Biosensors and Bioelectronics 62 (2014) 189–195

Contents lists available at ScienceDirect

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

Copper nanoclusters as a highly sensitive and selective fluorescence sensor for ferric ions in serum and living cells by imaging Haiyan Cao a, Zhaohui Chen b, Huzhi Zheng a,n, Yuming Huang a,n a The Key Laboratory of Luminescence and Real-Time Analysis, Ministry of Education, College of Chemistry and Chemical Engineering, Southwest University, Chongqing 400715, China b Basic Department of Rongchang Campus, Southwest University, Chongqing 402460, China

art ic l e i nf o

a b s t r a c t

Article history: Received 25 February 2014 Received in revised form 16 May 2014 Accepted 22 June 2014 Available online 27 June 2014

A simple, one-step facile route for preparation of water soluble and fluorescent Cu nanoclusters (NCs) stabilized by tannic acid (TA) is described. The as-prepared TA capped Cu NCs (TA-Cu NCs) are characterized by UV–vis spectroscopy, Fourier transform infrared (FT-IR) spectroscopy, luminescence, transmission electron microscopy (TEM), and X-ray photoelectron spectroscopy (XPS). The TA-Cu NCs show luminescence properties having excitation and emission maxima at 360 nm and 430 nm, respectively, with a quantum yield of about 14%. The TA-Cu NCs are very stable even in 0.3 M NaCl, and their luminescent properties show pH independent. The fluorescence (FL) of Cu NCs is strongly quenched by Fe3 þ through an electron transfer mechanism, but not by other metal ions. Furthermore, the FL of the TA-Cu NCs shows no changes with the addition of Fe2 þ or H2O2 individually. On this basis, a facile chemosensor was developed for rapid, reliable, sensitive, and selective sensing of Fe3 þ ions with detection limit as low as 10 nM and a dynamic range from 10 nM to 10 μM. The proposed sensor was successfully used for the determination of iron contents in serum samples. Importantly, the Cu NCsbased FL probe showed long-term stability, good biocompatibility and very low cytotoxicity. It was successfully used for imaging ferric ions in living cells, suggesting the potential application of Cu NCs fluorescent probe in clinical analysis and cell imaging. & 2014 Elsevier B.V. All rights reserved.

Keywords: Cu nanoclusters Fluorescence probe Cell imaging Chemosensor Ferric ions

1. Introduction Iron ions play crucial roles in biological systems because ferrous/ferric (Fe2 þ /Fe3 þ ) is one of the important redox pairs for electron transport in the respiratory chain and for a variety of enzymatic reaction in biological systems (Que et al., 2008). The unbalance of iron in human body leads to many diseases (Omara and Blakley, 1993; Allen, 2002). Thus the determination of iron ions in the biological system is very significant. Various analytical methods have been developed for determination of iron ions, including flame atomic absorption spectroscopy (FAAS) (Ajlec and Stupar, 1989), voltammetry (Van den Berg, 2005), colorimetric method (Luan and Burgos, 2012), chemiluminescence (Bowie et al., 2002), and fluorescence detection (Dwivedi et al., 2011; Ho et al., 2012; Bricks et al., 2005). Due to its high sensitivity, the fluorescent probes have been recognized as the efficient molecular tools to help monitor and visualize trace amounts of samples in live cells or tissues. Recently, n

Corresponding authors. Tel./fax: þ86 23 68254843. E-mail addresses: [email protected] (H. Zheng), [email protected] (Y. Huang). http://dx.doi.org/10.1016/j.bios.2014.06.049 0956-5663/& 2014 Elsevier B.V. All rights reserved.

nanoclusters (NCs) are becoming a focus of considerable interest and one of very promising fluorescent probes owing to their special advantages, such as low toxicity, high photoluminescence yield and good biocompatibility, which made them potential applications in single-molecule imaging, biolabeling, sensing, etc. (Schaeffer et al., 2008; Lu and Chen, 2012). However, the previous studies concentrated much effort on gold sub-nanometer clusters. Only in recent years, Ag and Cu NCs have attracted considerable interest due to their unique photoluminescent properties and thus the potential applications in fluorescence analysis (Lu and Chen, 2012). For example, a few studies have been manipulated on fluorescence Cu NCs, using some surface protecting ligands such as DNA (Rotaru et al., 2010; Chen et al., 2012; Zhou et al., 2011), proteins (Goswami et al., 2011), polymers (Kawasaki et al., 2011), and thiols (Yuan et al., 2011). These surface protecting ligands improve their stability and water solubility by providing a protecting layer on the surface of Cu NCs. However, these macromolecular ligands resulted in the formation of Cu NCs with large hydrodynamic radius, which limited the scope of potential applications (Adhikari and Banerjee, 2010; Bao et al., 2010). Most recently, benzotriazole was applied as a template to reach desired smaller size (Salorinne et al., 2012). However, relatively complicated

