Biosensors and Bioelectronics 70 (2015) 232–238

Contents lists available at ScienceDirect

Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios

A carbon dot-based “off–on” fluorescent probe for highly selective and sensitive detection of phytic acid Zhao Gao a,b, Libing Wang a, Rongxin Su a,b,n, Renliang Huang c,nn, Wei Qi a,b, Zhimin He a a

State Key Laboratory of Chemical Engineering, School of Chemical Engineering and Technology, Tianjin University, Tianjin 300072, PR China Collaborative Innovation Center of Chemical Science and Engineering (Tianjin), Tianjin 300072, PR China c School of Environmental Science and Engineering, Tianjin University, Tianjin 300072, PR China b

art ic l e i nf o

a b s t r a c t

Article history: Received 19 January 2015 Received in revised form 13 March 2015 Accepted 19 March 2015 Available online 20 March 2015

We herein report a facile, one-step pyrolysis synthesis of photoluminescent carbon dots (CDs) using citric acid as the carbon source and lysine as the surface passivation reagent. The as-prepared CDs show narrow size distribution, excellent blue fluorescence and good photo-stability and water dispersivity. The fluorescence of the CDs was found to be effectively quenched by ferric (Fe(III)) ions with high selectivity via a photo-induced electron transfer (PET) process. Upon addition of phytic acid (PA) to the CDs/Fe(III) complex dispersion, the fluorescence of the CDs was significantly recovered, arising from the release of Fe(III) ions from the CDs/Fe(III) complex because PA has a higher affinity for Fe(III) ions compared to CDs. Furthermore, we developed an “off–on” fluorescence assay method for the detection of phytic acid using CDs/Fe(III) as a fluorescent probe. This probe enables the selective detection of PA with a linear range of 0.68–18.69 μM and a limit of detection (signal-to-noise ratio is 3) of 0.36 μM. The assay method demonstrates high selectivity, repeatability, stability and recovery ratio in the detection of the standard and real PA samples. We believe that the facile operation, low-cost, high sensitivity and selectivity render this CD-based “off–on” fluorescent probe an ideal sensing platform for the detection of PA. & 2015 Elsevier B.V. All rights reserved.

Keywords: Carbon dot Fluorescence Off–on Sensor Phytic acid

1. Introduction Phytic acid (PA) is the principal storage form of phosphorus in many plant tissues, such as cereals, grains, soybeans, fruits and vegetables (Oatway et al., 2001; Wu et al., 2009). It is also ubiquitous in mammalian cells and plays a positive role in normal physiological processes. Nevertheless, there are still major gaps in the full understanding of the biological and physiological roles of this molecule and its metabolites. To understand the exact roles of PA and provide valuable information for its dietary intake and metabolism, the detection of PA in the fields of biological science, medicine, and food science is generally required (Wu et al., 2009). From the viewpoint of practical applications, an excellent method should not only be highly sensitive and selective but also simple and economical in operation. However, it is difficult for a single method to satisfy all of these requirements. In recent years, fluorescent sensing has received widespread attention because of its high sensitivity, simplicity, and rapid implementation (W. Chen n Corresponding author at: State Key Laboratory of Chemical Engineering, School of Chemical Engineering and Technology, Tianjin University, Tianjin 300072, PR China. Fax: þ 86 22 27407599. nn Corresponding author. Fax: þ 86 22 27407599. E-mail addresses: [email protected] (R. Su), [email protected] (R. Huang).

http://dx.doi.org/10.1016/j.bios.2015.03.043 0956-5663/& 2015 Elsevier B.V. All rights reserved.

et al., 2009a; Gao et al., 2014; Hou et al., 2014; Li et al., 2008; Shen et al., 2012, 2011). Up to now, limited reports are available on the fluorescent quantification of PA only involving the use of organic fluorescent dyes (Cao et al., 2011; Y. Chen et al., 2009b). Therefore, it is highly desirable to develop a simple, biocompatible and efficient fluorescent sensing system for the detection of PA. As a newly emerging class of fluorescent probes, carbon dots have attracted increasing attention over the past few years because of their advantageous features, such as favorable bright fluorescence, biocompatibility and excellent photo-chemical stability (da Silva and Goncalves, 2011; Ding et al., 2014). Recently, carbon dots have been reported for the optical sensing of many biologically important analytes, including proteins, anions, small molecules and metal ions (Guo et al., 2015; Yang et al., 2014a; Zhang and Chen, 2014). To provide a unique interaction between CDs and each analyte, surface functionalization of CDs is required to control their interfacial properties. In this regard, an efficient synthetic strategy was developed that involved the pyrolysis treatment of carbon precursors with different passivation reagents, achieving simultaneous fabrication and surface functionalization of CDs with diverse chemical groups (Liu et al., 2013; Shen and Xia, 2014; Yang et al., 2014b; Zhang and Chen, 2014). Specifically, to focus on the detection of metallic ions, Dong et al. reported a polyamine-functionalized carbon dot that was

