Introduction

Carbon dots (CDs), a new type of carbon nanomaterial with a diameter less than 10 nm, are regarded as a rising star among fluorescent materials owing to their outstanding fluorescence properties [1]. Because of outstanding advantages such as simple preparation, good fluorescence properties, good water solubility, and low toxicity [2,3,4], they have attracted tremendous interest in the fields of sensing [5], bioimaging [6], and drug delivery [7]. Because of their multimodal applications, various synthetic methods, such as the solvothermal method (including a hydrothermal method) [8, 9], the microwave method [10], and chemical oxidation [11], have been developed for the preparation of CDs. Among these methods, the solvothermal method has been applied most commonly as it is convenient and efficient. However, to obtain CDs with high quantum yield (QY), this commonly used method suffers from several limitations, such as the selection of precursors and the toxicity of solvents. Usually, the common used precursors in solventhermal method are citric acid coupled with urea [12], ethanediamine [13], tris (hydroxymethyl) methyl aminomethane [14] and etc. Moreover, some reactions even need to use toxic solvents such as toluene [15] or dimethyl fumarate [16] to obtain CDs with high QY. The use of these chemicals and toxic solvents does not meet the standard of green chemistry and is harmful to humans. Therefore, there is a need to explore a green synthetic approach for the preparation of CDs by use of green precursors and nontoxic solvents.

Lately, various natural raw materials, such as milk [17], pseudostem of the banana plant [18], onion waste [19], Lantana camara berries [20], and tulsi leaves [21], have been used as green precursors for the preparation of CDs by a green hydrothermal method. Because of the presence of multiple components, the CDs prepared with these natural raw materials have tunable surface functional groups. These groups can improve the photoluminescence properties of CDs, resulting in increased sensitivity and selectivity of CDs for target molecules. Xu et al. [22] prepared CDs by using ginkgo leaf as a green precursor. The CDs were used for ultrasensitive, highly selective detection of Pb2+ with a limit of detection (LOD) of 55 pM. In addition, these CDs prepared with green precursors and with water as a solvent have good water solubility and biocompatibility and can be applied in the bioimaging field [23]. Li et al. [24] prepared CDs by using Brassica compestris L. var. purpurea Bailey as a precursor and water as a solvent: The CDs had low cytotoxicity toward HepG2 cells and labeled them successfully. From these points, preparation of CDs with good physicochemical and photoluminescence properties by use of green precursors and a green synthetic method is very desirable.

The flowers of Osmanthus fragrans Lour., also named “Guihua” in China, are especially famous as their essential oil can be widely used in aromatherapy and to produce perfumes [25, 26]. They also exhibit many pharmacological activities, such as antioxidant and antitumor activities [25, 27]. The flowers of O. fragrans are rich in flavonoids, phenylethanoid glycosides, and phenolic acids [28]. The diversity of the constituents makes the flowers a potential natural raw material for the preparation of CDs. Moreover, flavonoids in the flowers can easily interact with metal ions such as Fe3+ and Pb2+ [29, 30]. Hence, CDs prepared with flowers of O. fragrans as green precursors may be used for the detection of metal ions. In addition, as the planting of O. fragrans in China is widespread, many flowers of O. fragrans are discarded and not fully utilized. Therefore, use of the flowers of O. fragrans as precursors is also an effective method to maximize the utilization of natural resources.

