Introduction

Antibiotics are drugs that are primarily used to treat bacterial infections. The first successful case of penicillin treatment in 1941 marked the beginning of its widespread use. The discovery of penicillin also led to rapid development in the pharmaceutical industry, with more than 160 new antibiotics or semisynthetic derivatives introduced by 19791,2,3. However, the overuse of antibiotics in healthcare, animal husbandry and aquaculture means that many incompletely metabolised antibiotics enter the environment in their original form or as active ingredients through excreta4. These residues contaminate waterways via various pathways, leaving low concentrations of antibiotics in the environment over long periods. Prolonged exposure to these drugs may lead to bacteria gradually developing resistance, which could eventually result in the emergence of multidrug-resistant ‘superbugs’ and ‘superviruses’, posing a serious threat to human safety. In recent years, antibiotics have received widespread public attention, primarily due to the fact that drug abuse results in antimicrobial resistance and environmental pollution, and antibiotic residues have been found in almost all aquatic environments around the world. Therefore, new technologies or materials are needed worldwide to mitigate environmental pollution caused by improper wastewater treatment6,7.

Although traditional analytical detection methods8,9 can accurately analyse and detect antibiotics in the environment, their disadvantages-such as high time consumption, cost and technical requirements, limited selectivity and the inability to be used for on-site monitoring-limit their use. Therefore, the focus of current research is on highly selective and sensitive analytical detection methods that can be used efficiently for the detection of antibiotics in the environment.

Carbon dots (CDs) are a novel type of carbon nanomaterial with unique fluorescent properties. Like semiconductor nanomaterials, they are spherical or spherical-like particles that are usually less than 10 nm in size. Due to their superior properties, such as water solubility, photostability, biocompatibility and biotoxicity, they have received widespread attention in various applications10,11. In terms of analytical detection, the use of fluorescent carbon quantum dot probes has many advantages, including a fast response time, high sensitivity, simple equipment and ease of operation. This method is ideal for pollutant detection12,13. However, single-intensity-based fluorescent probe analysis methods are susceptible to interference from many factors independent of the analyte, such as variations in the local concentration of the probe, fluctuations in the excitation source, light scattering from the sample matrix and instrumental factors, such as the specific microenvironment around the probe15,16. Consequently, there has been growing interest in studying ratiometric fluorescent probes, where the analyte can differentially affect the emission peak intensity of dual-emission probes. Shortcomings of single-intensity fluorescence sensing methods can be overcome by establishing an internal reference ratio to provide a self-calibrated response signal, thereby increasing the accuracy of quantitative analysis17,18,19. Currently, many scholars have demonstrated that doping with heteroatoms can effectively enhance the optical properties of carbon quantum dots. Additionally, doping with metals increases the functional properties of carbon quantum dots21,22,23,24,25.

In recent years, a great deal of research has been carried out by scholars on the preparation of carbon quantum dots and on analytical and detection methods for antibiotics in the environment. This research has focused on the following aspects, Deng et al.25synthesised antioxidant carbon dots (E-CDs) with excellent fluorescence properties using citric acid and ethylenediamine as raw materials. The quantum yield was 81.97%. The E-CDs exhibited a specific and sensitive response to •OH. The fluorescence of the E-CDs quenched by formed •OH could be restored through a competitive reaction with ampicillin (AMP). Roghayeh et al26. developed a ratiometric fluorescent sensor containing different coloured carbon dots (CDs) as dual fluorophores and a mesoporous, molecularly imprinted polymer receptor (B/YCDs@mMIP) for the detection of penicillin G (PNG) in milk.

In this study, the use of banana peel as a starting material for the synthesis of CDs is based on various considerations such as its rich chemical composition, environmentally friendly synthesis pathway, excellent product performance, and significant socio-economic benefits. This innovative method provides new ideas for the green synthesis of CDs and reflects the sustainable development concept of “turning waste into treasure”.

