Introduction

Graphene quantum dots (GQDs), as a recently discovered member of luminescent carbon nanomaterials, have attracted considerable interest in various areas of analytical chemistry. They are graphene fragments with a size distribution below 100 nm, which exhibit pronounced exciton confinement and quantum size effects. As a result, GQDs have non-zero band gap and show intense and stable photoluminescence. Furthermore, GQDs have some other interesting properties such as low toxicity, excellent solubility in various solvents, chemical inertia, and high electrical and thermal conductivity. It should be mentioned that unlike carbon dots which mainly contain amorphous carbon, GQDs have clear graphene lattices inside the dots. Moreover, compared to carbon dots, GQDs have higher surface areas, larger diameters and better surface grafting properties [1, 2].

Chlorine has been extensively used as disinfectants in treating drinking and swimming pool water. Chlorine is usually applied to water in its molecular or hypochlorite form. It initially undergoes hydrolysis to form free chlorine consisting of aqueous molecular chlorine, hypochlorous acid, and hypochlorite ion. High excess of active chlorine in water is harmful to human health and must be avoided [3, 4]. Therefore the monitoring of free chlorine concentration in water samples has appeared of great importance and various analytical methods including chromatography [5, 6], electrochemistry [7, 8], spectrophotometry [9, 10], fluorescence [11, 12] and chemiluminescence (CL) [13–17] have been reported for this purpose.

In recent years, due to the unique physicochemical properties of nanomaterials, many CL systems based on different kinds of nanomaterials have been developed [18, 19]. More recently, the application of luminescent carbon nanomaterials in CL [20–32] and electrochemiluminescence [33] reactions has been investigated. Carbon nanomaterials could participate in a CL reaction as emitting species, after direct oxidation [20–26]; as catalysts of reactions involving other luminophores [27–29]; or as emitters after CL energy transfer [24, 25, 30–32]. Hypochlorite has been used as an oxidant in various chemiluminescent reactions [34] but its effect on the luminescent carbon nanomaterials has not yet been studied.

In this work, we synthesized the GQDs by simple pyrolysis of glucose as a carbon source. It was found that ClO− as an oxidant is able to directly elicit a CL from the prepared GQDs. The CL intensity of the GQD/ ClO− reaction is greatly enhanced by addition of cetyl trimethyl ammonium bromide (CTAB). On the basis of this CL system, a simple and sensitive method was established for the determination of ClO− in tap and pool water samples.

Experimental

Apparatus

The CL signals were monitored by LUMAT LB 9507 chemiluminometer (Berthold; www.berthold.com). The luminometer was equipped with an automatic injector functioning on the basis of JET injection technology® and being used for injecting the oxidant. The fluorescence spectra were recorded by RF-540 spectrofluorimeter (Shimadzu, Japan; www.shimadzu.com). CL spectrum was also recorded by this instrument using flow mode with the excitation light source being turned off. UV–vis spectra were recorded on a Cary-100 Spectrophotometer (Varian; www.varianinc.com). The size and shape of GQDs were characterized by transmission electron microscopy (TEM, Leo 906, Zeiss, Germany). The Fourier transform infrared (FTIR) spectrum was recorded on a FT-IR spectrophotometer (Tensor 27, Bruker; www.bruker.com). X-ray powder diffraction (XRD) pattern was measured by a Siemens D 500 instrument (Germany). Raman spectrum was recorded using an Almega Thermo Nicolet Dispersive Raman Spectrometer with a 532 nm laser.

Reagents

All reagents were of analytical-reagent grade. Doubly distilled deionized water (obtained from Ghazi Serum Co. Tabriz, Iran) was used throughout the experiment. Glucose, di-sodium phosphate and CTAB were purchased from Merck (Darmstadt, Germany, www.merck-chemicals.com). A stock standard solution of 0.1 M of NaClO was daily prepared by diluting the appropriate amount of 5 % NaClO solution (standardized by iodometric titration) in 25 mL of deionized water. Working solutions were prepared by appropriate dilution of the stock solution.