In present study, flowers of O. fragrans were used as green precursors for the preparation of new CDs by a simple hydrothermal method for the first time. By use of this method, high-quality CDs with a QY of 18.53% can be produced on a large scale without the assistance of any chemical reagent. Optical and chemical properties of the CDs were characterized by UV–visible (UV–vis) absorption spectroscopy, fluorescence spectroscopy, X-ray diffraction (XRD), Fourier transform IR (FT-IR) spectroscopy, transmission electron microscopy (TEM), X-ray photoelectron spectroscopy (XPS), and energy-dispersive X-ray spectroscopy (EDS). Because CDs prepared with the flowers of O. fragrans as precursors have potential in metal ion detection, the CDs were applied for the fluorescent detection of Fe3+. They showed high sensitivity and selectivity toward Fe3+ with a linear range of 10 nM to 50 μM. The LOD of Fe3+ detection was 5 nM, which was lower than for previously reported CDs (prepared with green precursors) [26, 31]. Moreover, the CD–Fe3+ quenching mechanism was studied by UV–vis detection and fluorescence lifetime determination. The mechanism was identified as the inner filter effect (IFE). Furthermore, the CDs were further applied to the detection of Fe3+ in real water samples and human serum with a recovery range of 93.8–113.8%, which confirms the good analytical performance of this CD-based sensor. The fluorescence of CDs quenched by Fe3+ can be partly recovered by addition of ascorbic acid (AA), and this change can be used for further detection of AA by an “off–on” sensor. Finally, to explore toxicity and biocompatibility, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) measurement and cell imaging were conducted on a human lung cancer (A549) cell line. The results demonstrate that the CDs exhibited low toxicity and good biocompatibility with regard to A549 cells, and Fe3+ (AA) can also be qualitatively analyzed by monitoring of the changes in cell fluorescence. In general, the CDs have great potential for application in detection of Fe3+ and AA and can be used as a fluorescent probe for cell imaging.

Experimental

Chemicals and reagents

The flowers of O. fragrans were picked from Southwest Medical University (Luzhou, China) and dried at room temperature. Quinine sulfate (purity greater than 99%) was bought from Acros Co. (https://www.acros.com). AA and reagents related to ions were purchased from Kelong Chemical Co. (http://www.cdkelong.com/cn). A549 cells were obtained from the Cell Center of the Chinese Academy of Medicinal Sciences (http://cellbank.bioon.com.cn). Phosphate-buffered saline (PBS), penicillin–streptomycin, and MTT were obtained from Beijing Dingguo Biotechnology Co. (http://www.dingguo.com). Dulbecco’s modified Eagle’s medium (DMEM), fetal bovine serum (FBS) and trypsin were obtained from Gibco (https://www.fishersci.com). Ultrapure water was used in all aqueous solutions in the whole experiment process.

Instruments

The UV–vis absorption spectra of CDs were obtained with a UV-2600 spectrometer (Shimadzu, https://www.shimadzu.com). The fluorescence spectra of CDs were obtained with an LS-55 fluorescence spectrometer (PerkinElmer, http://www.perkinelmer.com.cn). A JEM 2100 transmission electron microscope (JEOL, https://www.jeol.co.jp) was used to acquire the TEM image. XRD measurements were performed with an Ultima IV X-ray diffractometer (Pabakytucal X′Pert Powder, http://www.spectris.com). FT-IR spectra were obtained with an IRAffinity-1S FT-IR spectrophotometer (Shimadzu, https://www.shimadzu.com). XPS spectra were obtained with an ESCALAB 250Xi X-ray photoelectron spectrometer (Thermo Electron, https://www.thermofisher.com). EDS was conducted with an EX-250 energy-dispersive X-ray analyzer (Horiba, https://www.horiba.com). Fluorescence images of cells were recorded with an IX-73 inverted fluorescence microscope (Olympus, http://cn.olympus.com). The fluorescence lifetimes were obtained with an FLS980 lifetime and steady-state spectrometer (Edinburgh Instruments, https://www.edinst.com).

Preparation of CDs

A facile one-step hydrothermal method was used to synthesize CDs by use of flowers of O. fragrans as green precursors. Firstly, 35 mL of ultrapure water was added to 2 g of dried O. fragrans flowers and the mixture was ultrasonicated for 30 min until the water changed from white to brown. Then the mixture was transferred to a 50-mL Teflon-lined stainless-steel autoclave and heated at 240 °C for 5 h. After cooling to room temperature, the upper yellow-brown solution was centrifuged at 10,000 rpm for 10 min to remove most of the impurities, and was then filtered through a 0.22-μm membrane filter. The solution was dialyzed through a dialysis membrane (3000-Da tubing) for 48 h, and then the final CDs were obtained.