The novelty and aim of the study were the use of banana peels as carbon source, potassium permanganate as source of manganese element to prepare Mn, N-CDs, which is environmentally friendly, cost-effective, non-toxic, and compared with traditional organic fluorescent molecules, the Mn, N-CDs has strong photostability and good resistance to photobleaching, solving the problem of signal attenuation in the detection of antibiotics in complex systems. The Mn, N-CDs provided an environmentally friendly, sensitive, stable, efficient and rapid analytical method for detecting penicillin antibiotics in the environment.

Experimental section

Reagents and instruments

Banana peels were purchased from supermarkets around the school. Erythromycin, roxithromycin, sulfamethoxypyridazine, clindamycin hydrochloride, clarithromycin and ampicillin sodium were purchased from Shanghai Myriad Chemical Science and Technology Co., Ltd.; potassium permanganate and potassium dihydrogen phosphate were purchased from Chengdu Kelong Chemical Reagent Factory; polyethylene glycol was purchased from Tianjin Guangfu Institute of Fine Chemical Industry; and hydrochloric acid was purchased from Zhejiang Longsheng Group, all of which were of analytically pure grade.Pulverizer (YB-800B, Yongkang Speed Front Industry & Trade Co., Ltd.), electronic balance (JA5003N, Shanghai Jinghai Instrument Co., Ltd.), vacuum drying oven (DZ-3BCII, Shanghai Aoanalytical Scientific Instrument Co., Ltd.), high-power numerical-control ultrasonic cleaner (KH-600KDE, Kunshan Wo Chuang Ultrasonic Instrument Co., Ltd.), electrothermal constant-temperature drum and air drying oven (DHG-9140 A, Shanghai Ranpai Industrial Co. 9140 A), Fluorescence spectrophotometer (F97PRO, Shanghai Prism Technology Co., Ltd.), Fourier transform infrared spectrometer (FTIR, WQF-530 A, Beijing Beifen Ruili Analytical Instruments), transmission electron microscope (TEM, JEOL JEM-F200, Japan), X-ray photoelectron spectrometer (XPS, Thermo Scientific K-Alpha, U.S.A.), desktop low-speed automatic balanced Centrifuge (TDZ4-WS, Changsha Meijiasen Instrument Co., Ltd.), Digital Magnetic Heating Jacket (CLT-1 A, Shanghai Li-Chen Bangxi Instrument Technology Co., Ltd.) were used in this experiment.

Preparation of Mn, N-CDs

Pretreatment of banana peels, rinsing, chopping, baking (220 °C, 1 h), Crushing into a fine powder. To prepare a polyethylene glycol aqueous solution with a concentration of 0.1 g/mL, 0.7 g of banana peel powder obtained from pretreatment was added to 70 mL of polyethylene glycol solution, 0.5 g of potassium permanganate was added, the mixture was stirred, the mixture was transferred to the reaction kettle and heated at 160 °C, the reaction was 7 h, and the kettle was removed and cooled at the end of the reaction. Then, the reaction mixture was filtered to remove solid particles by a 0.22 μm disposable filtration membrane and subsequently dialysed with 3000 Da dialysis bag for 48 h. The above solution transferred to a vacuum drying oven at 105 °C for 6 h. Finally, it was ground into powder with a mortar and pestle and stored in a sealed bag in a cool place.

Characterization of the Mn, N-CDs and fluorescence quantum yield determination

The fluorescence spectra of the Mn and N-CDs and their intensities were obtained with a fluorescence spectrophotometer (FS) to determine the optimal emission and excitation wavelengths. UV analysis was carried out via an ultraviolet-visible (UV) absorption spectrometer. The morphology and particle size of the Mn and N-CDs were characterized via transmission electron microscopy (TEM), and the ratios of their elemental compositions were determined via X-ray photoelectron spectroscopy (XPS). The functional groups on the surface of the prepared Mn, N-CDs were identified by a FTIR spectrophotometer. Mixed Mn, N-CDs powder with KBr powder in a ratio of 1:100, then thoroughly grind and mix the sample evenly, and then load it into a compression mold and press it into shape. The resolution of the FTIR used for measuring samples is 4 cm−1, and the sample is scanned 20 times with a scanning range of 400 cm−1 to 4000 cm−1. The quantum yield (QY) of quinine sulfate, which dissolved in 0.1 M H2SO4, was specified as a reference. The QY value of quinine sulfate was determined as 54% under excited wavelength of 360 nm. Based on the equation, calculating the QY value of Mn, N-CDs was calculated, QCDs=QR(ICDs/IR)(AR/ACDs)(ɳ2CDs/ɳ2R).