Z. Gao et al. / Biosensors and Bioelectronics 70 (2015) 232–238

synthesized by carbonization of citric acid with branched polyethylenimine. This carbon dot was demonstrated as an efficient fluorescent probe for the detection of Cu2 þ ions via the inner filter effect (Dong et al., 2012a,b). Recently, Zong et al. produced CDs within mesoporous silica spheres via the pyrolysis treatment of citric acid and complex salts, achieving good detection performance for both Cu2 þ ions and L-cysteine (Zong et al., 2014). Another successful example was use of ethylenediamine tetraacetic acid salts as precursors for the preparation of CDs, which showed a highly selective response to Hg2 þ ions and biothiols via an electron transfer process (Zhou et al., 2012). Moreover, a nitrogendoped carbon dot was prepared using folic acid as both the carbon and nitrogen sources and was also able to selectively detect Hg2 þ ions (Zhang and Chen, 2014). In these cases, it should be noted that the diversity of the surface characteristics of CDs, derived from the different starting materials in the pyrolysis process, led to versatile CD-based fluorescent probes with fascinating detection performances. Inspired by the fact that PA has a strong binding affinity for ferric Fe(III) ions (Table S1) (Crea et al., 2008), this study aimed to develop a new carbon dot with a high fluorescence response to Fe(III) ions and subsequently construct a fluorescent platform for the selective and sensitive detection of PA. Herein, we prepared carbon dots via a one-step pyrolysis process using citric acid as the carbon precursor and lysine as the passivation reagent. The morphologies, diameter distribution, chemical structures, and optical properties of the CDs were characterized by transmission electron microscopy and spectroscopic analysis. Moreover, we investigated the fluorescence intensity changes of the CDs upon the addition of PA, Fe(III) ions, and Fe(III)/ PA. The fluorescence mechanism was further elucidated with fluorescence decay, zeta potential measurements and resonance light scattering spectroscopy. We also investigated the effect of different metal ions on the fluorescence response of CDs. Finally, we developed a carbon dot-based “off–on” fluorescent probe for the detection of phytic acid. The detection performances, including sensitivity, selectivity, repeatability, stability, and recovery, were evaluated using the standard or real PA samples.

2. Materials and methods 2.1. Materials Phytic acid (PA, 99.8%) was purchased from Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan). Citric acid, lysine, ferric chloride hexahydrate (FeCl3  6H2O), sodium hydroxide, and 2-(N-Morpholino) ethanesulfonic acid (MES) were bought from Aladdin Chemical Reagent Co. Ltd. (Shanghai, China). Metal salts (KCl, NaCl, AgNO3, Zn(NO3)2  6H2O, CaCl2, MgSO4, MnCl2  4H2O, CdCl2, BaCl2, Cu(NO3)2  5H2O, CoCl2  6H2O, NiCl2  6H2O, Hg(ClO4)2  3H2O, CrCl3  3H2O, and AlCl3) were purchased from Guangfu Chemical Reagent Co. Ltd. (Tianjin, China). All reagents were of analytical grade and were used as received without further purification. Ultrapure water prepared from a Millipore water purification system was used throughout the experiments. 2.2. Preparation of carbon dots Carbon dots were prepared by hydrothermal treatment of citric acid and lysine. In a typical experiment, 4.2 g of citric acid and 1.46 g of lysine were added to 40 mL of water. The resulting solution was transferred to a poly(tetrafluoroethylene)-lined autoclave (100 mL) and maintained under autogenous pressure at 200 °C for 5 h. Then, the mixture was cooled to room temperature and dialyzed against ultrapure water using a 500 Da cut-off dialysis bag for 2 days to remove the unreacted small molecules. The

