III. RESULTS AND DISCUSSIONS

The synthesis of GQDs has been well developed through various approaches such as hydrothermal, electrochemical synthesis, chemical synthesis, chemical exfoliation from graphite fibers, and electron beam lithography. ^{Reference Yan, Cui, Li and Li30–Reference Lee, Kim, Park, Kim, Park, Kim, Bong, Kim, Chae, Jun, Hwang, Jung and Jeon39} Herein, GQDs were synthesized by chemical oxidation of pitch graphite fibers. ^{Reference Peng, Gao, Gupta, Liu, Romero-Aburto, Ge, Song, Alemany, Zhan, Gao, Vithayathil, Kaipparettu, Marti, Hayashi, Zhu and Ajayan37} Microstructure characterizations of the GQDs are summarized in Fig. 1. Transmission electron microscopy (TEM) image [Fig. 1(a)] shows the size of as-synthesized GQDs in the range of 2–5 nm. X-ray photoelectron spectroscopy (XPS) measurements were also performed to find out the composition of GQDs. As shown in Fig. 1(b), the C 1s peak at 284.3 eV and the O 1s peak at 531.9 eV for GQDs were obtained. The observation of C=C, C–O, C=O, and COOH from the high resolution C 1s XPS spectra indicated the presence of hydroxyl, carbonyl, and carboxylic acid groups in the GQDs.

FIG. 1.

Characterizations of as-synthesized GQDs. (a) A typical TEM image of GQDs. The scale bar is 50 nm. (b) The XPS of C 1s spectrum of GQDs. Inset is the XPS survey spectrum of GQDs. (c) Absorption curve of GQDs. (d) PL spectra of GQDs upon excitation from 420 to 480 nm.

Two absorption peaks at 230 nm and 270 nm [Fig. 1(c)] were observed for as-synthesized GQDs, which was similar to the previous report. ^{Reference Pan, Zhang, Li and Wu36,Reference Luo, Lu, Somers and Johnson40} The absorption at 230 nm originated from the n–n* transition. ^{Reference Pan, Zhang, Li and Wu36} The absorption feature of 270 nm, which was not observed in graphene and graphene oxide, could be attributed to n–π* transition of the C=O bond. Such an observed absorption may be explained by the large edge/bulk ratio of GQDs. ^{Reference Luo, Lu, Somers and Johnson40} The PL spectra of as-prepared GQDs excited at various wave lengths from 420 to 480 nm are shown in Fig. 1(d). Increasing the excitation wave length results in a red shift of the PL peak. The excitation wave length dependence of the emission wave length has been widely observed for carbon-based QDs including carbon dots and GQDs. ^{Reference Cao, Meziani, Sahu and Sun29–Reference Liu, Wu, Feng and Mullen38,Reference Sun, Zhou, Lin, Wang, Fernando, Pathak, Meziani, Harruff, Wang, Wang, Luo, Yang, Kose, Chen, Veca and Xie41,Reference Yang, Cao, Luo, Lu, Wang, Wang, Meziani, Liu, Qi and Sun42} The phenomenon is generally attributed to the emissive traps, electronic conjugate structures, free zigzag sites, and surface states. ^{Reference Wang, Gu, Lei, Wang, Xia and Hao28,Reference Tang, Ji, Cao, Lin, Jiang, Li, Teng, Luk, Zeng, Hao and Lau35,Reference Peng, Gao, Gupta, Liu, Romero-Aburto, Ge, Song, Alemany, Zhan, Gao, Vithayathil, Kaipparettu, Marti, Hayashi, Zhu and Ajayan37} When 450 nm light was used to excite the GQDs aqueous solution, the highest emission at 525 nm was observed. In this work, the excitation wave length was fixed at 450 nm for all PL measurements. Compared with the low chemical stability of bare Cd-based semiconductor QDs, the prominent chemical stability of GQDs was confirmed by the unchanged PL intensity (Fig. S1) with a duration of 21 days exposure to the aerobic conditions. Therefore, GQDs could provide a more reliable platform for chemical sensing.