Synthesis of GQDs

GQDs were synthesized by pyrolyzing glucose. In a typical procedure, 6.0 g of glucose was put into a glass beaker and heated on a hot plate at 180 °C until the glucose powder was melted. After about 1.5 min, the color of liquid was changed from colorless to orange. This color change implies the formation of GQDs. At this stage, the heating was stopped and 25 mL of deionized water was added into the orange liquid with continuous and vigorous stirring. The obtained GQD solution was stable for more than 1 month in the refrigerator at 4 °C.

General procedure for CL monitoring

CL measurements were conducted on the batch condition with the automatic injection system, which provides high speed and repeatable and efficient mixing of the reagents. An amount of 400 μL of GQDs, 100 μL of phosphate buffer (0.1 M, pH = 8) and 100 μL of CTAB solution (0.25 M) were added into the cell (a tube with 75 mm height and 12 mm diameter) and the final volume was reached to 700 μL with deionized water. After injection of 300 μL of standard or sample ClO− solution by the automatic injector, the CL signal was automatically monitored versus time. Maximum CL intensity was used as the analytical signal.

Procedure for water samples

No pretreatment was necessary for tap water. Pool water was diluted 5-fold. 300 μL of water samples was taken for analysis according to the general procedure.

Results and discussion

Synthesis and characterization of GQDs

In this work, a simple carbonization method was used for synthesis of GQDs using glucose as a carbon source. Glucose has previously been used for preparation of GQDs by microwave-assisted hydrothermal method [35]. However, we found that its direct pyrolysis is a much simpler procedure for obtaining GQDs. Various amounts of glucose was tested for pyrolysis and found that when 6 g of glucose was pyrolyzed and its volume was reached to 25 mL, the highest fluorescence was obtained. Also the influence of time and temperature of heating was studied and found that 1.5 min heating at 180 °C is the best conditions. At longer and shorter times, the fluorescence of obtained GQDs is decreased. The TEM image of GQDs synthesized by the suggested procedure shows that they are of spherical shape and nearly monodisperse with size distribution of 20 ± 5 nm (Fig. 1a). As shown in Fig. 1b, the XRD pattern of GQDs exhibits a broad (002) peak around 20° (0.44 nm), corresponding to the graphite structure. This is in a good agreement with the reported values for GQDs prepared by hydrothermal method [35]. FTIR spectroscopy was applied to characterize the functional groups of the prepared GQDs (Fig. 1c). The peaks at 1,643 and 1,710 cm−1 are attributed to C = C and C = O stretching vibrations, respectively and the peaks at about 1,364 and 2,894 cm−1 to the stretching vibration of C–H. The broad intense peak at 3,450 cm−1 is assigned to O-H stretching vibrations. Additionally, the peak at 1,076 cm−1 reveals the existence of C-O bonding.

Fig. 1
figure 1

a TEM image, b XRD pattern, c FT-IR spectrum and d Raman spectrum of the prepared GQDs

Full size image

Raman spectrum of the synthesized GQDs (Fig. 1d) exhibits two characteristic peaks around 1,370 and 1,605 cm−1, corresponding to the D and G band of graphene, respectively. The D band arising from a breathing mode of k-point phonons of A1g symmetry and a G corresponds to the first-order scattering of the E2g vibration mode. The prepared GQDs exhibit small ID/IG value of about 0.5, which indicates the high quality of crystalline graphitic system [1].

The UV–vis absorption spectrum of GQDs (Fig. 2a) exhibits typical absorption of graphene derivatives in the UV region with a peak at ~226 nm which was assigned to the π–π* transition of C = C and a peak at ~282 nm corresponds to the n–π* transition of C = O. The fluorescence emission of the GQDs was also investigated. The maximum fluorescence wavelength was ∼475 nm, which obtained with an excitation wavelength of 400 nm. As shown in Fig. 2b the GQDs exhibit an excitation-dependent photoluminescence behavior i.e. when the excitation wavelength changes from 350 to 450 nm, the emission peak shifts to longer wavelengths. The obtained results were similar to those previously reported in literature and confirm the proper synthesis of GQDs [35].