190

H. Cao et al. / Biosensors and Bioelectronics 62 (2014) 189–195

synthetic procedure and constrained accessibility to the materials greatly limited applications of the as-prepared Cu NCs. Furthermore, careful attention needs to be paid to the synthesis of tiny Cu clusters due to their easy oxidation. Thus, development of a simple and facile strategy for the synthesis of water soluble, extremely stable, and highly quantum efficient fluorescent NCs is highly valuable, but even more challenging. Herein, for the first time, we reported the successful synthesis of Cu NCs stabilized by tannic acid using one-pot method in which CuSO4 was reduced by ascorbic acid. The as-prepared Cu NCs exhibited the blue emission at 430 nm with a high quantum yield (QY) of 14%. It was found that the fluorescence (FL) of Cu NCs is strongly quenched with the addition of Fe3 þ through specific binding of Fe3 þ and tannic acid to form a complex (Ernst and Menashi, 1963). More importantly, H2O2 and other common metal ions, including Fe2 þ , Mg2 þ , Ca2 þ , Mn2 þ , Co2 þ , Ni2 þ , Zn2 þ , Cd2 þ , Pb2 þ , Hg2 þ , Cr3 þ , In3 þ , Na þ , and K þ , have minor effects on the fluorescence of Cu NCs. This indicates that the as-prepared Cu NCs show high selectivity toward Fe3 þ . Based on this new finding, a facile, green, sensitive, and selective chemosensor was developed to detect Fe3 þ in serum samples and image Fe3 þ in living cells, showing great potential in biological applications.

2. Materials and method 2.1. Materials All chemicals and reagents were of analytical grade and used as received without further purification. Tannic acid (denoted TA) was purchased from Qiangshun Chemical Reagent Co., Ltd. (Shanghai, China). FeCl3  6H2O and FeSO4  7H2O were purchased from Signopharm Chemical Reagent Co., Ltd. (Shanghai, China). Ascorbic acid was purchased from Chengdu Kelong Chemistry Reagent Factory (Sichuan, China). CuSO4, NaOH, HCl and citrate acid were obtained from Chongqing Chemical Reagent Company (Chongqing, China). The A549 cell used in the study was purchased from Beijing Dingguo Changsheng Biotechnology Co., Ltd. (Beijing, China). Ultra-pure water was prepared in the lab using a water treatment device. 2.2. Instrumentation The fluorescence spectrum and intensity were recorded on a Hitachi F-7000 Fluorescence spectrophotometer (Tokyo, Japan). The pH of the solutions was measured by a PHS-3D pH meter (Shanghai Precision Scientific Instruments Co., Ltd., China). The UV–visible spectra were measured on a Hitachi U-4100 spectrophotometer (Tokyo, Japan). Dynamic light scattering (DLS) was performed on a Zetasizer Nano-ZS90 (Malvern, UK) instrument for characterization of the size distribution of the Cu NCs in a solution. X-ray photoelectron spectroscopy (XPS) spectra were measured by a XSAM-800 X-ray photoelectron spectrometer (Kratos, UK). The morphology of Cu NCs was observed using transmission electron microscopy (TEM, LIBRA 200, ZEISS) at an acceleration voltage of 200 kV. Cellular images were obtained with an Olympus IX70 inverted fluorescence microscope system, equipped with a 100-W mercury lamp, 40  objective (Olympus) and a U-MWU filter set (330–385/400/420 nm, Olympus, Japan). 2.3. Preparation of copper nanoclusters In a typical preparation process, a solution of CuSO4 (0.2 mL, 0.1 M) and tannic acid (0.1 mL, 1 mM) in water (20 mL) was stirred at room temperature for 5 min. After ascorbic acid (0.2 mL, 1 M) was added to the mixture, the resulting pale blue mixture solution

