Ratiometric fluorescence nanoprobe based on nitrogen-doped carbon dots for Cu²⁺ and Fe³⁺ detection

Abstract

Heavy metal ions pollution in environmental waters has an increasing impact on human health. As two common metal ions, copper ions (Cu²⁺) and ferric ions (Fe³⁺) widely exist in nature and play a vital role in life process. Therefore, it is significant to design sensitive and simple detection approaches for Cu²⁺ and Fe³⁺. In our work, the ratiometric fluorescence analysis method (denoted as N-CDs/OPD) was established for Cu²⁺ and Fe³⁺ detection. The N-CDs exhibited a Cu²⁺ and Fe³⁺ fluorescence quenching response properties. The o-phenylenediamine (OPD) may be oxidized to 2,3-diaminophenazine (DAP) by Cu²⁺ and Fe³⁺. With addition of Cu²⁺ or Fe³⁺, the fluorescence of N-CDs (436 nm) was quenched and a new peak at 556 nm (DAP) appeared, which realized fluorescent ratiometric detection of Cu²⁺ and Fe³⁺. The Cu²⁺ concentration shows a good linear correlation versus fluorescence ratio (F₄₃₆/F₅₅₆) in the range of 10 to 30 µM (R² = 0.9981) with detection limit (LOD) of 0.86 µM. In addition, a good linear relationship between fluorescence ratio (F₄₃₆/F₅₅₆) and Fe³⁺ concentration in the range of 20 to 80 µM (R² = 0.9880) with LOD of 7.12 µM. This nanoprobe realizes the detection of authentic samples successfully, which is expected to serve as a testing kit for analysis in water samples.

Introduction

With the development of modern industry, heavy metal ions pollution in environmental waters has an increasing impact on animals, plants and human health. As two common metal ions, copper ions (Cu²⁺) and ferric ions (Fe³⁺) widely exist in nature and play a significant role in life process^1,2,3. The formation of intracellular oxidoreductase in the body requires sufficient amounts of copper⁴. However, excessive Cu²⁺ intake has been linked to several physical disorders, including Wilson’s disease and Alzheimer’s disease^5,6. The irreversible liver and kidney damage may be caused by long-term accumulation of copper⁷. Unfortunately, due to widespread use of Cu²⁺ in industrial production and daily life, which has been enrolled in the EPA’s pollutant list of toxic metal species⁸. Fe³⁺ plays critical roles in cellular metabolism, electron transfer, enzyme catalysis, oxygen hemoglobin production⁹. Both of its deficiency or excess can cause Huntington¹⁰ or Parkinson’s disease¹¹. Therefore, analysis of heavy metal ions, especially Cu²⁺ and Fe³⁺, is of great significance for environmental protection and clinical research.

Different analytical methods have been proposed for heavy metal ions detection, including atomic absorption spectrometry¹², electrochemical analysis¹³, inductively coupled plasma mass spectroscopy¹⁴, colorimetry¹⁵ and fluorometry^16,17. Among those ways, fluorescent probes show benefits of simplicity, high sensitivity, low cost and background signals^18,19. Traditional fluorescence methods are based on intensity of individual emission peak to quantitatively analyze ions, which is susceptible to environmental interferences and results in inaccurate detection^20,21,22. To ensure the accurate detection, ratiometric fluorescence probes have self-calibration capabilities. Two emission bands can act as internal reference, removing a majority of interfering factors, including probe concentration, environmental conditions photobleaching and instrumental efficiency^23,24. However, the largest number of ratiometric fluorescent probes usually focus on detecting a single heavy metal ion specifically. The increased and complex detection requirements of multiple heavy metal ions are urgently needed.

Among fluorescent probes, various fluorescent nanomaterials attract extensive interest, such as metal nanoclusters, quantum dots (QDs) and carbon dots (CDs), etc²⁵. The CDs are new carbon-based nanomaterials, the size of which less than 10 nm with spherical particles²⁶. Compared with QDs, CDs have better performance, including low toxicity, low cost, simple synthesis, photo stability, adjustable optical properties, good biocompatibility and water solubility²⁷. Herein, we innovatively developed a N-doped carbon dots (N-CDs)-based ratiometric fluorescent nanoprobe (N-CDs/OPD) for detection of Cu²⁺ and Fe³⁺ in aqueous solution (Fig. 1). The N-CDs were synthesized via one-step solvothermal treatment of methionine and formamide. Due to inter filter effect (IFE) between metal ion and carbon dots, the N-CDs exhibited a Cu²⁺ and Fe³⁺ fluorescence quenching response properties. After Cu²⁺ and Fe³⁺ mixed with this probe, the fluorescence of N-CDs (436 nm) was quenched and a new peak at 556 nm (DAP) appeared. Cu²⁺/Fe³⁺ would quench CDs’ fluorescence and oxidize o-phenylenediamine (OPD) to 2,3-diaminophenazine (DAP). This strategy realized the fluorescence ratiometric detection of Cu²⁺ and Fe³⁺ in authentic samples successfully, which broadens the potential application of carbon dots.