Detection of antibiotics by Mn, N-CDs

Mix 1mL of different concentrations of ampicillin sodium solution (0–20 µg/mL), 50 µL of Mn, N-CDs (0.01 g/mL), and 2 mL of acetate buffer solution (pH 3.0) with distilled water to a volume of 5 mL. Mix with an ultrasonic cleaner for 30 min, then measure the fluorescence intensity at excitation wavelength of 451 nm and emission wavelength of 515 nm and record the data (F).

In order to detect the interference of Mn, N-CDs fluorescent probes on ampicillin in complex samples, 1 ml of antibiotic solutions with a concentration of 10 µg/mL (erythromycin, roxithromycin, sulfamethoxypyridazine, clindamycin hydrochloride, clarithromycin and ampicillin sodium) were measured, respectively. Then add 50 µL of Mn, N-CDs fluorescent probe solution and dilute the mixed solution to 5 mL with deionized water. Mix with an ultrasonic cleaner for 30 min, transfer the reacted solution to a quartz colorimetric dish, and measure the fluorescence intensity (F) using fluorescence spectrophotometry. Record the fluorescence intensity of blank Mn, N-CDs solution was F0.

Conditioned reaction experiments

Mixed 1 ml of ampicillin sodium solution with 50 µL of CDs solution, then add 2 ml of acetate buffer solutions with 5 different pH values (pH 2, 3, 4, 5, and 6), and dilute to 5 ml with distilled water. The reaction was carried out in an ultrasonic cleaner for 45 min. Determined the optimal pH value for the antibiotic reaction by measuring the fluorescence intensity at excitation wavelength of 451 nm and emission wavelength of 515 nm.

A total of 5 groups of 1 mL ampicillin sodium solution, 50 µL carbon solution and 2 mL acetate buffer solution (pH 3) were mixed with distilled water, fixed to 5 mL, and reacted at five different sonication times (15 min, 30 min, 45 min, 60 min, and 75 min). The optimum sonication time for the antibiotic reaction was determined by measuring and recording the fluorescence intensity at an excitation wavelength of 451 nm and an emission wavelength of 515 nm via a fluorescence spectrophotometer.

Five different amounts of CDs (30 µL, 35 µL, 40 µL, 45 µL, 50 µL) were added to 1 mL of ampicillin sodium solution (pH 3, 2 mL acetate buffer solution), and the mixture was fixed to 5 mL of distilled water. The reaction was carried out in an ultrasonic cleaner for 45 min. With the help of a fluorescence spectrophotometer, the fluorescence intensity was measured at an excitation wavelength of 451 nm and an emission wavelength of 515 nm and recorded to obtain the optimum amount of carbon quantum dots for this antibiotic reaction.

Plotting of standard curves

The ampicillin sodium solution was diluted to 0.05, 5, 10, 15–20 µg/mL, and 1 mL of each of the above five different concentrations of the solution was added to a test tube. Under the same experiment (pH 3, 45 min, 40 µL of CDs), measured their fluorescence intensity and record the data, and draw a standard curve between concentration and fluorescence quenching efficiency to determine its linear relationship.

Testing of actual samples

Lake water in front of the school library was collected and centrifuged at 1000 rpm for 10 min. It was then filtered through a 0.22 μm filter to remove impurities. The prepared drug was then added to the lake water, and the concentration of the drug in the lake water was calculated and recorded. The reaction was performed under the optimal conditions determined above (pH, time, amount of carbon dots, etc.). Finally, the fluorescence intensity values were measured at the same wavelengths, recorded, and the concentrations of the drug and recovery were calculated.