Fluorescence assay of Fe3+ and AA

All the Fe3+ detection experiments were performed under the same conditions at pH 6 in 10 mM PBS and the fluorescence intensity was recorded at an excitation wavelength of 340 nm. Other ions (including Fe2+, Zn2+, Ba2+, Mn2+, Al3+, Co2+, Mo4+, K+, Cd2+, Pb2+, Cu2+, Ni2+, Mg2+, Ca2+, HCO3-, HPO4-, NO2-, and SO42- ) were used to determine the selectivity of the CDs for Fe3+ under the same experimental conditions. To evaluate the sensitivity of the CDs for Fe3+, different concentrations of Fe3+ (300 μL) were added to aqueous solutions of CDs (6 μg mL-1, 2.7 mL) and the mixture solutions were measured by the fluorescence spectrometer.

For detection of AA, 300 μL of different concentrations of AA was mixed with 2.7 mL of the CD–Fe3+ mixture solution (the concentration of Fe3+ was 0.5 mM and the concentration of the CDs was 6 μg mL-1). The fluorescence spectra were obtained at an excitation wavelength of 340 nm.

Determination of Fe3+ in real samples

The pretreatment of water samples and the human serum sample is described in the electronic supplementary material. To calculate the recoveries of Fe3+ in the real samples, different concentrations of Fe3+ standard solutions were added the CD solution directly by the standard addition method. The fluorescence intensities were measured with the fluorescence spectrometer at excitation and emission wavelengths of 340 and 410 nm, respectively. Moreover, the experiment was performed three times under the same experimental conditions to obtain the relative standard deviation.

Application of CDs in living cells

MTT measurement

To explore the biocompatibility and cytotoxicity of CDs, MTT measurement was done on A549 cells. Cells were seeded in 96-well plates and incubated for 24 h in a CO2 incubator. Then different concentrations of CDs were added to the wells and incubation was performed for another 24 h. Next, 10 μL of MTT (5.0 mg mL-1) was added to each well and the resulting mixture was incubated at 37 ° C for 4 h. The solutions in all wells were then removed and 150 μL of dimethyl sulfoxide was added to each well to dissolve the formazan complex. The plate was shaken for 10 min before testing. Finally, the absorbance was detected at 570 nm by use of a Synergy H1 microplate reader (Biotek, https://www.biotek.com).

Cell imaging

The A549 cells were cultured in DMEM medium (containing 10% FBS) in six-well plates and incubated for 24 h at 37 °C until confluency. The cells were then treated with CDs at 1 mg mL-1 (by addition of 10 μL) for 3 h. Then DMEM was removed and A549 cells were washed with PBS (pH 7.4) three times and then kept in PBS. The CD-labeled cells were then treated with Fe3+ (0.4 mM) and incubated for 30 min for further fluorescence imaging. Finally, AA solution (0.4 mM) was added to the cells (with CDs and Fe3+) and the resulting mixture was incubated for 30 min for fluorescence imaging. All the fluorescence images were collected at an excitation wavelength of 340 nm. The preparation process and applications of CDs are stated in Fig. 1.

Fig. 1
figure 1

The synthesis and application of carbon dots (CDs). AA ascorbic acid

Full size image

Results and discussion

Synthesis and characterization of CDs

Compared with traditional methods, the hydrothermal method is green and facile. Moreover, it is reported that the flowers of O. fragrans contain a large amount of flavonoids and glycosides, which are nontoxic and hydrophilic and so the CDs prepared with these constituents can easily and quickly disperse in water [32, 33]. Therefore, in the present work, CDs were synthesized via a one-step hydrothermal method by use of flowers of O. fragrans as green precursors and water as a solvent. To obtain CDs with high QY, three factors (the amount of the flowers of O. fragrans, the reaction temperature, and the reaction time) influencing the QY were optimized. As shown in Table S1, the highest QY obtained was 18.53% by use of 2 g of flowers of O. fragrans as raw materials, 240 °C as the reaction temperature and 5 h as the reaction time