Fig. 1

Schematic illustration of ratiometric fluorescent nanoprobe (N-CDs/OPD) for Cu²⁺ and Fe³⁺ detection.

Experimental section

Materials and apparatus

o-phenylenediamine (OPD), methionine and formamide and were obtained from Shanghai Aladdin Biochemical Techonology Co., Ltd. Phosphoric acid (H₃PO₄) and acetic acid (HAc) were purchased from Sinopharm Chemical Reagent Co. Cu²⁺ standard solution, Fe³⁺ standard solution, boric acid (H₃BO₃), NaOH, cobalt chloride hexahydrate (CoCl₂·6H₂O), ferrous sulfate heptahydrate (FeSO₄·7H₂O), nickel (II) sulfate hexahydrate (NiSO₄·6H₂O), barium chloride dihydrate (BaCl₂·2H₂O), chromium (III) nitrate nonahydrate (Cr(NO₃)₃·2H₂O), KCl, CaCl₂, NaCl, MgCl₂, MnCl₂ and ZnCl₂ were acquired from Shanghai Macklin Biochemical Technology Co., Ltd. The section of “apparatus” was in “supplementary information”.

Synthesis of N-doped carbon dots

N-doped carbon dots (N-CDs) were synthesized by solvothermal synthesis approach with precursors of methionine and formamide. 0.2093 g of methionine was dissolved in 12.5 mL of formamide with ultrasound for 30 min. This mixed solution was added into a Teflon-lined stainless steel autoclave (50 mL) for 1 h with 180 °C. After the system cooling to room temperature, the light brown supernatant was filtered by filter membrane of 0.22 μm to remove large particles. Then, the solution was purified with dialysis bag (molecular weight cut-off of 1000 Da) for 8 h. And finally, the purified product stored in a refrigerator at 4 °C for further use.

Fluorescent detection of Cu2+ and Fe3+

The stock solution concentrations of Cu²⁺ and Fe³⁺ standard solution was 10 mM, respectively. 150 µL of Britton-Robinson buffer [pH 5.3(Cu²⁺) 5.8(Fe³⁺), 40 mM], ultrapure water, 150 µL of o-phenylenediamine solution (1 mM), various concentrations of Cu²⁺/Fe³⁺ solution were blended and diluted to 1.4 mL. After 60 min/10 min still standing at room temperature, 100 µL of N-doped carbon dots was added to above system. Fluorescence measurements were carried out under the excitation wavelength at 360 nm. It was used as the ordinate that the ratio of fluorescence intensity at 436 nm and 556 nm wavelengths (F₄₃₆/F₅₅₆). Concentrations of Cu²⁺/Fe³⁺ were regarded as horizontal ordinate and drawing standard curve for Cu²⁺ and Fe³⁺ detection.

In order to realize the Cu²⁺ and Fe³⁺ detection in tap water, samples were spiked with various concentrations of Cu²⁺ (20 µM, 30 µM) and Fe³⁺ (40 µM, 60 µM). After that, the samples were measured by this probe using the same experimental procedures as above. These fluorescent spectra were measured and concentrations of spiked samples were calculated by corresponding standard curve.

Selectivity of the proposed nanoprobe

150 µL of Britton-Robison buffer (pH 5.3, 40 mM), 150 µL of o-phenylenediamine solution (1 mM), the same concentration of Cu²⁺/Fe³⁺ standard solution, or pure water sample as blank sample, various metal interference ions (K⁺, Na⁺, Ca²⁺, Mg²⁺, Ba²⁺, Zn²⁺, Co²⁺, Mn²⁺, Fe²⁺, Ni²⁺ and Cr³⁺) mixed well and diluted to 1.4 mL. 100 µL of N-CDs were added to the above mixture after standing for 60 min. These fluorescent spectra were collected under 360 nm excitation wavelength.

