Characterizations of N,S-CDs
First, the morphology of the synthesized N,S-CDs was investigated via transmission electron microscopy (TEM). Figure 1a shows a TEM image in which the N,S-CDs have a near-spherical appearance and good dispersion. The particle size of N,S-CDs was mainly distributed from 2.5 to 5.5 nm, and the average size was approximately 3.9 nm (Fig. 1a, insert). The HRTEM image in the upper right panel of Fig. 1a shows that the N,S-CDs have high crystallinity and distinguishable lattice stripes with a spacing of 0.22 nm, matching the lattice constants in the (100) plane of graphitic carbon [23], which is similar to previous studies on carbon dots [24].
FTIR and XPS were then carried out to analyze the structures and compositions of N,S-CDs. As illustrated in Fig. 1b, the absorption peak at 3435.4 cm− 1 is attributed to the stretching vibrations of O–H and N–H, and the three cusp peaks at 1645.5, 1597.7, 1399.1 cm− 1 are attributed to the stretching vibrations of C = O, C = C, and –C–N, respectively. The peak at 530.5 cm− 1 is attributed to the bending vibration of C–S [22]. In addition, the XPS full spectrum in Fig. 1c shows that the N,S-CDs were mainly composed of C (283.5 eV), N (398.5 eV), O (530.4 eV), and S (162.7 eV), and the corresponding atomic contents were 70.5, 3.5, 22.9, and 3.1%, respectively. Furthermore, Fig. 1d shows the high-resolution XPS peaks for C 1s, which indicates the existence of C = O (288.0 eV), C–N (286.1 eV), and C–C (284.6 eV). The N 1s XPS spectrum indicates the presence of C–N–C (399.6 eV) (Fig. 1e). The high-resolution S 2p XPS spectrum shows a peak at 168.6 eV (Fig. 1f), which represents –C–SO3– bonds [25]. The XPS results indicate that the synthesized CDs were successfully doped with N and S elements.
Based on the above results, the surface of N,S-CDs contains hydrophilic groups, such as –NH, –COOH, and –OH, which not only allow the N,S-CDs to be easily modified on the surface but also increase their hydrophilicity and stability in aqueous solution.
Optical properties of N,S-CDs
The optical properties of N,S-CDs were characterized by fluorescence and UV–vis spectroscopies. As shown in Fig. 2a, a strong excitation peak was present at 408 nm. The UV–vis spectrum of N,S-CDs shows two absorption peaks at 218 and 285 nm, which were ascribed to the π–π* transition of the aromatic sp2 domains including C = C and C = N bonds and the n–π* transition of the C = O bond [26], respectively. The aqueous solution of N,S-CDs was light yellow under natural light and exhibited bright blue–green fluorescence under excitation at 365 nm (Fig. 2a, inset). Figure 2b shows the fluorescence spectra of N,S-CDs with excitation at 290–450 nm. When the excitation wavelength was gradually increased from 290 to 450 nm, two emission peaks at 390 and 500 nm were obtained at excitation wavelengths of 310 and 410 nm.
Due to their dual-wavelength emission properties, N,S-CDs can be developed as ratiometric probes. Figure 2c shows the fluorescence profiles of the N,S-CDs before and after adding Hg2+. Notably, the fluorescence at 390 nm remained stable, while the emission peak at 500 nm decreased when Hg2+ was added. In addition, Fig. 2d shows significant quenching of the green fluorescence of N,S-CDs with Hg2+, indicating that the probe could be used for Hg2+ detection. Thus, the stable fluorescence emission of 390 nm could be applied as a reference for the variable Hg2+-controlled fluorescence intensity of 500 nm as a ratiometric strategy.
Then, the effects of pH, ionic strength, and time on the fluorescence stability of N,S-CDs were evaluated. Fig. S1 shows a negligible change in the ratio of F390/F500 when the pH was between 4 and 12. However, the emission intensity of F390 remained stable, whereas that of F500 decreased as that the ratio increased when the pH was between 2 and 3. The pH responsiveness of the fluorescence of N,S-CDs likely results from the transition between C–NH2 under acidic conditions and C = NH under alkaline conditions [22]. As shown in Fig. S2, the ratio was constant at different concentrations of NaCl solution. In addition, the F390/F500 ratio changed negligibly after 1 h of continuous laser irradiation, indicating a high resistance to photobleaching (Fig. S3). In addition, Fig. S4 shows that the fluorescence intensity of the N,S-CDs did not decrease considerably in irregular measurements over 60 days, indicating that these CDs have good fluorescence stability. These results indicate that the synthesized N,S-CDs are stable in an acidic environment, with excellent salt resistance, photobleaching resistance, and fluorescence stability, and are a luminescent material that has stable fluorescence performance, with the potential for practical application.
Furthermore, the biocompatibility of the prepared N,S-CDs was assessed, as shown in Fig. S5. After a 24-h incubation period with a series of concentrations of N,S-CDs, the viability of 4T-1 cells remained at an impressive 92%, even in the presence of 100 mg/mL N,S-CDs. This result highlighted the low cytotoxicity and excellent biocompatibility of N,S-CDs, suggesting promising application prospects in the biomedical field.