The morphology of the CDs was revealed by TEM. As shown in Fig. 2a, the TEM image of the CDs revealed that the average diameter of the CDs was about 2.23 nm, and it can be clearly seen that the size distribution was relatively uniform and the size differentiation was relatively small (Fig. 2b). In addition, the crystalline nature of the CDs was confirmed by XRD analysis. The XRD pattern of the CDs in Fig. 2c shows a broad diffraction peak centered at 2θ = 26.7° attributed to the (002) lattice spacing of the graphitic carbon, and this indicated that the CDs contained graphite-like structures [34, 35].

Fig. 2
figure 2

a Transmission electron microscopy image of carbon dots (CDs). b Particle size distribution of CDs. c X-ray diffraction pattern of CDs. d Fourier transform IR spectrum of CDs

Full size image

The composition and functional groups of the CDs were identified by FT-IR spectroscopy. In Fig. 2d the absorption bands at 3418 and 3239 cm-1 correspond to O–H and N–H stretching vibrations. These two functional groups can increase the hydrophilicity and stability of the CDs [36, 37]. Moreover, the small peak at 2953 cm-1 is identified as the stretching vibration of C–H bonds. And the peaks at 1663 cm-1, 1584 cm-1 and 1405 cm-1 were attributed to the C=O, C=C and CH2 stretching bending vibrations, respectively [38]. In addition, the band at 1069 cm-1 is ascribed to the stretching vibration of C–O. These results indicate that –OH, –NH2, and –COOH may exist on the surface of CDs. To further explore the elemental composition of the CDs, the EDS spectrum (Fig. S1) was obtained, and the results show that the CDs were composed of carbon (64.57%), nitrogen (3.1%), and oxygen (31.33%).

Furthermore, XPS was used for surface elemental analysis of the CDs. As depicted in Fig. 3a, the three strong typical peaks in the XPS spectrum of the CDs at 284.82, 399.84, and 531.89 eV belong to C 1s, N 1s, and O 1s. This indicates that the surface of the CDs was mainly composed of carbon, oxygen, and nitrogen [39, 40]. In detail, there were four main peaks in the high-resolution XPS spectrum of C 1s centered at 284.51, 285.69, 286.31, and 288.05 eV, which were assigned to C=C, C–N, C–OH, and C=O, respectively (Fig. 3b) [41]. However, the N 1s expanded spectrum (Fig. 3c) can be deconvoluted into three distinct carbon states at 387.90, 399.47, and 400.99 eV, which correspond to C–N–C, N–H, and C3–N, respectively [42]. In addition, as shown in Fig. 3d, two peaks appeared in the O 1s expanded spectrum, at 530.99 and 532.28 eV, which are associated with C=O (sp2) and C–O (sp3), respectively [43]. These results indicate that the CDs contained C=C, –NH2, and –COOH groups, which was in accordance with the results of FT-IR spectroscopy.

Fig. 3
figure 3

a X-ray photoelectron spectroscopy spectra of carbon dots (CDs). b C 1s spectrum of CDs. c N 1s spectrum of CDs. d O 1s spectrum of CDs

Full size image

Optical properties of CDs

The absorption and fluorescence properties of the CDs were studied by UV–vis absorption spectroscopy and fluorescence spectroscopy. Figure 4a shows the UV absorption spectrum of the CDs from 200 to 600 nm. It can be obviously seen that there was an obvious absorption peak at around 280 nm, this peak was corresponded to the n–π* electronic transition of C=O banded on CDs [44]. Figure 4b reveals that the CDs exhibited strong blue fluorescence at 410 nm when excited at 340 nm. Moreover, from the inset in Fig. 4b, it can be easily seen that the CD solution was almost colorless in daylight and emitted bright blue fluorescence under excitation at 365 nm. Figure 4c shows the different emission wavelengths of the CDs obtained with different excitation wavelengths (ranging from 300 to 420 nm with a 10-nm increment). Furthermore, the corresponding normalized spectra are displayed in Fig. 4d. As the excitation wavelength increased, the emission wavelength redshifted. This phenomenon resulted from the different surface states of CDs produced by different functional groups on the surface of the CDs and different sizes of the CDs [45].