Fig. 2
figure 2

a UV–vis absorption spectrum of GQDs. b Fluorescence spectra for GQDs excited at 350–450 nm wavelength range, with increments of 25 nm

Full size image

Kinetic profile of CL

Our preliminary experiments indicated that ClO− is able to directly elicit a CL from GQDs, though the CL intensity was weak (Fig. 3a). However, in CTAB micellar medium CL is greatly enhanced (about 18-fold) (Fig. 3b). The control experiments showed that the coexisting substrates have low or no CL emission upon reaction with ClO−. The kinetic profile of GQDs/ClO− CL system (Fig. 3) showed that the CL reaction was very fast and the intensity reached a maximum value within 1 s after injection of the oxidant.

Fig. 3
figure 3

Kinetic curve for GQDs/ClO− CL system (a) alone, (b) in the presence of CTAB. Conditions: GQDs, 400 μL, phosphate buffer (0.01 M, pH = 8), CTAB (0.025 M) and ClO− (10−3 M)

Full size image

Possible mechanism of CL reaction

In order to explain the CL reaction mechanism and identify the emitting species, following experiments were performed. Firstly, the CL spectrum of GQDs/ClO−/CTAB system was recorded using a spectrofluorimeter in flow mode. The obtained spectrum (Fig. 4a) indicated that there was an emission band in the range of 400–600 nm with a maximum at ∼510 nm. This CL spectrum is similar to the GQD fluorescence spectrum (Fig. 4b). Therefore, the CL can be attributed to the excited GQDs.

Fig. 4
figure 4

a CL spectrum of GQDs/ClO− /CTAB system, obtained with continuous flow of reagents: GQD solution, phosphate buffer (0.02 M, pH = 8) and CTAB (0.05 M) in one line and ClO− (2 × 10−3 M) in other line; b fluorescence spectrum of GQDs, at λex = 450 nm

Full size image

Moreover, the fluorescence spectrum of GQDs after mixing with ClO− was recorded. According to the results, after addition of ClO− to GQD solution, the fluorescence intensity at 510 nm decreased. So it can be concluded that a chemical reaction occurred between GQDs and ClO−. Finally the effect of some radical scavengers including NaN3, thiourea, quercetin and ascorbic acid on the CL intensity of GQD/ClO− /CTAB system was investigated. The results of this study together with possible reaction pathways are described in Electronic Supplementary Material (ESM).

Effects of variables on the CL intensity

According to the above-mentioned findings, a new CL system could be developed for the determination of ClO−. To establish the optimal conditions for the analysis of ClO−, the effects of several analytical variables such as the pH value, buffer and surfactant concentration and the amount of GQD solution on the CL intensity were investigated.

The CL intensity is highly sensitive to the pH value. In order to obtain the suitable medium for CL reaction, the pH of GQD/ClO− /CTAB reaction mixture was varied in the range of 4–11. The results in Fig. S1a (ESM) indicated that the CL intensity increased with pH and reached a maximum around pH 8. This may be attributed to the fact that the production of peroxyacids from the surface carboxylate groups of the GQDs would be most efficient at this pH where the parent carboxylic acid is fully dissociated while the chlorine is still in the form of hypochlorous acid [13].

Moreover, the effect of phosphate buffer concentration on the CL signal was studied and the results are shown in Fig. S1b (ESM). As seen, the CL intensity increased by increasing phosphate buffer concentration up to 0.01 M, and then remained constant.