233

final CD dispersion solution with a concentration of approximately 19.8 g L  1 was obtained and stored at 4 °C for subsequent use. 2.3. Characterization High resolution transmission electron microscopy (HRTEM) was performed on a JEM-2100 microscope (JEOL Ltd., Akishima, Japan) operating at an accelerating voltage of 200 kV. Fourier transform infrared (FTIR) spectra were recorded on a Nicolet-560 FTIR spectrometer (Nicolet Co., Madison, WI, USA) with an accumulation of 20 scans and a resolution of 4 cm  1. UV/vis spectroscopy was performed on a TU-1810 spectrophotometer (Purkinje General Instrument Co. Ltd., Beijing, China) using a quartz cell with a 1.0 cm optical path. The fluorescence (FL) and resonance light scattering (RLS) spectra were recorded on a Cary Eclipse fluorescence spectrometer (Agilent Technologies, Santa Clara, CA, USA). Fluorescence lifetimes were measured using a Fluorolog-3 spectrofluorometer (Horiba Jobin Yvon Inc., Lille, France) operating in the time-correlated single photon counting mode (TCSPC). A NanoLED diode emitting pulses at 279 nm was used as an excitation source. Fluorescence decay analysis software, DAS6, was used to fit the model functions to the experimental data. 2.4. Fluorescence quenching of CDs The CD dispersion solution was diluted with a MES buffer solution (10 mM, pH 5.7) to a final concentration of 0.198 mg L  1. Then, 20 μL of ferric chloride solution (10 mM) was added to 2 mL of the CD dispersion (0.198 mg L  1), followed by vigorous shaking on a vortex mixer and incubation at 25 °C for 15 min. For comparison, different metal cations with a final concentration of 100 μM were added to 2 mL of the CD dispersion (0.198 mg L  1) and incubated at 25 °C for 15 min. The fluorescence spectra were recorded at an excitation wavelength of 350 nm and excitation/ emission slits of 5 nm. 2.5. Fluorescent detection of phytic acid In a typical assay, 20 μL of ferric chloride solution (10 mM) was incubated with 2 mL of the CD dispersion (0.198 mg L  1) at room temperature for 15 min as described previously. Then, the standard or real (corn grains, the details on sample preparation were described in Supporting information) phytic acid solutions with different concentrations were added to the resulting mixture. After incubation for 10 min, the fluorescence intensities of the solutions were measured with an excitation at 350 nm. The selectivity of phytic acid detection was investigated by the addition of other interferences following the same procedure. All experiments were performed at room temperature and repeated at least three times.

3. Results and discussion The CDs were prepared via the hydrothermal treatment of citric acid with lysine at 200 °C for 5 h, yielding a brown yellow CD dispersion. In this system, citric acid was chosen as the carbon source and lysine was used as the surface passivation reagent. On one hand, citric acid is thought to be an excellent carbon source for the synthesis of CDs because of its well known low carbonization temperature (Dong et al., 2012b). On the other hand, previous studies have demonstrated that amine-containing molecules, such as 1,2-ethylenediamine, diethylamine, and diethylenetriamine, are able to passivate CDs with some amines, thus increasing the quantum yield and selectively recognizing target analytes, such as Hg2 þ , Cu2 þ , and Cr6 þ ions (Zhang and Chen, 2014; Zheng et al., 2013; Zong et al., 2014). In this study, lysine has two amines and

234

Z. Gao et al. / Biosensors and Bioelectronics 70 (2015) 232–238

Scheme 1. Schematic illustration of fluorescence response of CDs in the presence of Fe(III) ions and Fe(III)/PA. Mode a: photo-induced fluorescence emission yielding bright blue luminescence. Mode b: photo-induced electron transfer (PET) in which an electron in the excited state enters the unfilled d orbit of Fe, leading to fluorescence quenching of the CDs.

one carboxyl group. We predict that this unique chemical structure of lysine, different from that of the reported passivation reagents, will allow for a significant fluorescence response to Fe(III) ions because of its strong affinity for lysine-derived carboxyl,

aromatic hydroxyl or multiple oxygenated groups on the surface of CDs. Once the complexation between Fe(III) ions and CDs is complete, fluorescence quenching of the CDs should be achieved (Scheme 1, mode b). To confirm this speculation, the structural

Fig. 1. (a) TEM image of the CDs prepared by the hydrothermal treatment of citric acid with lysine. Inset: particle size distribution obtained from TEM image. (b) High resolution TEM images of the CDs. (c) FTIR spectrum of the CDs. (d) UV/vis absorption, fluorescence excitation and emission spectra of an aqueous suspension of the CDs. Inset: photographs of CDs in aqueous solutions under visible (left) and UV (right) light.