Fig. 4
figure 4

a UV–vis absorption spectrum of carbon dots (CDs). b Fluorescence excitation spectrum and emission spectrum of CDs. The insets display images of the CDs in daylight and under 365-nm light. c Excitation-wavelength-dependent emission spectra from 300 to 420 nm. d Normalized emission spectra of CDs with different excitation wavelengths. FL fluorescence

Full size image

Fluorescence response of CDs to Fe3+

Optimization of the experimental conditions

The effects of pH, ionic strength, and solvent on the fluorescence intensities of the CDs were explored first. Respective results are given in Fig. S2. The optimized conditions for maintaining the CDs’ performance were as follows: (a) pH 6 was the optimum pH; (b) the optimum ionic strength was 0 M NaCl (almost no effect on the detection); and (c) the solvent for CD dispersion was water. Further, the stability of the CDs in water was studied for 120 min. Figure S2d shows that the fluorescence intensity was almost invariable for 120 min, indicating that the CDs had good stability.

Furthermore, to exclude some possible interferences and screen the best detection conditions for Fe3+, the following parameters for Fe3+ detection (the Fe3+ concentration used was 50 μM) were optimized: (a) the concentration of CDs, (b) the pH, (c) the concentration of NaCl, and (d) the type of solvent. Respective plots are given in Fig. S3. The optimized conditions for Fe3+ detection were as follows: (a) CD concentration of 6 μg·mL-1, (b) pH 6, (c) 0 M NaCl, and (d) water solvent. These results demonstrated the best conditions (pH, ionic strength, and solvent) for Fe3+ detection were in accordance with the best conditions for maintaining the CDs’ performance.

Selectivity and sensitivity for the detection of Fe3+

The sensing performance of the CDs was assessed by selectivity experiments with various ions. The fluorescence intensities were investigated by our adding different ions (Fe2+, Zn2+, Ba2+, Mn2+, Al3+, Co2+, Mo4+, K+, Cd2+, Pb2+, Cu2+, Ni2+, Mg2+, Ca2+, HCO3-, HPO4-, NO2-, SO42-, and Fe3+) in 0.3 mM concentration to CD solutions (6 μg·mL-1). From Fig. 5a, it is clearly seen that Fe3+ showed the strongest fluorescence quenching effect on CDs, whereas there were no remarkable fluorescence changes for the other ions. Moreover, to further determine the selectivity of CDs for Fe3+, 0.3 mM Fe3+ was added to a solution of CDs containing other ions in 0.3 mM concentration. The results revealed that the CDs had high selectivity for the detection of Fe3+ and can be used as a good fluorescent sensor for Fe3+ detection. This good selectivity of the CDs for Fe3+ was mainly attributed to photoelectron transfer caused by the strong affinity interaction between Fe3+and the –COOH and –OH groups on the surface of the CDs. Fe3+ had faster chelating kinetics in comparison with the other ions [46, 47]. All these reasons can cause the electrons of the CDs in the excited state to be transferred to the unfilled orbital of Fe3+and resulted in nonradiative electron–hole recombination, causing fluorescence quenching [39].

Fig. 5
figure 5

a Selectivity of carbon dots (CDs) for Fe3+ over other ions. b Fluorescence spectra of CDs in the presence of different concentrations of Fe3+ (0–500 μM). c F0/F versus Fe3+concentration in the range from 10 nM to 50 μM, where F0 and F are the fluorescence areas of the CDs in the absence and presence of Fe3+. FL fluorescence