was stirred for 6 h at 50 °C to obtain yellowish Cu NCs. A dialysis membrane (MWCO: 3500 Da; pore size: ca. 0.35 nm) was then used to separate the Cu NCs from any residual un-reacted species. The as-prepared TA stabilized Cu NCs (denoted TA-Cu NCs) were stored in a refrigerator at 4 °C until use. 2.4. General procedure for fluorescent detection In a typical process, a series of working standard Fe3 þ solutions with different concentrations were added to TA-Cu NCs solution (final concentration of 3 μM). The mixture was mixed well, and then FL spectra measurements and photographs were taken. The FL signal was monitored by a photomultiplier tube (PMT, operated at 700 V) of the Type F-7000 Fluorescence spectrophotometer and was recorded by a computer. The relative fluorescence intensity [(I0  I)/I0] versus Fe3 þ concentration was used for calibration. Here, I0 and I are the fluorescence intensities of the TA-Cu NCs before and after adding analytes, respectively. At every Fe3 þ concentration, the measurement was repeated thrice, and the average FL signal was obtained. 2.5. Procedure for the determination of serum iron Pretreatment of serum samples was adapted from Kyaw (1976), with minor modification. Briefly, 2.0 mL of water was added into the 1.0 mL fresh serum, and then 2.0 mL of 20% trichloroacetic acid solution was added. The mixture was mixed well and put into an oscillator with continuous shaking for 45 min, then centrifuged at 2500 rpm for 15 min. The supernatant was collected and 20 μL of 3% H2O2 was added to 4.0 mL of supernatant to oxidize the Fe2 þ ion to Fe3 þ ion for further analysis. The fluorescent determination of serum sample was performed by the same way as Fe3 þ standard except that Fe3 þ solution was replaced by pretreated serum samples. 2.6. Cell viability assay The cytotoxicity of TA-Cu NCs was examined by the cellcounting kit-8 (CCK-8, Dojindo Laboratories, Japan) assay (Yu et al., 2014; Wang et al., 2013). First, A549 cells were seeded in 96-well plates at a density of 3.25  104 cells mL  1. After 24 h incubation, the medium was then replaced by the medium containing TA-Cu NCs with various concentrations (0, 0.00001, 0.0001, 0.001, 0.01, 0.1, 1, and 10 μM), and the cells were incubated for another 24 h. Next, the cells were washed thrice with phosphate buffered saline (PBS), and freshly prepared CCK-8 (10 μL) solution in culture medium (90 μL) was added to each well. After 1 to 4 h incubation, the CCK-8 medium solution was carefully removed. Then, the plate was gently shaken for 10 min at room temperature, and the optical density (OD) of the mixture was measured at 450 nm. The cell viability was assessed by the following equation:

Cell viability (%) = (Sample OD − Blank OD)/(Control OD − Blank OD) × 100% 2.7. Cell imaging The A549 cells were grown in modified Eagle's medium (MEM) supplemented with 10% fetal bovine serum (FBS) at 37 °C and 5% CO2. Firstly, the cells were washed with PBS, followed by incubating with 0.2 mM TA-Cu NCs in 320 μL of medium for 5 h at 37 °C, and then washing with PBS five times. To detect Fe3 þ in living cells, the cells were incubated with 0.2 mM TA-Cu NCs for 5 h and then stimulated with 0.05 mM Fe3 þ and 0.1 mM Fe3 þ for 1 h,

H. Cao et al. / Biosensors and Bioelectronics 62 (2014) 189–195

respectively. All the concentrations mentioned here are the final values. Cell imaging was then carried out on an Olympus IX70 inverted fluorescence microscope system after the cells were washed with PBS five times.

3. Results and discussion 3.1. Synthesis and characterization of the TA capped Cu NCs Generally, the reduction of metal ions in aqueous solution results in large nanoparticles (NPs) rather than small NCs due to spontaneous aggregation of NCs, usually leading to a significant

191

loss of their properties (Li and Kaner, 2006; Iijima and Kamiya, 2009). Therefore, the stabilization of NCs is a crucial aspect that must be considered during their synthesis from the view point of analytical applications. Furthermore, selection of suitable capping ligands is of key importance for obtaining small and highly fluorescent metal NCs because the nature of the ligands used for capping the particle surface can significantly affect their emission properties (Wu and Jin, 2010; Shang et al., 2011). The TA is selected as capping ligand in the present study because it is one of the important polyphenols with excellent metal-binding capacities, and has proven to be effective in the surface modification of metal NPs (Huang et al., 2010; Yi et al., 2011; Tsai et al., 2004). The Cu NCs were prepared by reducing Cu2 þ to Cu0 clusters with ascorbic