Optimization of experimental conditions
Several experimental conditions were optimized to improve the sensitivity of the ratiometric method for Hg2+ detection. As shown in Fig. S6, the F390/F500 ratio of the two emission peaks of the N,S-CDs did not change considerably when the pH was again kept between 4 and 12. The fluorescence ratio appeared to change after adding Hg2+ under the same conditions, and the ratio was maximized when the pH was 4. Therefore, a pH of 4 was selected as optimal for the ratiometric fluorescence detection of Hg2+. In addition, the reaction time was investigated. As shown in Fig. S7a, the fluorescence intensity at 390 nm remained constant, and the fluorescence at 500 nm decreased rapidly after 1 min of reaction. The F390/F500 ratio increased rapidly and reached an equilibrium minimum within 1 min of adding Hg2+ (Fig. S7b), suggesting rapid adsorption and quenching of Hg2+ by N,S-CDs.
Ratiometric Hg2+ sensing
Under optimized conditions, the quantitative detection of Hg2+ was assessed, revealing a distinctive pattern in Fig. 3a where the fluorescence intensity of F500 decreased upon adding Hg2+, while F390 remained stable. As shown in Fig. 3b, the F390/F500 ratio steadily increased with increasing Hg2+ concentration from 0 to 100 µM. Notably, Fig. 3c shows a strong linear correlation between the F390/F500 ratio and Hg2+ concentration within the 2–30 µM range, with a linear regression equation of y = 0.9242 + 0.004x and an R2 value of 0.9954. The limit of detection (LOD) for Hg2+ using N,S-CDs, calculated based on the 3σ/k principle, was determined to be 0.21 µM, where σ represents the standard deviation from 12 blank solution repetitions, and k is the slope of the calibration curve. Compared to alternative Hg2+ detection methods (Table S1), the synthesized N,S-CDs exhibited commendable sensitivity and performance, further enhanced by their straightforward synthesis process and inexpensive raw materials. These results confirm that this method can effectively simplify the synthesis of ratiometric N,S-CDs with dual-channel detection capability.
Selectivity for Hg2+ detection
Specificity is one of the most pivotal performance indicators for evaluating sensing systems, as it gauges the probe’s ability to respond to a target analyte in a complex matrix selectively. In the context of fluorescent probes, the ideal scenario involves a precise and exclusive response to a specific metal ion. However, the reality can be more nuanced, with multiple ions potentially eliciting changes in the signal, necessitating rigorous testing to affirm the system's specificity. As demonstrated in Fig. 3d, a stark contrast emerged when an array of metal ions was introduced to the N,S-CDs. Among them, only Hg2+ triggered a pronounced augmentation in the fluorescence intensity ratio of 390 to 500 nm (F390/F500). This unambiguous enhancement served as compelling evidence of the ratiometric fluorescent probe’s specificity towards Hg2+, highlighting its ability to discern and respond solely to the targeted ion in a myriad of potential distractors. Moreover, the robustness of this detection method was validated by its resilience against interference from co-existing ions. Despite these potentially confounding factors, the fluorescence-based detection of Hg2+ persisted and was unaffected, underscoring the exceptional anti-interference capabilities of the established ratio detection strategy. This feature is particularly advantageous in real-world applications, where samples are often contaminated with multiple ions, making developing such highly specific and interference-resistant sensors all the more crucial.
Mechanism of Hg2+ detection
To verify the mechanism of the fluorescence quenching of N,S-CDs by Hg2+, we analyzed the UV absorption spectra, hydrodynamic diameters, and fluorescence lifetimes of N,S-CDs in the absence and presence of Hg2+. As shown in Fig. 4a, the fluorescence lifetimes of N,S-CDs in the presence and absence of Hg2+ were 2.47 ns and 1.99 ns, respectively, with little change, indicating that the effect of Hg2+ on the N,S-CDs was not due to dynamic fluorescence quenching [27, 28]. The absorption spectra showed that in the presence of Hg2+, the absorption peak of N,S-CDs at 285 nm shifted to a shoulder peak at 280 nm, probably due to the binding of N,S-CDs to Hg2+ and the transfer of electrons (Fig. 4b). To further analyze the fluorescent quenching mechanism, we measured the hydrodynamic diameter of the samples before and after Hg2+ detection using the N,S-CDs, and Fig. 4c shows that the average hydrodynamic diameter of N,S-CDs increased from 3.5 to 747.1 nm after adding Hg2+. This indicated that N,S-CDs were bound to Hg2+ and generated a non-fluorescent complex [29], which increased the ephemeral diameter. Based on those results, the detection of N,S-CDs on Hg2+ involved a static quenching mechanism, which is in line with our conclusions in this study.
Application in real samples
Tests were conducted on actual water samples to evaluate the practical applicability of N,S-CDs. As presented in Table S2, satisfactory recoveries ranging from 94.5 to 107.8% were achieved, accompanied by relative standard deviations (RSD) of less than 3.3%. These results demonstrated that the N,S-CDs-based ratiometric sensor provides a reliable and viable method for detecting Hg2+ in actual samples with high precision and excellent selectivity.