The Inherent Relationship between Carbon Nanodots and Carbon Source in the Determination of Metal Ions

Carbon nanodots (CDs) have exhibited excellent sensing capability for various metal ions. However, it is dicult to determine the selectivity of CDs to metal ions. In this work, we chose appropriate carbon source to design CD sensors against Cu(II) and Ag(I). Glycine, histidine and leucine have been conrmed to form complexes with Cu(II) and Ag(I), and were applied to prepare CDs using microwave heating method. The as-prepared CDs inherited the specic ion-binding capability from their carbon source and could response to both Cu(II) and Ag(I). The response sensitivity corresponded to the binding energy between the carbon source and metal ions. These experimental results are very important for the further design of CD sensors for a large variety of analytes.


Introduction
Carbon nanodots (CDs) are one of the most widely researched uorescence nanomaterials. Compared with other uorescence nanomaterials, CDs are particular in their stable photoluminescence (PL) performance, green synthetic methods, abundant raw materials, as well as good biocompatibility [1][2][3]. Fluorescent CDs can be applied in many research elds, especially in environmental and biomedical sensing applications. CDs are sensitive to a list of analytes in solution, including small organic molecules and bio-macromolecules [4][5][6]. PL turn-on or turn-off phenomenon appears and can be utilized to realize the qualitative and quantitative analysis of various analytes.
Interestingly, CDs are born to be the uorescent sensors of various metal ions. Almost all of the CDs can response to one or several kinds of metal ions. The metal ion sensors based on CDs are reliable and accurate, with a high stability and sensitivity. However, there are still two problems demanding to be resolved immediately for the CD sensors. The rst problem is about the sensing mechanism of CD for various metal ions, which is inde nite and requires further research and discussion [7]. Two mechanisms are widely accepted, one of which is the uorescence inner lter effect [8,9] and the other is the binding induced energy transfer between metal ions and CDs [10,11]. The second problem appears in the design of CD sensors. Most of the current CD sensors for metal ions were born through later screening methods.
In simple words, CDs were prepared rst from various carbon source, and then the ion selectivity was determined through contrasting the uorescence variation of CDs in the presence of various metal ions.
These two problems limit the development of CD sensors. This research is focused on the second problem.
In fact, most of the sensors are designed through a classical lock-and-key approach inspired by nature.
The sensing process is accomplished through the speci c interaction between sensors and analytes, just like a lock and a key. In order to realize the speci c recognition, the sensors must be designed with special structure or functional groups to selectively bind with analytes. The lock-and-key approach gives us a clue to design CD sensors. CDs are prepared from carbon source and can inherit some important structural characters or functional groups from carbon source. Therefore, it is possible to control the ion selectivity of CDs through choosing appropriate carbon source, which has speci c binding capacity with certain metal ions.
Amino acids have been con rmed to form stable complexes with Ag(I) and Cu(II) [12][13][14][15][16]. Herein, in this research, Ag(I) and Cu(II) sensors based on CDs were designed through choosing amino acids as carbon source. The inherent relationship between CDs and carbon source in the determination of metal ions was investigated and discussed. Three kinds of amino acids, including glycine (Gly), histidine (His) and leucine (Leu), were chosen to prepare CDs. Experimental results showed that all these CDs were sensitive to both Ag(I) and Cu(II). Moreover, the detection sensitivity depends on the binding energy of the carbon source to various metal ions. Scheme 1 shows the synthetic procedures of CDs, as well as their sensing mechanism to metal ions.

Materials
Urea, glycine, histidine and leucine were ordered from InnoChem Science&Technology Co. Ltd. (Beijing, China). All the chemicals were used directly without further puri cation. Dialysis bag was ordered from Baoman Biological technology Co. Ltd. (Shanghai, China). Ultrapure water was used throughout all experiments.

Instruments
Photoluminescence (PL) measurements were recorded by RF-5301PC (Shimadzu, Japan). Both the excitation and emission slits were set as 3 and 3 nm, respectively. PL spectra were processed using normalized method, and the normalized uorescence intensity was obtained through dividing each PL intensity of the PL spectra by the maximum value of their own. Microwave reactions were carried out using a domestic microwave oven with the power of 750 W.

Synthesis of G-CDs, H-CDs and L-CDs
All of the CDs were prepared using a microwave irradiation method in our previous work [6]. In a typical procedure, 4.0 mmol amino acid and 16.7 mmol urea were dissolved in 10 mL ultrapure water. The mixed solution was heated in a microwave oven for 4 min, and then crude CD powder was gotten. The resultant solid powder was re-dissolved in 10 mL ultrapure water to prepare CD solution. The CD solution was centrifuged at 12 000 rpm for 4.5 min to remove insoluble impurities. The supernatant was then dialyzed against ultrapure water using a dialysis bag (molecular weight cutoff = 8 000-14 000) for 48 hours to remove residual reactants and precursors. The resultant dialysate was freeze-dried and nally brown CD powder was collected. The CDs prepared from glycine, histidine and leucine were called as G-CDs, H-CDs and L-CDs, respectively.

PL Assay of Metal Ions
The PL spectra of CDs were followed in the presence of various metal ions with a certain concentration.
In order to minimize the interference of CD concentration, the absorbance of CD solution was calibrated to 0.1. On this basis, a de nite volume of metal-ion solution (0.5 mL) with various calculated concentrations was added into CD solution (3.5 mL). The PL of the as-prepared CDs were all excitationdepended. The PL spectra of the CD solution were recorded at the optimal excitation wavelength, where the PL intensity of CD can reach the maximum.