Z. Gao et al. / Biosensors and Bioelectronics 70 (2015) 232–238

3.1. Structural and optical characterization of CDs TEM analysis revealed that the synthesized CDs are highly dispersible in aqueous solution and have a narrow size range of 2– 5 nm. The average particle size of the CDs is approximately 3.2 nm (Fig. 1a). As shown in the HRTEM image in Fig. 1b, most of the CDs were observed to be amorphous carbon nanoparticles, whereas a few particles possessed well-resolved lattice fringes, which is consistent with the previously reported results (Zhang and Chen, 2014; Zhu et al., 2013). To investigate the chemical structure of the CDs, FTIR spectroscopy was employed to characterize the resulting samples. As shown in Fig. 1c, the FTIR spectrum has two peaks at 3364 cm  1 and 1570 cm  1, which are attributed to the stretching and bending vibrations of N–H groups, respectively (Zhu et al., 2013). Additionally, three peaks at 3325 cm  1, 3090 cm  1, and 1712 cm  1 were observed, possibly indicative of the O–H stretching vibrations and C ¼O stretching vibrations, respectively (Zong et al., 2014). The FTIR results indicate the presence of amino (NH2), carboxyl (COOH) and hydroxyl (OH) groups on the surface of the CD particles. These groups are able to stabilize the CDs in an aqueous environment and also help to generate energy gaps that emit light efficiently when photo-stimulated (Sun et al., 2006). The

Normalized Intensity

1.0

I, CDs II, CDs/Fe(III) III, CDs/Fe(III)/PA IV, Fe(III)/PA V, CDs/PA

0.8 0.6 0.4 0.2 0.0

good stability and dispersibility of the CDs in aqueous solutions, together with their excellent biocompatibility, render them attractive as fluorescent probes for the in vivo/in vitro detection of biologically important compounds (e.g., PA in this study). We further investigated the optical characteristics of CDs using UV/vis and photoluminescence spectroscopy. As shown in Fig. 1d, an aqueous solution of CDs exhibits bright blue luminescence under irradiation from a UV lamp with a wavelength of 365 nm. Similar to the previously reported values (Zhai et al., 2012; Zhu et al., 2013), the maximum excitation wavelength and emission wavelength of the CDs were 350 and 450 nm, respectively (Fig. 1d, Scheme 1, mode a). As shown in Fig. S2, the fluorescence emission wavelengths of the CDs were found to be excitation-independent, indicating the uniformity of the size and surface states (Zhai et al., 2012). The fluorescence intensity of the CDs remained unchanged under continuous 365 nm illumination for up to 3 h (Fig. S3), indicative of good photo-stability. Additionally, the UV/vis spectrum showed that the CDs exhibit a significant UV/vis absorption band centered at approximately 350 nm and a shoulder peak at 230 nm (Fig. 1d). The absorption peak at 350 nm corresponds to the optimum fluorescence excitation peak and should be attributed to the n  πn transition of the chemical groups on the surface of the CDs (Zheng et al., 2013; Zhu et al., 2013). The shoulder peak is a typical characteristic of CDs, arising from the π–πn transition of aromatic C ¼C bonds (Pan et al., 2010). Generally, the optical properties of CDs are known to be dictated by a combination of a few factors, i.e., size, shape and surface state (Zhu et al., 2013). In this case, the

CDs CDs/Fe(III) CDs/Fe(III)/PA

10000

Counts

and optical characterization of the synthesized CDs and their fluorescence response to Fe(III) ions were investigated in this study.

235

1000

100

400

450

500

550

600

650

Wavelength / nm

20

40

60

80

100

Time / ns

30

Zeta Potential (mV)

21.4 mV 20 10 0 -10

-9.72 mV -14.26 mV

-20

CDs

CDs/Fe(III)

CDs/Fe(III)/PA

Fig. 2. (a) Fluorescence emission spectra of CDs under different conditions. (b) Time-resolved fluorescence decays of CDs (black), CDs/Fe(III) complex (red) and CDs/Fe(III)/PA mixture (blue). (c) Zeta potentials of the CDs under different conditions. (d) Resonance light scattering spectra of the CD dispersion under different conditions (I: CDs; II: CDs/ Fe(III); III: Fe(III)/PA; IV: CDs/Fe(III)/PA). Inset in d): RLS spectra of the CD dispersion in the presence of Fe(III) ions and PA with different concentrations. Default concentrations of Fe(III) and PA were 100 μM and 16 μM, respectively (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.).