Results and discussion

Characterization and optical properties of N-doped carbon dots

The N-doped carbon dots (N-CDs) were synthesized by solvothermal method with a simple procedure. As depicted in Fig. 2A, the TEM image indicates N-CDs have a monodisperse spherical structure with average diameter of 2 nm. In order to certify element components of N-CDs, XPS measurement was performed. The XPS full spectrum of N-CDs reflects five characteristic peaks, which represent C 1s, N 1s, O 1s, S 2s and S 2p, respectively (Fig. 2B). Atomic percentages of C, N, O and S were 52.23%, 8.06%, 31.01% and 8.71% by calculation, which showing the major components of N-CDs and doped by S and N. The high-resolution XPS spectra of N-CDs (C 1s, N 1s, O 1s and S 2p) are displayed in Fig. 2C–F. As shown in Fig. 2C, the high-resolution XPS spectrum of C 1s was deconvolved into two characteristic peaks, which were attributed to C = O (288.5 eV) and C-O/C = C (285.6 eV) bonds. High-resolution spectra of N 1s (Fig. 2D) displayed three kinds of N: pyrrole N (401.1 eV), amino N (400.3 eV) and pyridine N (399.3 eV)^28,29. The high-resolution XPS spectra of O 1s (Fig. 2E) locates peaks at 532.1 eV and 533.0 eV, which associate with C-O and C = O bonds. As demonstrated in Fig. 1F, the S 2p spectrum illustrates two major peaks of 164.8 eV and 163.7 eV, which correspond to S 2p1/2 and S 2p3/2 bonding³⁰.

Fig. 2

(A) TEM image of N-CDs, (B) XPS analysis of N-CDs, (C–F) High-resolution XPS spectra of C 1s, N 1s, O 1s and S 2p of N-CDs.

The FT-IR spectrum also shows the surface functional groups of N-CDs (Fig. 3A). A broad absorption bands at 3200–3600 cm^− 1 is attributed to stretching vibrations of O–H/N–H³¹. The peak located around 2876 cm^− 1 and 1301 cm^− 1 are regarded as stretching vibration of C–H^32,33. Moreover, these specific peaks located at 1049 cm^− 1, 1387 cm^− 1, 1602 cm^− 1 and 1665 cm^− 1 represent the stretching vibration of C–O, C–N, C = C and C = O groups, respectively^29,34,35,36. These results demonstrate the surface of carbon dots is decorated by oxygen containing and amino groups, and CDs doping with N are in accordance with FT-IR data. As demonstrated in Fig. 3B, these fluorescence emission spectra of N-CDs with different excitation wavelengths were measured. The emission peaks of N-CDs exhibit a slight red shift from 388 nm to 444 nm with the excitation wavelength from 310 to 370 nm, indicating the excitation-dependent emission of N-CDs. The property of excitation-dependence can be caused by surface heterogeneity of functional groups³⁷.

Fig. 3

(A) FT-IR spectrum of N-CDs, (B) Fluorescence emission spectra of N-CDs with different excitation wavelengths in range of 310 nm to 370 nm. (C) Fluorescence excitation (a) and emission (b) spectra of N-CDs, UV-vis spectra of Cu²⁺(c) and Fe³⁺(d). (D) Fluorescence emission spectra of N-CDs mixed with different metal ions (from top to bottom: Zn²⁺, Mn²⁺, Fe²⁺, Mg²⁺, Na⁺, Ca²⁺, K⁺, blank, Co²⁺, Fe³⁺, Cu²⁺).

When N-CDs were excited at 360 nm, a maximum fluorescence emission intensity was acquired at 436 nm (Fig. 3C). As depicted in inset photograph of Fig. 3C, the as-prepared N-CDs appear pale brown color under daylight and emit bright blue fluorescence with 365 nm UV light. The N-CDs may be acted as a fluorescence probe for Cu²⁺ and Fe³⁺ detection due to their optical property. To examine the selectivity of N-CDs for Cu²⁺ and Fe³⁺, different interfering metal ions were measured with N-CDs (Fig. 3D). The fluorescent intensity of N-CDs with relevant metal ions had little changes, while fluorescence quenching occurred with Cu²⁺ and Fe³⁺. These comparison data of fluorescence spectra demonstrate the excellent specificity of N-CDs.

Feasibility analysis and condition optimization

To testify the feasibility of this probe, the sensing mechanism need to be studied. The as-prepared N-CDs can be exploited as a probe to detect cupric ion and ferric ion. We speculated that the fluorescence quenching mechanism of N-CDs was static quenching, which can be strongly determined by fluorescence lifetime measurement. Static quenching usually does not change the fluorescence lifetime³⁸. As demonstrated in Fig. 4A, compared with lifetime of N-CDs in the presence of Cu²⁺ and Fe³⁺, the three lifetime curves almost overlap, which illustrates the static quenching effect. In addition, the absorption peak of Cu²⁺ and Fe³⁺ (Fig. 3C) and excitation peak of N-CDs overlap significantly. Based on the almost unaffected lifetime curve of N-CDs by metal ion, it is certified that the existence of inter filter effect (IFE) between metal ion and carbon dots.