Results and discussion

Characterization and analysis of the Mn, N-CDs

FT-IR analysis of the Mn, N-CDs

FT-IR allowed the identification of the presence of functional groups in the prepared samples. Figure 1 shown the FT-IR spectra of the Mn, N-CDs. The broad absorption band at 3600–3400 cm−1 was a characteristic peak formed by O-H and N-H. the absorption peaks at 3000–2800 cm−1 and 1110.15 cm−1 were attributed to -COOH. The absorption peak at 3500 ~ 3300 cm−1 corresponded to N-H, at 2886.54 cm−1 corresponded to -CH₃, at 1463.40 cm−1 corresponded to C-H, and at 528.19 cm−1 corresponded to Mn-O. The above results shown that the CDs contained hydrophilic oxidizing groups such as hydroxyl, carboxyl and amino groups and indicated that Mn and N atoms had been successfully doped into the CDs.

Fig. 1
Fig. 1
Full size image

Infrared spectra of Mn, N-CDs.

TEM analysis of the Mn, N-CDs

TEM was used to characterize the size and morphology of the Mn, N-CDs. Figure 2 shown electron micrographs of the samples at 200 nm, 1 μm and 10 μm, respectively, and the figures revealed that the prepared Mn, N-CDs were spherical with a relatively uniform particle size distribution, and agglomerations of these Mn, N-CDs were not observed, and the dispersion was relatively good. The average nanoparticle size was 5 nm.

Fig. 2
Fig. 2
Full size image

TEM spectra of Mn, N-CDs at 200 nm, 1 μm and 10 μm, respectively. The below illustration: the filter diameter distribution map of the Mn, N-CDs.

XPS analysis of the Mn, N-CDs

From the Fig. 3(a), it could be seen that the XPS spectra of the Mn, N-CDs had four major characteristic peaks at 285.28, 537.38, 399.08, and 640.62 eV, which were C1s, O1s, N1s, and Mn2p, and the percentages were 67.59%, 30.11%, 1.2%, and 1.1%, respectively. The C1s were divided into three main peaks, 283.78 eV (C-C/C = C) and 285.18 eV (C = N/C-O) (Fig. 3b), the N1s also was divided into two main peaks, 398.03 eV (Mn-N) and 399.70 eV (Graphitic-N) (Fig. 3c), and the O1s was divided into three main peaks, 530.08 eV (Mn-O), 531.57 eV (-OH) and 531.98 eV (C = O) (Fig. 3d). The Mn2p peak at 640.62 eV was attributed to Mn-O. The above XPS analysis revealed that Mn and N atoms were successfully doped into the CDs. The appearance of these peaks indicated that the Mn, N-CDs were dominated by carbocyclic rings and contained abundant water-soluble functional groups, which was similar to the characterization results of FT-IR.

Fig. 3
Fig. 3
Full size image

(a) XPS survey scan, (b) C1s XPS, (c) N1s XPS, (d) O1s XPS of Mn, N-CDs.

Spectral properties of Mn, N-CDs

The optical properties of the Mn, N-CDs were investigated using fluorescence (FS) and ultraviolet–visible (UV–vis) absorption spectroscopy. As illustrated in Fig. 4(b), a characteristic absorption peak was observed at 281 nm, resulting from the π–π* electron transition of the aromatic sp² carbon in the Mn, N-CDs. Figure 4(a) shows that the optimum excitation wavelength for the Mn, N-CDs was 450 nm and the optimum emission wavelength was 515 nm. The inset in Fig. 4(b) shows that the Mn, N-CDs solution appeared yellowish under sunlight and blue under a 365 nm UV lamp. The fluorescence quantum yield of the Mn, N-CDs was found to be 22.25%, using quinine sulfate as a reference.

Fig. 4
Fig. 4
Full size image

(a) Excitation (red line), emission (black line) and (b) UV-visible absorption spectra of Mn, N-CDs. Illustration: A photograph of Mn, N-CDs in visible light (left) and ultraviolet light (right).