Full size image

From Fig. 1, it is very apparent that the fluorescence of the CDs almost disappears under 365-nm light after addition of 0.3 mM Fe3+ to the CD solution. To further study the sensitivity of the CDs for Fe3+, the fluorescence of the CDs was investigated by our adding different concentrations of Fe3+ (0–500 μM) under the optimum conditions. Figure 5b shows that as the concentration of Fe3+ increased, the fluorescence intensity of the CDs gradually decreased. The relationship between F0/F and the concentration of Fe3+ is shown in Fig. 5c, where a good linear relationship was obtained in the range from 10 nM to 50 μM with a correlation coefficient (R2) of 0.9992. The linear equation was F0/F = 0.0234C + 1.0412, where C is the concentration of Fe3+. On the basis of three times the standard deviation of 11 measurements of the blank signal of the CD solution, the LOD was found to be 5 nM, which is lower than previously reported in the literature (listed in Table S2). These results indicate that the CDs showed high sensitivity for Fe3+, and the wide linear range and low LOD confirmed that the CDs can be used as a good fluorescent sensor.

Detection of Fe3+ in real samples

As Fe3+ is one of the most essential trace elements in the human body and water samples, its detection in real water samples and human serum samples is a good way to verify the applicability of the CDs. In the present work, tap water, two different river water samples (river water 1 representing Tuo River water and river water 2 representing Min River water), and human serum were chosen as the real samples to check the realistic application and the feasibility of the CDs. Different concentrations of Fe3+ (5, 10, and 50 μM) were added to the processed samples; the related recoveries are shown in Table 1. The recoveries of these samples ranged from 93.76% to 113.80% and the relative standard deviations were less than 1%. From these results, it can be summarized that the CDs had good performance in real samples and they can be used to effectively detect Fe3+ in water and biological samples.

Table 1 Determination of Fe3+ in real samples with the proposed method (n = 3)
Full size table

Quenching mechanism study of the CD–Fe3+ system

To investigate the quenching mechanism between CDs and Fe3+, the UV–vis absorption spectra of Fe3+and the fluorescence spectra of the CDs were studied. Figure 6a shows that there was obvious overlap between the excitation spectrum of the CDs and the absorption spectrum of Fe3+. Because of this overlap, the enhancement of absorbance of Fe3+ can be successfully converted into fluorescence quenching of CDs [48]. As the IFE usually occurs when the excitation spectrum or emission spectrum of the donor has obvious overlap with the absorption spectrum of the acceptor, Fe3+can quench the fluorescence of the CDs probably via the IFE [49]. The fluorescence decay spectra of the CDs with and without Fe3+ were investigated to further verify the mechanism. As shown in Fig. 6b, the fluorescence decay time of the CDs in the absence and presence of Fe3+ was analyzed by double-exponential fitting. Two decay lifetimes of 2.06625 and 7.56736 ns (χ2 = 0.997) were obtained, whereas the lifetimes changed to 1.78886 and 7.48038 ns (χ2 = 0.991) after addition of Fe3+ to the CD solution. The fluorescence lifetimes changed little after the addition of Fe3+ on fitting of the data, which indicated that static quenching existed and further confirms the existence of the IFE between the CDs and Fe3+ [48, 50, 51].

Fig. 6
figure 6

a Overlap between the absorption band of Fe3+ and the excitation or emission band of the carbon dots (CDs). b Fluorescence decay curves of CDs in the presence and absence of Fe3+. FL fluorescence

Full size image

Fluorescence off–on detection of AA

Because of its strong oxidization, Fe3+ has an effect on substances with strong reducibility. Because AA is highly reductive, the redox reaction between Fe3+ and AA will lead to the oxidation of the enediol group in AA along with the reduction of Fe3+ to Fe2+ (shown in Eq. S1) [52, 53]. Therefore, the weak fluorescence in the CD–Fe3+ system may be partially recovered as some Fe3+ ions are converted into Fe2+ ions (shown in Fig. 7a). Therefore, the CD–Fe3+ system can be applied to detect AA by turning fluorescence off–on.