Fig. 4

(A) Fluorescence lifetime curves of N-CDs in absence and presence of Cu²⁺ and Fe³⁺. (B) Fluorescence spectra of a: N-CDs, b: N-CDs + Cu²⁺, c: N-CDs + OPD + Cu²⁺, d: N-CDs + Fe³⁺, e: N-CDs + OPD + Fe³⁺. (C) UV–Vis absorption spectra of a: N-CDs, b: N-CDs + Cu²⁺, c: N-CDs + OPD + Cu²⁺, d: N-CDs + Fe³⁺, e: N-CDs + OPD + Fe³⁺.

To realize the ratiometric fluorescent detection of Cu²⁺ and Fe³⁺, we investigated the response mechanisms by the UV-vis absorption and fluorescence spectra. As shown in Fig. 4B, after Cu²⁺ and Fe³⁺ mixed with N-CDs/OPD system, the fluorescence of N-CDs (436 nm) was quenched and a new peak at 556 nm (DAP) appeared. Because Cu²⁺/Fe³⁺ would quench the fluorescence of N-CDs and oxidize OPD to DAP. In Fig. 4C, N-CDs exhibits an obvious UV-vis absorption peak at 319 nm. The N-CDs/OPD appeared new broad absorption peaks attributed to DAP in the presence of Cu²⁺ and Fe³⁺, which overlaps significantly with the excitation spectrum of N-CDs. These phenomena may be relative to Cu²⁺/Fe³⁺ react with OPD, causing the generation of new substance. These results can explain the further fluorescence quenching occurs for N-CDs induced by Cu²⁺ and Fe³⁺ after the addition of OPD in Fig. 4B. In terms of theoretical mechanisms, Cu²⁺ and Fe³⁺ both have a strong oxidation in acidic conditions, which could oxidize OPD to DAP in this sensing system. The relevant reaction process is shown in Fig. 1. Cu²⁺ oxidized OPD to DAP and it was reduced to Cu⁺/ Cu⁰. Fe³⁺ oxidized OPD to DAP and it was reduced to Fe²⁺.

Optimized experimental conditions (pH values, concentration of OPD and reaction time) were employed to obtain optimal response for Cu²⁺ and Fe³⁺ detection. As illustrated in Fig. 5A and C, as pH values increase, F₄₃₆/F₅₅₆ (the fluorescent peak of N-CDs at 436 nm and DAP at 556 nm) reduces and attains a stable condition. Therefore, pH 5.3 and pH 5.8 are selected as proper pH values for Cu²⁺ and Fe³⁺ detection respectively. To investigate whether pH has an effect on the fluorescence of N-CDs, the fluorescence response of N-CDs under different time with pH 7.0 and pH 5.3 were measured (Fig. S1). The N-CDs have an excellent fluorescent stability in weak acidic and neutral environment. Meanwhile, the OPD concentrations of Cu²⁺ and Fe³⁺ measurement have similar change trend (Figs. S2, S3). Thus, 100 µM of OPD was chosen for next test. Furthermore, the reaction time of Cu²⁺ and Fe³⁺ are shown in Fig. 5B and D. These values of F₄₃₆/F₅₅₆ for Cu²⁺ and Fe³⁺ remain steady after 60 min and 10 min, respectively. Hence, 60 min and 10 min are selected as optimal reaction time for Cu²⁺ and Fe³⁺ detection.

Fig. 5

The effect of (A) pH values and (B) reaction time with Cu²⁺ detection. The effect of (C) pH values and (D) reaction time with Fe³⁺ detection.