The stability of Mn, N-CDs

To verify the stability of the Mn, N-CDs, the change in the fluorescence intensity of the Mn, N-CDs was measured after being stored at 4 °C for one month. As shown in Fig. 5(a), the fluorescence intensity remained stable without significant changes. This indicated that the Mn, N-CDs had strong photostability.

Under illumination, the fluorescence intensity of the Mn, N-CDs changes with time, as shown in Fig. 5(b). The decreasing trend of fluorescence intensity of Mn, N-CDs with time is more obvious within 0 ~ 72 min. Within 72 ~ 240 min, the decreasing trend of fluorescence intensity becomes slower, indicating that Mn, N-CDs have excellent photostability.

Acetate buffer solutions with pH values ranging from 2.0 to 12.0 were added to the Mn, N-CDs in sequence. The experimental results are shown in Fig. 5(c). When the pH of the acetate buffer solution was between 2 ~ 6, the fluorescence quenching efficiency was relatively stable, significantly decreased at pH 7, which might be the transition of surface functional group protonation state, reversal of surface charge state, mutation of aggregation behavior, and changes in the competitive relationship between radiation/non radiation composite pathways and slightly decreased but not very significantly above pH 8. This indicated that the Mn, N-CDs could be used as fluorescent nanoprobes in acidic and alkaline media.

Fig. 5
Fig. 5
Full size image

(a) Effect of different storage time, (b) The influence of illumination time, and (c) pH on fluorescence efficiency of Mn, N-CDs.

Selectivity of Mn, N-CDs

The changes in fluorescence intensity were obtained via the reactions of six different antibiotics with the prepared Mn, N-CDs in neutral, acidic and basic buffer solutions. From the analysis of the data in Fig. 6, it can be concluded that the effects of the Mn, N-CDs on ampicillin were significant, whereas the effects on the other five antibiotics were not obvious. As shown in Fig. 6, ampicillin sodium had the best fluorescence quenching efficiency under acidic conditions.

Fig. 6
Fig. 6
Full size image

Effect of different antibiotics on fluorescence intensity of Mn, N-CDs (n = 3).

Optimization of detection conditions

To improve the performance of the Mn, N-CDs for the detection of ampicillin sodium, several variables affecting the detection were specifically optimized, including the pH, response time and amounts of Mn, N-CDs. The quenching efficiency was F/F0 throughout the experiment, where F and F0 are the fluorescence intensities of the Mn, N-CDs with and without the addition of sodium ampicillin, respectively. When the pH was varied between 2.0 and 6.0 with the acetate buffer solution, there was a significant change in the fluorescence quenching efficiency of the solution, and the optimized value was pH 3, as shown in Fig. 7 (a). Similarly, the response time and amount of Mn, N-CDs were optimized to 45 min and 40 µL, as shown in Fig. 7 (b) and (c), respectively.

Fig. 7
Fig. 7
Full size image

(a) fluorescence quenching efficiency of Mn, N-CDs at different pH values, (b) Effects of different reaction time on the fluorescence quenching efficiency of Mn, N-CDs, (c) Effect of different amounts of Mn, N-CDs on fluorescence quenching efficiency (n = 3).

Linearity and selectivity

To verify the sensitivity of the method. Under the optimal conditions (pH 3, 45 min, 40 µL), different fluorescence intensities were obtained by adding different concentrations of ampicillin sodium, the data were recorded, and a standard curve was plotted, as shown in Fig. 8. When the concentration of ampicillin sodium was in the range of 0.05 ~ 20 µg/mL, the F/F0 and the concentration of ampicillin sodium had a good linear relationship with an equation of F/F0 = 0.19311 + 0.03723 C, R2 was 0.99882, and the LOD for ampicillin sodium was calculated as 0.038 µg/mL on the basis of 3σ/S (where S represents the slope of the first fitting equation, σ is the standard deviation of the blank solution, and n = 10). Therefore, the prepared Mn, N-CDs as fluorescent probes had high sensitivity to ampicillin sodium and good potential for use in environmental detection.

Fig. 8
Fig. 8
Full size image

The linear fitting plot of the fluorescence intensity of Mn, N-CDs and ampicillin sodium (n = 3).