236

Z. Gao et al. / Biosensors and Bioelectronics 70 (2015) 232–238

above observations suggest that surface state emission plays a decisive role in the photoluminescence of the CDs. 3.2. Fluorescence response of CDs and mechanism analysis Fig. 2a shows the fluorescence emission spectra of CDs under different conditions. The fluorescence of the CDs was found to be significantly quenched in the presence of Fe(III) ions when 20 μL of ferric chloride (10 mM) solution was added to 2 mL of the CD dispersion (0.198 mg L  1). In this case, the fluorescence intensity decreased to 43.8% of the original value within 15 min (Fig. 2a, curves I and II). We expect that the coordination interactions between Fe(III) ions and the carboxyl and/or aromatic hydroxyl groups on the surface of the CDs lead to the formation of CDs/Fe (III) complexes. Upon photo-excitation, the excited state of the CDs can relax its energy in a non-radiative electron transfer manner, in which an electron in the excited state enters the unfilled d orbit of Fe, leading to the fluorescence quenching of CDs through a photoinduced electron transfer (PET) process (Scheme 1, mode b). Previous studies have demonstrated that the surface state has a significant impact on the band gap and excited states of CDs, thus resulting in the change in fluorescence emission and lifetimes (Bao et al., 2011; Zheng et al., 2011). To further confirm the quenching mechanism, we measured the time-resolved fluorescence decays using a TCSPC technique to investigate the excited states of the CDs in the absence/presence of Fe(III) ions (Fig. 2b). The data were found to be fit well to a classical three-exponential function (Table S2). In comparison to the CDs alone, the CDs/Fe(III) complex has a much shorter lifetime (13.7 ns versus 15.5 ns). These results provide powerful evidence in support of our previously mentioned speculation. According to the above analysis, the fluorescence quenching of the CDs should be attributed to the dynamic photoinduced electron transfer that arose from the complexation between the CDs and Fe(III) ions. This PET quenching mechanism was also found in the cases of Cu2 þ and Hg2 þ ions, whereas Cr6 þ and Zn2 þ ions quench the fluorescence of CDs through the inner filter effect and internal charge transfer, respectively, because of their different electronic configurations (Chandra et al., 2012; Dong et al., 2012a; Zhang and Chen, 2014; Zheng et al., 2013). It is well known that multiphosphates show high affinity for ferric ions because of their multiple ligands (Simmons, 2010). Among these multiphosphates, phytic acid (the molecular structure of which is shown in Fig. S1) has a high density of negatively charged phosphate groups (Selle et al., 2000). Therefore, we infer that ferric ions are able to selectively and preferentially bind with phytic acid, thus leading to fluorescence recovery of the CDs in the

presence of Fe(III) ions. Based on this point of view, we attempted to employ the CDs/Fe(III) complex as a fluorescent light-on probe for the detection of phytic acid. As expected, upon the addition of PA to the CDs/Fe(III) complex, the fluorescence intensity at 450 nm significantly increased from 43.8% to 82% of the original value within 10 min (Fig. 2a, curve III). In a control experiment, no detectable fluorescence was observed upon the addition of PA to the Fe(III) solution alone (Fig. 2a, curve IV). Additionally, we found no change in fluorescence intensity when the PA solution was added to the CD dispersion alone (Fig. 2a, curve V). These results confirm that the preferential binding of PA with Fe(III) ions released from the CDs/Fe(III) complex leads to the recovery of fluorescence (Scheme 1). The fluorescence decay analysis shows that the lifetime of the resulting CDs/Fe(III)/PA mixture was restored to 15.6 ns (Fig. 2b, Table S2), which is similar to that of the CD dispersion alone (15.5 ns). This result suggests that the introduction of PA allows for the release of Fe(III) ions from the CDs/Fe(III) complex because of its higher affinity for PA than for CDs. Moreover, we measured the zeta potentials of the CDs at different conditions to monitor their surface characteristics. As shown in Fig. 2c, the zeta potential of the CDs shifted from  9.72 to 21.4 mV when ferric ions were added, whereas the value deceased to  14.26 mV upon the addition of PA to the CDs/Fe(III) complex solution. The results further confirm that the Fe(III) ions initially associate with the carboxyl and aromatic hydroxyl groups on the surface of the CDs and then release from the CDs/Fe(III) complex to bind with PA upon its introduction, which is accompanied by a significant “off–on” fluorescence response (Scheme 1). The formation of a PA–Fe complex was further confirmed by resonance light scattering (RLS) analysis. The RLS spectra were recorded by synchronous scanning of the excitation and emission monochromators from 200 to 800 nm. As shown in Fig. 2d, the RLS intensities of the CD dispersion in the absence/presence of Fe(III) ions are very weak (curves I and II). However, the RLS intensities increased dramatically with increasing concentration of PA with almost the same RLS pattern during evolution (Fig. 2d, inset). The RLS pattern of the CDs/Fe(III)/PA mixture is similar to that of the Fe (III)/PA mixture solution (Fig. 2d, curves III and IV). This observation confirms the strong binding of Fe(III) ions to PA molecules, which leads to the formation of the Fe(III)/PA complex. We also measured the time-dependent fluorescence responses of CDs upon addition of Fe(III) ions. Fig. 3a, curve I, shows the fluorescence quenching of CDs in the presence of Fe(III) ions as a function of incubation time. The fluorescence intensity decreased dramatically at the initial stage and reached a plateau after 15 min,