6

Absorbance

5 4 3 2 1

TA TA-Cu NCs

0 200 250 300 350 400 450 500 550 600 650 Wavelength(nm)

O1s

C1s C=C, 284.9 eV

C=O, 531.9 eV C-O, 532.4 eV

Counts

Counts

C=O, 286.6 eV O-C=O, 288.71 eV

276

279

2p3/2

282 285 288 291 Binding Energing(eV)

932.3

525

294

1.0

950.5

Counts

Fluorescence Intensity(a.u.)

2p1/2

925

930

935 940 945 950 Binding Energing(eV)

955

O-C=O, 533.3 eV

960

0.8

528 531 534 537 Binding Energing(eV)

540

a b c d

0.6 0.4 0.2 0.0 300

350

400 450 Wavelength(nm)

500

550

Fig. 1. (A) UV  vis absorption spectra of TA (black line) and TA-Cu NCs solution (red line). (B) TEM image of TA-Cu NCs, Scale bar ¼10 nm. (C) XPS spectrum in the C1s region of TA-Cu NCs. (D) XPS spectrum in the O1s region of TA-Cu NCs. (E) XPS spectrum in the Cu2p region of TA-Cu NCs. (F) FL spectra (excitation spectra, a, b; emission spectra, c, d) of 1 mM TA-Cu NCs solution in the absence (a, c) and presence (b, d) of 0.5 mM Fe3 þ ions. The inset shows the photos of 1 mM TA-Cu NCs solutions in the absence (left) and presence (right) of 0.5 mM Fe3 þ ions illuminated by UV light of 365 nm. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

192

H. Cao et al. / Biosensors and Bioelectronics 62 (2014) 189–195

acid, which is an environment friendly reducing agent (Xiong et al., 2011). From Fig. 1A, the absorption band of the TA-Cu NCs gave three peaks at 230 nm, 268 nm, and 290 nm. This is largely different from the characteristic SPR band of large Cu NPs at 560– 600 nm (Mott et al., 2007; Salzemann et al., 2004) or those of the TA ligands (Fig. 1A). These molecular-like optical transitions absorbance bands in UV–vis absorption spectra are due to the quasi-continuous electronic energy band structure and quantum confinement effects of Cu NCs, which is significantly different from large NPs (Wei et al., 2011; Jin et al., 2005; Rao and Pradeep, 2010). The TEM image (Fig. 1B) showed that the average size of TA-Cu NCs is 2.2 70.5 nm. FT-IR spectroscopy revealed that the TA and TA-Cu NCs showed similar locations and appearance of the major bands (Fig. S1), confirming the successful introduction of TA onto the surface of Cu NCs, which was further confirmed by XPS analysis of the as prepared TA-Cu NCs (Fig. 1C–E). The C1s spectrum (Fig. 1C) shows two peaks at 284.9 and 288.7 eV, which are attributed to C ¼C and C–O ¼O, respectively (Fan et al., 2008). The O1s spectrum (Fig. 1D) exhibits three peaks at 531.9, 532.4, and 533.3 eV, which are attributed to metal–O (Cu–O), surface –OH/C ¼ O, and C–O–C/C– OH groups, respectively (Ma et al., 2013). The oxidation state of copper in samples was studied by XPS. Two intense peaks are located at 932.3 eV and 950.5 eV, which are attributed to 2p2/3 and 2p2/1 features of Cu(0) (Fig. 1E). This finding is consistent with that of a previous study; the intense peak of zero-valent Cu NCs of approximately 2.2 nm in size is located at 932.4 eV in XPS (Brege et al., 2009). Interestingly, a previously reported peak (  934 eV), associated with Cu(II) electrons, was not observed in our study. This suggests the lack of significant oxidation of the Cu NCs. It should be indicated that the presence of Cu þ on the surface of the Cu NCs could not be excluded because the 2p3/2 binding energy of Cu(I) is only  0.1 eV different from that of Cu(0) species (Wei et al., 2011). Therefore, the valence state of the as-prepared TA-Cu NCs most probably lies between 0 and þ1 (Goswami et al., 2011; Wei et al., 2011). Fig. 1F shows the excitation and emission spectra of dilute aqueous TA-Cu NCs solution at room temperature. The spectrum shows an emission maximum at 430 nm upon excitation at 360 nm. Also, excitation at wavelength between 310 nm and 360 nm gives rise to an emission maximum at 430 nm (Fig. S2A). This suggests that no change in emission maximum was observed by altering the excitation wavelength, implying that the obtained emission is a real luminescence from the excited states and it is not due to scattering effects. Tannic acid can be easily oxidized under the basic condition. We compared the fluorescence spectra of TA, TA-NaOH and TA-Cu NCs. The results suggested that the strong blue emission arose from TA-Cu NCs, rather than from TA or the oxidized product of TA (Fig. S2B). The emission wavelength of TA-Cu NCs is nearly equal to that reported for 2-mercapto-5-npropylpyrimidine-stablized Cun (n≦8) cluster (Wei et al., 2011). The QY of Cu NCs was determined to be 14%, using quinine sulfate as the reference, which is much higher than that of the previously reported fluorescent Cu NCs (Wei et al., 2011; Vilar-Vidal et al., 2010). The as prepared Cu NCs are very stable. This is probably because the TA protects the Cu NCs from aggregation even in the presence of a given high concentration of salt (0.3 M NaCl) (Fig. S3A). Also, the emission peak remains almost unchanged at pH values from 2 to 9 (Fig. S3B), indicating that the Cu NCs could work in a wide range of pH. They are found to be very stable, showing the same emission spectra after 60 days of storage at room temperature (Fig. S3C). Temperature-dependent fluorescence of TA-Cu NCs has been measured to investigate the thermal stability of Cu NCs. It can be seen from Fig. S3D that the emission of the Cu NCs in an aqueous solution slightly decreased upon increase in