Surfactants at concentrations above its critical micellar concentration (CMC) usually enhance the CL emission in many reactions [36]. Therefore, the effects of three kinds of surfactants including cationic (CTAB), anionic (sodium dodecyl sulfate, SDS), and nonionic surfactants (Triton X-100), on the GQD/ClO− CL system were investigated. The CL signal was remarkably enhanced in the presence of CTAB. However, SDS and Triton X-100 did not show significant influence on the GQD/ ClO− system. Therefore, CTAB was chosen as a suitable surfactant for the CL enhancement. The maximum CL intensity was obtained at 2.5 × 10−2 M CTAB concentration (Fig. S1c, ESM). Two phenomena may be involved in the enhancement effect of CTAB micelles on the CL system. First, the micelle surfaces which are rich in positive charge can attract and concentrate the superoxide anion radicals to providing more reaction opportunities. Second, the micelles protect the excited state GQDs and decrease the non-radiative transitions which lead to a significant increase in the CL intensity.

Finally, the effect of the amount of GQD solution on the CL intensity was investigated in the range of 50–700 μL of stock solution. As shown in Fig. S1d (ESM), the CL intensity increased by increasing GQD volume up to 400 μL and then remained relatively constant. Therefore, this amount was selected as the optimum value for experiments.

Analytical figures of merit

Under the optimized experimental conditions, the CL intensity was proportionally increased with concentration of free chlorine (Fig. 5a). The linear range was 5 × 10−7 to 1 × 10−3 M with a detection limit of (3 s) of 3 × 10−7 M (Fig. 5b). The regression equation was I = (7 × 107) C +727, R2 = 0.9999, where I is CL intensity and C is concentration of ClO− (M). The relative standard deviation (RSD) was 1.6 % for five determinations of 1 × 10−5 M ClO−. Table 1 compares the developed method with some other reported CL methods for the ClO− determination. As can be seen, the analytical performances of this method are better than or comparable with those of other methods. Moreover, our method is quite simpler and less expensive than most of other methods. Finally, it should be mentioned that GQDs/ClO− as a novel CL system may have other potential analytical applications.

Fig. 5
figure 5

a CL signals and b calibration curve for ClO− in the range from 5 × 10−7 to 10−3 M. Conditions: GQD (400 μL), phosphate buffer (0.01 M, pH = 8), CTAB (0.025 M)

Full size image
Table 1 Comparison of the developed CL method for the determination ClO− with some previously published CL methods
Full size table

Interferences

In order to evaluate the selectivity of the proposed method, the effect of some metal ions and anions commonly present in water samples on the determination of 1.0 × 10−5 M ClO− was investigated. The tolerable concentrations for interfering species in relative error of <5 % were summarized in Table 2. These results demonstrate that the method possesses a good selectivity for the determination ClO− in water samples.

Table 2 Tolerance limit of interferences on the determination of 1 × 10−5 M ClO−
Full size table

Analysis of water samples

The developed method was applied to the determination of free chlorine in tap and pool water samples. The obtained results are shown in Table 3. In order to confirm the accuracy of method, the results were compared with those obtained by N,N-diethyldiphenylenediamine (DPD) standard method. According to statistical analysis by Student t-test, there is no significant difference between the results of two methods (Table 3). In addition, the samples were spiked with known amounts of ClO−, and then they were analyzed according to the general procedure. The obtained results (Table 3) indicated that there are no significant differences between the added and found values.

Table 3 Determination of ClO− in tap and pool water samples
Full size table

Conclusion

In this work, we demonstrated that GQDs obtained from pyrolysis of glucose could be directly oxidized by ClO− to give rise to the CL emission. The CL was supposed to be originated from the processes of electron transfer annihilation and resonance energy transfer. Additionally, it was found that CTAB could remarkably enhance the CL intensity of GQD/ClO− system via concentrating of the reactive anion radicals and providing a protective environment for excited state species. The GQD/ClO− /CTAB CL system was applied to the determination of ClO− in tap and pool water samples with good accuracy and high precision. Considering the simplicity and reliability of this CL system, it can be used in other analytical applications.