1.4 1.0

1.0

II 1.3

0.8

0.6

F/F0

F /F0

1.2

F /FA

0.8

0.6

1.1

I

0.4

1.0

0.4 0

2

4

6

8

10

12

14

16

Time / min Fig. 3. (a) Time-dependent fluorescence responses of the CD dispersion upon addition of 100 μM Fe(III) ions (curve I) and the time-dependent fluorescence recovery of CDs in the presence of Fe(III) ions (100 μM) upon addition of 16 μM PA (curve II). (b) The fluorescence responses of CDs towards different metal ions. The final concentrations of CDs and metal ions were held constant at 0.198 mg L  1 and 100 μM, respectively.

Z. Gao et al. / Biosensors and Bioelectronics 70 (2015) 232–238

indicating that the interaction between CDs and Fe(III) reached equilibrium. Additionally, a suitable concentration of Fe(III) ions was optimized as 100 μM, as shown and discussed in Fig. S4. Furthermore, we investigated the fluorescence response of CDs upon addition of different metal ions. As shown in Fig. 3b, in addition to Fe(III) ions, no obvious fluorescence changes were observed in the presence of other metal ions. The results indicate that the as-synthesized CDs exhibit a highly selective response towards Fe(III) ions. This excellent selectivity is attributed to the strong interaction between Fe(III) ions and the unique surface groups of the CDs, as previously discussed, as well as the PETbased quenching process.

PA 2SO4 NO3 CH3COO Cl 3+ Al 2+ Ni 2+ Cu 2+ Zn 2+ Mg + K + Na

3.3. CD-based “off–on” fluorescence detection of PA According to the above findings, we attempted to develop an “off–on” assay method for the detection of PA. To achieve sensitive detection, we investigated the effect of pH value on the fluorescence intensity of CDs. As shown in Fig. S5, at pH values below 5, the fluorescence intensity is low, possibly because the carboxyl groups on the surface of the CDs become protonated, which weakens the electrostatic repulsion between CDs and renders them unstable. In contrast, a basic environment should lead to the formation of insoluble ferric hydroxide; therefore, we adjusted the pH values to a weak acidic condition using MES buffer (pH 5.7, 10 mM) for subsequent detection experiments. Furthermore, to understand the kinetic characteristics of the interaction between the CDs/Fe(III) complex and PA, we measured the time-dependent fluorescence responses of CDs upon addition of 50 μM PA. As shown in Fig. 3a, curve II, after the addition of PA to the CDs/Fe(III) complex solution, the fluorescence of the CDs was restored within 10 min at 25 °C and then remained stable, which suggests that the PA can bind Fe(III) ions gradually and displace them from the surface of the CDs. Under the optimal conditions, the sensing performance of the CDs/Fe(III) fluorescent probe was evaluated by adding different concentrations of PA into the CDs/Fe(III) mixture solution (CDs: 2 mL, 0.198 mg L  1; Fe(III): 100 μM). With increasing amounts of PA, the fluorescence intensity of the CDs was gradually restored (Fig. 4a). The fluorescence recovery factor, (F  FA)/FA, was plotted as a function of PA concentration, where FA and F are the fluorescence intensities of the CDs/Fe(III) mixture at 450 nm in the absence and presence of PA, respectively. As shown in Fig. 4b, the value of (F  FA)/FA increased gradually with increasing PA concentration and a linear calibration in the range from 0.68 to

237

-0.1

0.0

0.1

0.2

0.3

0.4

0.5

0.6

(F-FA)/FA Fig. 5. Fluorescence response of the CD dispersion in the presence of Fe(III) ions (100 μM) to various substances. The concentrations of these substances were 50 μM.