Detection of actual samples

Different concentrations of ampicillin sodium (1 µg/mL, 5 µg/mL and 10 µg/mL) and Mn, N-CDs were added to the treated lake water for the reaction under the optimal experimental conditions. As shown in Table 1, the spiked recoveries of ampicillin sodium at different concentrations were 96.0%, 98.20% and 97.40%, with RSDs of 1.38%, 2.41%, and 1.40%, respectively. Therefore, the Mn, N-CDs produced from banana peels as carbon sources and potassium permanganate as the doped metal element Mn have great potential for use as novel fluorescent probes for the detection of penicillin antibiotics in environmental water samples.

To confirm that this method presented an advance in the state of the art, a comparison with other sensors is shown in Table 2. The fluorescent probe synthesis in this study exhibits unique characteristics in terms of high sensitivity, simple operation, and wide linear detection range compared to other reported fluorescent probes.

Table 1 Preparation of Mn, N-CDs as fluorescent probes for the detection of ampicillin sodium in real water samples. (n = 3).
Full size table
Table 2 Comparison with reported methods for determination antibiotic. (n = 3).
Full size table

Mechanisms of ampicillin sodium detection by Mn, N-CDs

According to the literature, the current quenching mechanisms of CDs included dynamic quenching, static quenching and internal filtration effects32,33,34,35. Dynamic quenching was mainly caused by collisions between the luminescent material and the quencher, and the fluorescence lifetime changed significantly with or without the use of quencher36,37. Static quenching was mainly caused by the non-luminous or weakly fluorescent matrix complex formed by the interaction between the luminescent material and the quencher, and the fluorescence lifetime was basically unchanged with or without the quencher38,39. The internal filtration effect was a type of selective filtration phenomenon, which was essentially the process of re-absorption of light emitted by CDs40,41,42.

Therefore, the UV-vis absorption spectra of the Mn, N-CD solutions before and after the addition of ampicillin sodium (1 mL, 10 µg/mL) were determined in this experiment. As shown in Fig. 9 (a), the UV-vis absorption spectra of the Mn, N-CD solutions red-shifted after the addition of sodium ampicillin, which could be attributed to π-π* jump, and the static quenching caused by the complexation of Mn, N-CDs and sodium ampicillin to produce nonluminous or weakly fluorescent ground state complex and the UV-visible absorption spectrum did not overlap with the fluorescence spectrum, indicating the absence of reabsorption effects. Meanwhile, the fluorescence lifetime curves before and after the addition of sodium ampicillin to Mn, N-CDs were also determined, and as shown in Fig. 9 (b), their fluorescence lifetimes remained essentially unchanged (τ0 = 5.15137 ns and τ=5.13916 ns, τ0/τ≈1), which reinforced the mechanism of static quenching. The fluorescence quenching process of the Mn, N-CDs on antibiotics was shown in Fig. 10.

Fig. 9
Fig. 9
Full size image

(a) UV absorption spectra of Mn, N-CDs (black line) and ampicillin sodium + Mn, N-CDs (red line) and PL spectra of Mn, N-CDs, (b) Fluorescence lifetime decay curves of Mn, N-CDs and ampicillin sodium + Mn, N-CDs.

Fig. 10
Fig. 10
Full size image

Preparation of Mn, N-CDs and their application in antibiotic sensing research process.

Conclusion

In this study, Mn N-CDs with excellent fluorescence properties were synthesised using a one-step hydrothermal method. Banana peels were employed as the biomaterial and potassium permanganate as the metal dopant. Characterisation of the Mn, N-CDs using FT-IR, TEM, and XPS showed that elemental Mn and N had been successfully doped into the Mn, N-CDs, which were found to be small in size with abundant surface functional groups and excellent fluorescence properties. The experiments concluded that adding sodium ampicillin significantly altered the fluorescence intensity of the Mn, N-CDs under acidic conditions. Spiking experiments on ampicillin sodium resulted in a small error (RSD < 5%) and high spiking recovery (88.60–93.20%). It can therefore be concluded that the prepared Mn, N-CDs are valuable for detecting antibiotics in the environment.