Fluorescent ratiometric detection of copper ions and ferric ions

The analytical performance of this probe for Cu²⁺ and Fe³⁺ was evaluated under optimized experimental conditions. As depicted in Fig. 6A, keeping the concentrations of N-CDs and OPD constant, the fluorescence emission peak at 436 nm (N-CDs) reduces whereas 556 nm (DAP) increases with enhancive amount of Cu²⁺ (0–100 µM). The dependence of Cu²⁺ concentration and fluorescence ratio (F₄₃₆/F₅₅₆) is shown in Fig. S4. When the Cu²⁺concentration enhanced, F₄₃₆/F₅₅₆ declined and achieved equilibrium. As displayed in Fig. 6B, the nice linear correlation is obtained between F₄₃₆/F₅₅₆ and Cu²⁺ concentration in range of 10 to 30 µM with R² = 0.9981. The limit of detection (LOD) was determined by 3Sd/K, where Sd represents standard deviation of blank samples and K represents slope of calibration line. The limit detection of Cu²⁺ was calculated to be 0.86 µM. Similarly, as shown in Fig. 6C, the fluorescent ratiometric response of this probe for Fe³⁺ shows same variation trend with rising Fe³⁺ concentrations (0–120 µM) with optimal experimental conditions. A steady reduction in fluorescence ratio is recorded with Fe³⁺ concentration increasing (Fig. S5). In Fig. 6D, a good linear relationship (R² = 0.9880) is obtained between fluorescence ratio (F₄₃₆/F₅₅₆) and Fe³⁺ concentration in range of 20 to 80 µM with LOD = 7.12 µM.

Fig. 6

(A) Changes in fluorescence intensity of nanoprobe (N-CDs/OPD) in different concentrations (0, 10, 15, 20, 25, 30, 40, 70 and 100 µM) of Cu²⁺. (B) The linear plot of F₄₃₆/F₅₅₆ against Cu²⁺ concentration (10, 15, 20, 25 and 30 µM). (C) Changes in fluorescence intensity of nanoprobe (N-CDs/OPD) in different concentrations (0, 10, 20, 40, 50, 60, 80, 100 and 120 µM) of Fe³⁺. (D) The linear plot of F₄₃₆/F₅₅₆ against concentration of Fe³⁺ (20, 40, 50, 60 and 80 µM).

In Fig. 7, to access the selectivity of Cu²⁺ and Fe³⁺ detection, we performed fluorescent response of N-CDs/OPD toward various potentially metal ions, including K⁺, Na⁺, Ca²⁺, Mg²⁺, Ba²⁺, Zn²⁺, Co²⁺, Mn²⁺, Fe²⁺, Ni²⁺ and Cr³⁺. Figure 6 shows that only Cu²⁺ and Fe³⁺ could lead to significant increase of fluorescence intensity ratio and the influence of other metal ions is negligible. To examine the precision and accuracy of this nanoprobe, tap water samples were spiked with Cu²⁺ and Fe³⁺ ions and analyzed with this method in triple replicates with same conditions. These relative standard deviations (RSD) and recoveries results for the spiked samples are listed in Table 1. These results indicate a good consistency between added and found concentrations of Cu²⁺ and Fe³⁺. These recoveries in tab water samples are between 98.2% and 109.3%, and the RSD values are less than 1.40%. In order to strengthen scientific significance of this probe, standard addition detection of Cu²⁺ and Fe³⁺ in domestic wastewater were performed in Table S1, which acquired relatively good results. Therefore, these results demonstrate the probe has a good accuracy and precision for the analysis of tracking Fe³⁺ and Cu²⁺ in real water. Compared with other analytical methods in Table S2, the detection performance of current method was superior or comparable to previously reported probes for Cu²⁺ and Fe³⁺ detection. The above results indicate that the simple and sensitive fluorescent ratiometric system has potential for monitoring the level of copper ions and ferric ions in water.

Fig. 7

Fluorescent intensity ratio of N-CDs/OPD in presence of different metal ions for Cu²⁺ and Fe³⁺ detection (interfering and target ions from left to right: blank, Cu²⁺, Fe³⁺, K⁺, Na⁺, Ca²⁺, Mg²⁺, Ba²⁺, Zn²⁺, Co²⁺, Mn²⁺, Fe²⁺, Ni²⁺ and Cr³⁺, concentrations: 100 µM).

Table 1 Results of the standard addition detection of Cu²⁺ and Fe³⁺ in tap water.

Conclusion

In conclusion, a cost-effective N-doped carbon dots-based ratiometric fluorescent nanoprobe (N-CDs/OPD) is designed for the detection of copper ions and ferric ions. The OPD could be oxidized to DAP by Cu²⁺ and Fe³⁺. The N-CDs have fluorescence quenching response characteristics of Cu²⁺ and Fe³⁺ because of inner filter effect. After addition of Cu²⁺ and Fe³⁺, the fluorescence of N-CDs (436 nm) is quenched and a new peak at 556 nm (DAP) appears, which realizes fluorescence ratiometric change. Therefore, this strategy enables the fluorescence ratiometric determination of Cu²⁺ and Fe³⁺ in authentic samples successfully, which has potential application for point-of-care testing.