18.69 μM was obtained (R2 ¼ 0.981). Within this concentration range, the relative standard deviation (RSD) values were in a range of 1.93–6.64% (Table S3). The limit of detection was estimated to be 0.36 μM, as calculated according to the 3SD/K criterion (based on three times signal-to-noise ratio) (Ripp, 1996), where K is the slope of the calibration curve and SD represents the standard deviation of a blank (n ¼11). Such assay sensitivity is comparable or superior to those of spectroscopic methods for PA determination performed in aqueous solutions (Table S4). Moreover, the fluorescence assay demonstrated in this study is also a facile, low-cost and efficient method. These advantages render this CD-based “off–on” fluorescent probe an ideal sensing platform for the detection of PA. Additionally, the selectivity of this fluorescent probe was investigated by examining the fluorescence responses of the CDs/Fe (III) complex toward various substances, including different metal ions, anions, inositol and ethylenediaminetetraacetic acid (EDTA). As shown in Fig. 5 and S6, no obvious signal change was observed for these substances, whereas the addition of PA resulted in significant fluorescence recovery, as described previously. The high selectivity of this CD-based fluorescent probe provides a potential for the efficient detection of PA in complex matrix samples. Finally,

0.6

120

PA

40

0.4 0.4

0.3

(F-FA )/FA

80

(F-FA ) /FA

Intensity

0.5

0.2 0.1

0

400

450

500

550

Wavelength / nm

600

0.0

0.3 0.2 0.1 0.0

0

10

20

0

30

4

8

12

PA / µM

40

16

50

20

60

PA / µM

Fig. 4. (a) The fluorescence emission spectra of CDs in the presence of Fe(III) ions (100 μM) upon addition of PA with different concentrations. (b) Plot of the fluorescence recovery factor (F  FA)/FA versus PA concentration (Inset is a linear region). FA is the fluorescence intensity of CDs in the presence of 100 μM ferric ions, and F is the recovered fluorescence intensity after the addition of PA. The error bars represent the standard deviation of three measurements.

238

Z. Gao et al. / Biosensors and Bioelectronics 70 (2015) 232–238

to investigate the reliability in practical analysis, the proposed fluorescence method was used for determining the recovery ratios by adding different concentrations of PA into the real corn grain sample. As shown in Fig. S7 and Table S5, the recovery ratios of these measurements ranged from 92.0% to 98.3%, indicating that the proposed CDs/Fe(III) probe is suitable for the quantitative determination of PA in real samples.

4. Conclusions In summary, we have demonstrated the successful preparation of a new carbon dot with strong fluorescence response to Fe(III) ions using citric acid and lysine as a carbon source and surface passivation reagent, respectively. The complexation between Fe (III) and the chemical groups (e.g., COOH, OH) on the surface of the CDs resulted in significant fluorescence quenching via a photoinduced electron transfer process. As expected, the addition of PA into the CDs/Fe(III) complex dispersion led to fluorescence recovery because of the higher affinity of PA for Fe(III) ions compared to CDs, which allows for the release of Fe(III) ions from the surface of the CDs and binding to PA molecules. Based on this fluorescence quenching-recovery phenomenon, we developed an “off–on” fluorescence assay method for the detection of PA using CDs/Fe(III) as a probe. This fluorescent probe demonstrates facile operation, high sensitivity and selectivity and provides great potential for the efficient detection of phytic acid.

Acknowledgments This work was supported by the Ministry of Science and Technology of China (Nos. 2012AA06A303, 2012YQ090194, and 2012BAD29B05), the National Natural Science Foundation of China (Nos. 51473115 and 21276192), and the Ministry of Education (Nos. B06006 and NCET-11-0372).

Appendix A. Suplementary Information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.bios.2015.03.043.

References Bao, L., Zhang, Z.L., Tian, Z.Q., Zhang, L., Liu, C., Lin, Y., Qi, B., Pang, D.W., 2011. Adv.

