs is rich in hydrophilic groups, so the obtained SQDs have high water solubility and can be used as uorescent probes in aqueous media.

Optical properties of SQDs
To further explore optical properties of SQDs, the absorption, excitation and emission spectra of SQDs were measured. As shown in Fig. 3A, a weak peak observed at 259 nm is possibly belonged to the n-p* transition of S atoms. 38 The uorescence emission spectra of SQDs provided in Fig. 3B demonstrate excitation-dependent emission behavior, in which the emission intensity gradually enhances when the excitation wavelengths increase from 315 to 355 nm and then declines as excitation wavelengths further increase from 355 to 395 nm, accompanied by a red shi from 433 to 464 nm. This phenomenon, consistent with previous reports, possibly results from the inhomogeneous size distribution of particles. 29 Meanwhile, the maximum emission peak appears at 440 nm under the excitation at 355 nm, which is the characteristic blue uorescence of SQDs. Additionally, the QY of SQDs was calculated to be 6.30% at 355 nm excitation using quinine sulfate as the standard.

Fluorescence stability of SQDs
Fluorescence stability is of great signicance to the practical sensing application of SQDs. Thus, the effects of pH value, concentrations of NaCl, UV irradiation time and temperature on the uorescence intensity of SQDs were investigated before the further application of SQDs as uorescent probe in sensing. As described in Fig. 4A, the normalized uorescence intensities of SQDs remain stable and strong as the pH value varies from 3 to 11, implying the SQDs show excellent optical stability even under extreme pH conditions. Fig. 4B displays the uorescence emission intensity at 440 nm of SQDs has negligible change when incubated with increasing NaCl concentration from 0 to 1.0 M, which suggests SQDs have good property of resisting salt effect. Moreover, continuous UV irradiation for 60 min only causes slight change in uorescence intensity of SQDs (Fig. 4C), indicating that SQDs have good anti-photobleaching capability.
In addition, it can be seen from Fig. 4D that the uorescence intensity of SQDs is obviously affected by temperature, showing relatively poor temperature stability of SQDs. Therefore, the SQDs based uorescent probes should be used at a constant temperature. In consideration of the accuracy of detection and the requirements of practical analysis, room temperature is chosen to carry out the subsequent experiment in this work. These results demonstrate the excellent uorescence stability of SQDs around room temperature which guarantees the stable analytical performance.

Optimization of detection conditions
To achieve high sensitivity of the detection of TC, the effect of pH value of PBS buffer and incubation time on uorescence quenching ratios (F 0 /F) were investigated, respectively (F 0 and F represent the uorescence intensity of SQDs in the absence and presence of TC, respectively). As depicted in Fig. 5A, obvious quenching phenomenon can be observed aer addition of 50 mM TC into SQDs solution over the pH range of 3-11 and the maximum value of F 0 /F is obtained when the pH value of PBS buffer is 7.0. As such, pH 7.0 was chosen to carry out the subsequent experiment. Moreover, as illustrated in Fig. 5B, the values of F 0 /F rapidly increase in the range of 0-1 min and then remain constant aer 1 min when SQDs was incubated with 20 or 50 mM TC, indicating that TC can rapidly quench the uorescence of SQDs. Thus, 1 min was selected as the optimal incubation time.

Fluorescence selectivity of SQDs
Selectivity is a critical factor to evaluate the performance of optical sensors. To evaluate the selectivity of SQDs, the uorescence quenching ratios towards some possible interfering species, including different antibiotics (CTC, OTC, NOR, SMZ, SMX, AMO, SM, GEN, ROX), various biomolecules (VB1, VC, Glu, UA, DA, GSH, Cys, His, Gly, Phe, Arg, Lys, Tyr) and common metal ions (Al 3+ , Ni 2+ , Co 2+ , Ba 2+ , Fe 3+ , Cr 3+ , Cu 2+ , Hg 2+ , Pb 2+ , Cd 2+ ) were investigated. All the measurements were conducted under the same conditions. As shown in Fig. 6, the uorescence quenching ratios of the nanoprobe exhibit an obvious change towards TCs (TC, CTC and OTC), while the responses towards other species are negligible. It can be concluded that the SQDs possess outstanding selectivity toward TCs, which prove the great feasibility of TCs determination based on SQDs uorescent probe. In this study, TC with the best quenching effect and the most widely used is selected as a representative of TCs for further specic analysis.