The presence of function groups (–COOH) containing oxygen in the GQDs provides the reactive sites for absorption and desorption of incoming chemical molecules, indicating the potential application of GQDs as a platform for the chemical detection of metal ions. To evaluate the selectivity of the sensing system for various metal ions, the changes in the PL of the GQD in the presence of representative metals ions including Cr³⁺, Ba²⁺, Ca²⁺, Cd²⁺, Co²⁺, K⁺, Mn²⁺, Ni²⁺, Pb²⁺, Zn²⁺, Cu²⁺, Fe³⁺, Ag⁺, and Hg²⁺ were examined at the same conditions as shown in Fig. 2. The change in the PL intensity is defined by I/I ₀, where I is the PL intensity of the mixture of 0.9 mL of GQDs stock solution and 0.1 mL of metal ion solution and I ₀ is the PL intensity of the mixture of 0.9 mL of GQDs stock solution and 0.1 mL of MQ water. The final concentration of GQDs was controlled at 0.126 mg/mL. After mixing with metal ions for 5 min, an unchanged PL intensity of the mixture indicated a fast response of metal ions with GQDs. When the concentration of metal ions was controlled at 1 mM, the tremendous PL quenching of pristine GQDs was observed for most of the metal ions including Cr³⁺, Ba²⁺, Co²⁺, Mn²⁺, Ni²⁺, Pb²⁺, Cu²⁺, and Fe³⁺ [Fig. 2(a)]. Due to the easy absorption of metal ions on the pristine GQDs, the charge transfer between metal ions and GQDs becomes possible. Therefore, it is not a surprise to observe the PL quenching of GQDs toward multiple metal ions due to the suppressed photoexcited charge recombination of GQDs in the presence of high concentration of metal ions, since metal ions can interact with GQDs through the electrostatic force. Decreasing the concentrations of metal ions will increase the possibility of charge recombination. As shown in Fig. 2(b), the quenching of PL intensity of GQDs was much weaker for 0.1 mM metal ions, compared with that for 1 mM metal ions. The results indicate that the expected selectivity for sensing Cu²⁺ ions could not be realized by using pristine GQDs. Nevertheless, the PL intensity of GQDs could be quenched by Cu²⁺ ions. The PL intensity of GQDs in the presence of Cu²⁺ ions versus Cu²⁺ ions concentration was thoroughly investigated as shown in Fig. 2(c). It was found that PL intensity was almost totally quenched when the concentration of Cu²⁺ ions was larger than 0.2 mM derived from Fig. 2(c). The relative PL intensity of the mixture displays a linear relationship (R = 0.991) to the concentration of Cu²⁺ ions over the range of 0–0.20 mM with a limit of detection of 0.33 µM. The sensitivity of our GQDs is similar to the previous study. ^{Reference Wang, Gu, Lei, Wang, Xia and Hao28} However, our GQDs exhibit the PL response to a wide range of metal ions, compared with the high sensitivity of GQDs toward to Cu²⁺ ions in the previous report. ^{Reference Wang, Gu, Lei, Wang, Xia and Hao28} This may be attributed to different chemical environments of GQDs generated from the synthesis.

FIG. 2.

The PL intensity of GQDs quenched in the presence of different metal ions. (a) Relative PL intensity of the mixture of 0.9 mL of GQDs stock solution and 0.1 mL of metal ion solution, compared with that of 0.9 mL of GQDs stock solution and 0.1 mL of MQ water. The final concentration of metal ions was 1 mM. (b) Relative PL intensity of the mixture of 0.9 mL of GQDs stock solution and 0.1 mL of metal ion solution, compared with that of 0.9 mL of GQDs stock solution and 0.1 mL of MQ water. The final concentration of metal ions was 0.1 mM. (c) Plot of relative PL intensity of GQDs versus the concentrations of Cu²⁺ ions. (d) The recovery of quenched PL intensity of GQDs after treatment of biothiol cysteine. For the PL recovery experiments, the final concentrations of GQDs, metal ions, and cysteine were controlled at 0.112 mg/mL, 0.2 mM, and 1 mM, respectively.