Mater. 23, 5801–5806. Cao, S.H., Dong, N.A., Chen, J.W., 2011. Phytochem. Anal. 22, 119–123. Chandra, S., Pathan, S.H., Mitra, S., Modha, B.H., Goswami, A., Pramanik, P., 2012. RSC Adv. 2, 3602–3606. Chen, W., Peng, C.F., Jin, Z.Y., Qiao, R.R., Wang, W.Y., Zhu, S.F., Wang, L.B., Jing, Q.H., Xu, C.L., 2009a. Biosens. Bioelectron. 24, 2051–2056. Chen, Y., Chen, J., Luo, Z., Ma, K., Chen, X., 2009b. Microchim. Acta 164, 35–40. Crea, F., De Stefano, C., Milea, D., Sammartano, S., 2008. Coordin. Chem. Rev. 252, 1108–1120. da Silva, J., Goncalves, H.M.R., 2011. Trac-Trends Anal. Chem. 30, 1327–1336. Ding, C.Q., Zhu, A.W., Tian, Y., 2014. Acc. Chem. Res. 47, 20–30. Dong, Y., Wang, R., Li, G., Chen, C., Chi, Y., Chen, G., 2012a. Anal. Chem. 84, 6220–6224. Dong, Y., Wang, R., Li, H., Shao, J., Chi, Y., Lin, X., Chen, G., 2012b. Carbon 50, 2810–2815. Gao, Z., Su, R., Qi, W., Wang, L., He, Z., 2014. Sens. Actuators B-Chem. 195, 359–364. Guo, Y.M., Zhang, L.F., Zhang, S.S., Yang, Y., Chen, X.H., Zhang, M.C., 2015. Biosens. Bioelectron. 63, 61–71. Hou, J., Zhang, H., Yang, Q., Li, M., Song, Y., Jiang, L., 2014. Angew. Chem. Int. Ed. 53, 5791–5795. Li, M., He, F., Liao, Q., Liu, J., Xu, L., Jiang, L., Song, Y., Wang, S., Zhu, D., 2008. Angew. Chem. Int. Ed. 47, 7258–7262. Liu, J.M., Lin, L.P., Wang, X.X., Jiao, L., Cui, M.L., Jiang, S.L., Cai, W.L., Zhang, L.H., Zheng, Z.Y., 2013. Analyst 138, 278–283. Oatway, L., Vasanthan, T., Helm, J.H., 2001. Food Rev. Int. 17, 419–431. Pan, D., Zhang, J., Li, Z., Wu, C., Yan, X., Wu, M., 2010. Chem. Commun. 46, 3681–3683. Ripp, J., 1996. Analytical detection limit guidance & laboratory guide for determining method detection limits Wisconsin Department of Natural Resources, Laboratory Certification Program. Selle, P.H., Ravindran, V., Caldwell, A., Bryden, W.L., 2000. Nutr. Res. Rev. 13, 255–278. Shen, P., Xia, Y., 2014. Anal. Chem. 86, 5323–5329. Shen, W., Li, M., Wang, B., Liu, J., Li, Z., Jiang, L., Song, Y., 2012. J. Mater. Chem. 22, 8127–8133. Shen, W., Li, M., Xu, L., Wang, S., Jiang, L., Song, Y., Zhu, D., 2011. Biosens. Bioelectron. 26, 2165–2170. Simmons, J., 2010. Water Air Soil Pollut. 209, 123–132. Sun, Y.-P., Zhou, B., Lin, Y., Wang, W., Fernando, K.A.S., Pathak, P., Meziani, M.J., Harruff, B.A., Wang, X., Wang, H., Luo, P.G., Yang, H., Kose, M.E., Chen, B., Veca, L. M., Xie, S.-Y., 2006. J. Am. Chem. Soc. 128, 7756–7757. Wu, P., Tian, J.C., Walker, C.E., Wang, F.C., 2009. Int. J. Food Sci. Technol. 44, 1671–1676. Yang, X.M., Luo, Y.W., Zhu, S.S., Feng, Y.J., Zhuo, Y., Dou, Y., 2014a. Biosens. Bioelectron. 56, 6–11. Yang, X.M., Zhuo, Y., Zhu, S.S., Luo, Y.W., Feng, Y.J., Dou, Y., 2014b. Biosens. Bioelectron. 60, 292–298. Zhai, X., Zhang, P., Liu, C., Bai, T., Li, W., Dai, L., Liu, W., 2012. Chem. Commun. 48, 7955–7957. Zhang, R., Chen, W., 2014. Biosens. Bioelectron. 55, 83–90. Zheng, H., Wang, Q., Long, Y., Zhang, H., Huang, X., Zhu, R., 2011. Chem. Commun. 47, 10650–10652. Zheng, M., Xie, Z., Qu, D., Li, D., Du, P., Jing, X., Sun, Z., 2013. ACS Appl. Mater. Interfaces 5, 13242–13247. Zhou, L., Lin, Y., Huang, Z., Ren, J., Qu, X., 2012. Chem. Commun. 48, 1147–1149. Zhu, S., Meng, Q., Wang, L., Zhang, J., Song, Y., Jin, H., Zhang, K., Sun, H., Wang, H., Yang, B., 2013. Angew. Chem. Int. Ed. 52, 3953–3957. Zong, J., Yang, X., Trinchi, A., Hardin, S., Cole, I., Zhu, Y., Li, C., Muster, T., Wei, G., 2014. Biosens. Bioelectron. 51, 330–335.