Fluorescence analysis of TC
The relationship between uorescence intensity at 440 nm and the different concentrations of TC was evaluated by adding     (Table 1), the uorescent probe fabricated in this work shows unique advantages in simple operation, high sensitivity, and wide linear detection range.

Mechanism investigation
To further explore the uorescence quenching mechanism of SQDs, some related experiments were carried out. First of all, the uorescence lifetime of SQDs was studied and the uorescence decay curves of SQDs in the absence and presence of TC are presented in Fig. 8A. The average luminescence lifetimes of SQDs in the absence and presence of TC are calculated to be 2.26 ns and 2.22 ns, respectively. Apparently, TC hardly affects the uorescence lifetime of SQDs, indicating the absence of uorescence resonance energy transfer (FRET) and dynamic quenching effect (DQE) because the uorescence lifetime of SQDs will be shortened in the presence of TC when these two mechanisms exist. 47,48 It can be speculated that the uorescence quenching may result from the SQE or IFE between SQDs and TC. It is considered that the spectral overlap between absorption spectra of the quencher and the excitation spectra of the uorophore will result in IFE. 49 Therefore, the UV-Vis absorption spectrum of TC and excitation spectrum of SQDs were further recorded (Fig. 8B). It can be seen that the excitation spectrum of SQDs overlaps well with the UV-Vis absorption spectra of TC, which further conrms the existence of IFE. Furthermore, the Stern-Volmer equation was also utilized to describe the uorescence quenching. 50 where F 0 and F represent the uorescence intensity of uorophore in the absence and presence of quencher, respectively. K sv and c q represent quenching constant and the concentration of quencher, respectively. K q represents the quenching rate constant and s 0 is the uorescence lifetime of uorophore. K sv is approximately 2.9 Â 10 4 M À1 from the slope of the linear regression equation in Fig. 7B, and K q is calculated to be 1.28 Â 10 13 M À1 s À1 based on the values of above K sv and the uorescence lifetime of SQDs (2.26 ns), which is signicantly higher than the possible value of DQE (1.0 Â 10 10 M À1 s À1 ). The constant uorescence lifetime combined with the high quenching rate constant conrm the existence of SQE. 50-52 From the above discussion, it can be concluded that the quenching of SQDs by TC is mainly based on a combination of IFE and SQE.

Detection of TC in milk
In order to verify the practical applicability of SQDs based uorescent probe in real samples analysis, the prepared SQDs were employed for the analysis of TC in milk followed by the standard addition method. As displayed in Table 2, the TC cannot be detected in milk samples and the recoveries range from 92.57% to 105.40% with low relative standard deviations (RSD) of 1.34-2.45%. These practical analysis and standard addition results indicate relative high accuracy and reproducibility for TC determination, implying the fabricated uorescent

Conclusion
In summary, a novel uorescent probe based on blue emitting SQDs were fabricated for sensitive and selective detection of TC in milk samples. The synthesized SQDs showed prominent and favorable uorescence stability. Furthermore, the fabricated uorescent probe exhibited unique selectivity for TCs due to effective SQE and IFE between SQDs and TCs. TC was selected as a representative testing example for TCs. Under the optimum conditions, the fabricated nanoprobe exhibited a good linearity for TC from 0.1 to 50.0 mM with limit of detection (LOD) of 28.0 nM. Finally, the SQDs based uorescent probe was also used to determine TC in milk samples with satised results. The SQDs based uorescent probe shows huge potentials in rapid, selective and sensitive determination of contaminates in food samples.

Conflicts of interest
The authors declare that they have no competing interests.