temperature above 50 °C. This may be due to partially losing of protecting groups from the surface of Cu NCs. These results show that the TA-Cu NCs promise long-term storage stability, high concentration of salt tolerance and a wide range of pH tolerance, and lower toxicity relative to most semiconductor toxic quantum dots. Importantly, compared to Au NCs and Ag NCs, the TA-Cu NCs are significantly cheaper and can be widely used in analytical chemistry. Furthermore, the preparation of Cu NCs is free of the toxic reductants (e.g., NaBH4) and does not require complicated reactions. 3.2. FL response of the TA-Cu NCs to Fe3 þ and FL quenching mechanism Our approach was based on hypothesis that Fe3 þ easily combined with the surface of Cu NCs via forming the complex with tannic acid of the TA-Cu NCs (Ernst and Menashi, 1963), resulting in a strong FL quenching of the TA-Cu NCs possibly via an electron transfer mechanism. As can be seen from Fig. 1F, upon adding Fe3 þ , the strong blue emission of the TA-Cu NCs solution can be quenched obviously. The corresponding FL spectra indicate that the addition of Fe3 þ ions decreases the FL intensity of the TA-Cu NCs, but has no effect on the FL wavelength (Fig. 1F). On the other hand, the emission of TA-Cu NCs shows minimal spectral overlap with the absorption of Fe3 þ ions (Fig. S4). Hence, resonance energy transfer would not be the dominant mechanistic pathway. It has been reported that the presence of a hydrophilic BSA helped stabilize the Cu quantum clusters (QCs). Because BSA contains a high-affinity site (carboxylate group) for Pb2 þ ion, the luminescence quenching in the presence of Pb2 þ occurs due to the QCs aggregation induced by the complexation between BSA and the Pb2 þ ion (Goswami et al., 2011). However, this proposed mechanism was contradictory to the results of our DLS measurements (Fig. S5). Clearly, the addition of Fe3 þ ions from 0.1 to 8 μM does not change the hydrodynamic diameter of the TA-Cu NCs significantly. No aggregation was observed for all these samples. It can thus be concluded that the reduction in fluorescence intensity was not caused by aggregation of TA-Cu NCs into larger particles. An electron transfer mechanism is probably accounting for the FL quenching. Our results suggest that the TA-Cu NCs show a high selectivity towards other common metal-ions, including alkali, alkaline earth, and transitional metal ions (see later selectivity discussion). These distinctive responses should be due to the strong electron-accepting ability of Fe3 þ , and thus Fe3 þ can capture electrons more easily. The outer electronic structure of Fe3 þ is 3d54s0, and the five d orbits are half-filled. This leads to its relatively high charge density with the stronger electron-withdrawing ability (Li et al., 2013). Upon addition of Fe3 þ ions, they can be quickly adsorbed onto the surface of TA-Cu NCs through forming the complexes between electron-deficient Fe3 þ and electron-rich TA (Ernst and Menashi, 1963), leading to the static FL quenching of TA-Cu NCs. This has been confirmed by a large association constant between Fe3 þ and TA-Cu NCs. In the concentration range from 0.01 to 0.1 μM Fe3 þ , a classic Stern–Volmer equation, I0/I ¼1.02 þ6.36  105 [Fe3 þ ], r ¼ 0.9902, can be fit, giving a large association constant of 6.36  105 M  1 (Fig. S6A). This clearly suggests a highly effective static quenching of Fe3 þ by the FL emission of TA-Cu NCs. The association constant between Cr3 þ and TA-Cu NCs was also estimated. As can be seen from Fig. S6B, a classic Stern–Volmer equation is I0/I¼ 1.01 þ499.03 [Cr3 þ ] (r ¼0.9922), giving an association constant of 4.99  102 M  1. We measured the FL lifetimes of the TA-Cu NCs solutions before and after addition of Fe3 þ using a time-correlated single photon counting (TCSPC) technique. As can be seen from Table S1, both the TA-Cu NCs and TA-Cu NCs Fe3 þ systems show three lifetimes. The average lifetimes of the TA-Cu NCs Fe3 þ system showed no