Previous investigations indicate that biothiol cysteine has a strong binding to Cu²⁺ ions, in which two cysteine molecules react with one Cu²⁺ (2Cys + Cu(II) → Cys-Cu(II)-Cys, Fig. S2). ^{Reference Liu, Zhu, Eggers, Nersissian, Faull, Goto, Ai, Sanders-Loehr, Gralla and Valentine43,Reference Rigo, Corazza, Paolo, Rossetto, Ugolini and Scarpa44} Herein, the PL intensity of the mixture of GQDs and metal ions in the presence or absence of biothiol cysteine as the parallel measurements was studied. Details are listed in the experimental section. The concentrations of GQDs were 0.112 mg/mL for all measurements. For each parallel measurement, the concentrations of metal ions and cysteine were identical. The PL intensity of GQDs quenched by metal ions was denoted by I _q. The recovery of PL intensity (I _r) was observed for quenched GQDs solution after adding cysteine. Our control experiments given in Fig. S3 indicate that the PL intensity of 1 mM cysteine is near zero at the identical conditions for PL measurements. Therefore, such a PL recovery could be attributed to the ability of cysteine to capture the metal ions absorbed on the surface of GQDs and suppress the charge transfer between metal ions and GQDs. Herein, 0.2 mM of the final concentration for all metal ions was used. The results are displayed in Fig. 2(d). For Cr³⁺, Ba²⁺, Co³⁺, Mn²⁺, Ni²⁺ ions, the intensities of PL signals after the addition of cysteine have negligible change, compared with I _q. The metal ions, Cu²⁺, Pb²⁺, and Fe³⁺, showed distinct response after cysteine treatment. The value of I _r/I _q for Cu²⁺ ions was 11.3, which was much larger than 2.74 and 1.52 for Pb²⁺ and Fe³⁺, respectively. Such a large difference clearly demonstrates sensitive detection of Cu²⁺ ions could be achieved by adding cysteine to recover PL intensity of GQDs quenched by Cu²⁺ ions.

To address the universality of the aforementioned phenomenon, the PL recovering experiments were also conducted at various concentrations of metal ions (Cu²⁺, Fe³⁺, and Pb²⁺). The recovery factor was defined as the values of (I _r − I _q)/I _q to evaluate the detection of pristine GQDs for Cu²⁺ ions. All measurements were performed for five times. The recovered factors for metal ion concentrations of 0.05, 0.1, and 0.2 mM are summarized in Fig. 3. Three metal ions, Cu²⁺, Fe³⁺, and Pb²⁺, displayed a relative high recovery to cysteine [Figs. 3(a)–3(c)], which was consistent to results shown in Fig. 2(d). The recovery factor (8.52) for 0.2 mM Cu²⁺ ions was 5.56 and 13.43 times higher than that for Pb²⁺ and Fe³⁺ ions with 0.2 mM concentration [Fig. 3(a)], respectively. Similar relationship holds for the metal ion concentrations of 0.1 mM and 0.05 mM as shown in Figs. 3(b) and 3(c), respectively. The summary of recovery factors for three metal ions with various concentrations is displayed in Fig. 3(d), indicating the detection of Cu²⁺ ion could be realized through this simple two-step method as described.

FIG. 3.

Recovered PL intensities of the mixtures of metal ions and GQDs by cysteine at various concentrations of metal ions: (a) 0.2 mM; (b) 0.1 mM; (c) 0.05 mM. (d) Summary of recovery factors for various metal ions. The final concentrations of GQDs and cysteine for all recovery experiments were controlled at 0.112 mg/mL and 1 mM, respectively.

The evaluation of the performance of GQDs as the Cu²⁺ ion sensor with the real water was also carried out by using current method. The river water taken from the Chanba wetland of Xi'an was treated with filtration (0.22 µm membrane) and followed centrifuging at 12,000 rpm for 30 min. The concentration of GQDs was diluted to 0.07 mg/mL by the pretreated river water and MQ water, respectively. Figure 4 shows that the sensing performance of GQDs in the presence of river water displayed a similar phenomenon of PL decrease, compared with that of GQDs aqueous solution prepared from MQ water only. The concentrations of Cu²⁺ ions were in the range of 0–100 µM for the measurements. The PL decrease rate of GQDs dispersed in MQ water was much faster than that in the presence of river water, indicating the PL interference of minerals and organic species in the river water for Cu²⁺ ion detection. However, a linear sensing behavior (R = 0.998) for Cu²⁺ ion sensing in the range of 0–100 µM was observed in Fig. 4, demonstrating that the pristine GQDs could function as the suitable sensing platform for determining the Cu²⁺ ion concentrations in real water. Cu²⁺ ions with a concentration as low as 1 µM are easily detectable at the current system. Therefore, satisfying the sensitivity requirement for detection of the maximum concentration of Cu²⁺ ions in the drinking water (1.0 mg/L, 15.6 µM) defined by Ministry of Health, P.R. China, and this simple approach has the potential practical application in monitoring Cu²⁺ ions level for drinking water.

FIG. 4.

Plot of relative PL intensity of GQDs with a concentration of 0.07 mg/mL versus the concentrations of Cu²⁺ ions in the range of 0–100 µM. The green hollow circles represent the PL measurements carried out in the MQ water only. The purple solid circles represent the changes in the PL intensity after the addition of different amounts of Cu²⁺ ions in the presence of pretreated river water.