Fig. 7
figure 7

a Fluorescence spectra of carbon dots (CDs), CDs and Fe3+, CDs, Fe3+, and ascorbic acid (AA), and CDs and AA. b The fluorescence intensity at different response times. c Fluorescence spectra of CDs in the presence of Fe3+ on addition of AA at different concentrations (0–500 μM). dF/F0 versus AA concentration, where F0 and F are the fluorescence areas of the CD–Fe3+ system in the absence and presence of AA. FL fluorescence

Full size image

To optimize the detection conditions for AA, the effect of the response time was investigated. As shown in Fig. 7b, the fluorescence intensity immediately reduced after addition of 0.5 mM Fe3+ and then remained stable. This phenomenon can be explained by there being a rapid reaction between the groups on the surface of the CDs and Fe3+, while after introducing 1 mM of AA, Fe3+ and AA reacted with each other and the fluorescence intensity increased gradually.. The rising trend continued for 4–12 min, and then the fluorescence intensity became stable after 12 min. Therefore, the optimum recovery time was selected as 12 min.

The detection ability toward AA was studied by our adding different concentrations of AA (0–500 μM) to the CD–Fe3+ system (containing 0.5 mM Fe3+) under the optimized reaction conditions. As shown in Fig. 7c, as the AA concentration increased, the fluorescence intensity gradually recovered. Figure 7d shows that there was a good linear relationship between F/F0 and the concentration of AA (range from 0 to 200 μM), with a correlation coefficient of 0.9926. The linear equation was F/F0 = 0.0041C + 1.0231, where C is the concentration of AA. The LOD of AA was 5 μM (3σ/K). These results indicate that the CD–Fe3+ system can be used as a fluorescent off–on sensor for the detection of AA.

Cell cytotoxicity assay and cell imaging

To explore the toxicity and biocompatibility of CDs, and further try to apply the CDs for cell imaging, the cytotoxicity of CDs in A549 cells was studied by MTT measurements [54]. As depicted in Fig. 8a, the CDs had no obvious cell cytotoxic effect on cells in the concentration range from 25 to 1000 μg mL-1. Because the CDs showed low toxicity toward the cells, they can be used for cell imaging and biomedical applications. Figure 8b–d shows that before addition of the CDs, the cells did not fluoresce, whereas after 3 h of incubation with the CDs, the cytoplasm of the cells showed bright blue fluorescence. This change can be ascribed to the CDs easily passing through the cell membrane and entering the cytoplasm. However, as shown in Fig. 8e, the blue fluorescence of cells became very weak after incubation with 0.4 mM Fe3+ for 30 min. The fluorescence was partially recovered after treatment with 0.4 mM AA for 30 min (Fig. 8f). Hence, the CDs can be used as a fluorescent on–off sensor for the qualitative analysis of Fe3+, and the CD–Fe3+ system can be applied as a fluorescent off–on sensor for the qualitative analysis of AA in biological systems.

Fig. 8
figure 8

a Effects of carbon dots (CDs) on the proliferation of A549 cells. b Images of A549 cells in bright field. c Fluorescence image of blank control (not incubated with CDs). d Fluorescence image of cells after incubation with CDs. e Fluorescence image of cells in the presence of CDs and Fe3+. f Fluorescence image after further incubation with ascorbic acid

Full size image

Conclusions

In summary, a green hydrothermal method was used to prepare CDs with high-intensity blue fluorescence by use of flowers of O. fragrans as green precursors. The CDs had high stability and good water solubility and were an effective fluorescent sensor for the direct detection of Fe3+ with a LOD of 5 nM. The satisfactory results for the detection of Fe3+ in real water samples and a human serum sample proved the practicability of the sensor. Moreover, on the basis of the selectivity and sensitivity of the CDs for Fe3+, the CD–Fe3+ system was further developed as a fluorescent off–on sensor for the detection of AA. The LOD of AA was 5 μM. In addition, the CDs can be used for the fluorescence imaging of A549 cells with low toxicity and good biocompatibility. Moreover, Fe3+ (AA) can also be qualitatively analyzed by monitoring of the changes in cell fluorescence. Therefore, the CDs have great potential for application in sensing, bioimaging, and even disease diagnosis in analytical and biological fields as an eco-friendly sensor. However, the QY of the CDs and the recovery effect on AA need to be further explored.