Characterization of CDs
The morphology of the CDs can be observed in HR-TEM images. Typical spherical CDs are presented in Fig. 1, with an average diameter of about 10 nm. DLS tests reveal that the as-prepared CDs possess a good size distribution, and the size results keep pace with those from the TEM tests. The PL of the asprepared CDs are typically excitation-dependent due to the quantum con nement [17,18]. With the gradual increase of excitation wavelength from 310 to 390 nm, the uorescence intensity increases rst and then decreases after reaching a maximum value, while the emission wavelength is gradually redshifted (Fig. 2). The optimal excitation wavelength (λ ex ) of these three CDs, as well as its corresponding

PL Assay of Metal Ions
In order to evaluate the metal ion selectivity of CDs, the PL spectra of CDs were monitored in the presence of various metal ions with the same concentration, including Al(III), Ba(II), Ca(II), Cr(III), Mn(II), Pb(II), Hg(II), Ni(I), Co(II), Fe(III), Cu(II), and Ag(I). Results revealed that all of the CDs showed uorescence uctuation in varying degrees in the presence of various metal ions (Fig. 3). Most of the metal ions lead to the uorescence quenching of the CDs. The response was caused by the complicated interaction between the CDs and metal ions, including electrostatic interaction and coordination effect. There exist carboxyl, amino, and hydroxyl groups on the surface of CDs [19][20][21]. The negative carboxyl groups of CDs can electrostatically bind with metal cations [22]. In this respect, the fact that all of the metal ions could have an effect on the CDs is easy to understand. However, the signi cant in uence in the variation of uorescence intensity should originate from the speci c coordination effect [23][24][25]. Among these given metal ions, only Fe(III), Cu(II), and Ag(I) had signi cant effects on the CDs. Many CDs have been reported to be sensitive to Fe(III), due to their oxygen-contained functional groups coordinated with Fe(III) [26,27]. Cu(II) or Ag(I) responsive CDs have also been reported in some research, respectively [28][29][30][31].
However, there are hardly any literatures about the CDs sensitive to both Cu(II) and Ag(I), simultaneously. Therefore, the particular response should depend on the carbon source we chose to prepare CDs. Gly, His and Leu have been con rmed to form stable coordination complexes with Cu(II) and Ag(I) [32][33][34][35][36][37]. The CDs prepared from these three amino acids could also inherit the coordination capability.
In order to investigate the inherent relationship between the CDs and their corresponding carbon source in the determination of metal ions, we also monitored the uorescence intensity of CDs with the variation of Cu(II) ang Ag(I) concentration (Fig. 4). With the increase of Cu(II) and Ag(I) concentration, the uorescence intensity of CDs decreased regularly (Fig. 5). There existed a good linear relationship between the uorescence intensity of CDs and the concentration of metal ions in the range of low concentrations. Afterwards, the uorescence intensity of CDs tended to be constant, indicating that the binding of metal ions reached a saturated state. Comparatively speaking, the uorescence intensity of H-CDs quenched more compared with the other two CDs, and L-CDs took the second place. The uorescence quenching degree of CDs can re ect their sensitivity to metal ions. Therefore, among these three CDs, H-CDs are the most sensitive CDs to Cu(II) and Ag(I); L-CDs comes second. Moreover, the sensitivity of the CDs to Ag(I) and Cu(II) keeps pace with the binding ability of their corresponding aminoacid carbon source with Ag(I) and Cu(II). The binding ability between amino acids and metal ions can be evaluated through calculating their binding energy, which have been reported in some literatures [38][39][40]. Table 1 shows the binding energies of different amino acids to silver and copper [13], and the binding energy is Gly < Leu < His. Table 1 Relative binding energies (kcal/mol) of amino acids to silver(I), copper(II) and the proton [13] Amino

Fluorescence Quenching Mechanism
In order to investigate the mechanism of the metal-ion induced uorescence quenching of CDs, the uorescence lifetime (FL) of the CDs was monitored before and after adding Cu(II) (Fig. 6). The uorescence decay curves of three CDs can be tted by a single-exponential formula respectively with a similar lifetime (FL G−CD = 3.2567 ns; FL L−CD = 3.4580; FL H−CD = 2.8621). The FL of the CDs almost stayed unchanged in the presence of Cu(II), indicating a time-independant mechanism. Therefore, a static quenching mechanism could account for the response behavior of the CDs to Cu(II) [41,42]. A non uorescent ground-state complex should be formed between the CDs and metal ions, which immediately returns to the ground-tate without emission of a photon after absorbing light.

Conclusions
In summary, we prepared three kinds of CDs through adopting speci c carbon source in order to predetermine their ion-sensing capability. The carbon source, including glycine, histidine and leucine, has been con rmed to bind with Cu(II) and Ag(I). The as-prepared CDs inherited the binding capability from their corresponding carbon source. The binded metal ions quenched the uorescence of the CDs through a static mechanism, based on which quantitative determination of Cu(II) and Ag(I) can be realized. Moreover, there existed an inherent relationship between the ion-determining capability of the CDs and the binding energy between their corresponding carbon source and metal ions. These results provided valuable reference value for the design of the sensors based on CDs.

Data Availability
The data used to support the ndings of this study are included within the article.

Authors' Contributions
All authors contributed to the study conception and design.

Compliance with Ethical Standards
Competing Interests The authors have no relevant nancial or non-nancial interests to disclose.
Code Availability Not applicable.