H. Cao et al. / Biosensors and Bioelectronics 62 (2014) 189–195

dependence of Fe3 þ concentrations, clearly suggesting the static FL quenching of TA-Cu NCs by Fe3 þ . The effect of the TA-Cu NCs size on their FL and quenching properties was investigated. In order to do this, three Cu NCs with different sizes were synthesized (Fig. S7). As can be seen from Fig. S8, the FL intensity decreased obviously with increasing size of the TA-Cu NCs. From Table S2, the quenching constant decreased with increasing size of the TA-Cu NCs, showing that the quenching abilities decreased with increasing size of the TA-Cu NCs.

monitoring the FL intensity as a function of time. It is clear that only Fe3 þ can significantly decrease the FL intensity of the TA-Cu NCs (Fig. 2B). The quenching equilibrium reaches in less than 5 min, showing fast interaction between the TA-Cu NCs and Fe3 þ . The FL signal remained almost constant with increasing reaction time to 30 min. Therefore, 5 min reaction time was selected in subsequent experiments. Finally, the effect of the TA-Cu NCs concentration was investigated over the range of 0.5–10 μM (Fig. S9). It was found that the relative FL intensity [(I0 I)/I0] increased with increase of the TA-Cu NCs concentration up to 3 μM, above which it decreased. Finally, 3 μM of the TA-Cu NCs was chosen. Under the selected conditions, a quantitative analysis of Fe3 þ was performed to test the sensitivity. The FL intensity of the TA-Cu NCs displayed a gradual decrease at 430 nm with increasing Fe3 þ concentration (Fig. 3). Plotting the value of (I0  I)/I0 versus the concentrations of Fe3 þ gave a calibration curve (inset in Fig. 3). Fe3 þ can be detected as low as 10 nM (with an S/N ratio of 3). A linear relationship was observed with Fe3 þ concentrations from 0.01 to 10 μM with two ranges (Fig. 3). A good linear relationship was found in the studied concentration range of Fe3 þ . To test the selectivity of the developed method for Fe3 þ , the influence of various metal ions and anions was tested, including Fe2 þ , Mg2 þ , Ca2 þ , Mn2 þ , Co2 þ , Ni2 þ , Zn2 þ , Cd2 þ , Pb2 þ , Hg2 þ , Cr3 þ , In3 þ , Al3 þ , Na þ , K þ , NH4 þ , F  , Cl  , Br  , NO3  , SO42  , CO32  , ClO3  , SO32  , PO43  and Ac  . The results indicated that all the tested ions except Al3 þ have no obvious effects on the FL emission of Cu NCs. However, when using 0.2 mM NaF as masking reagent, the interference derived from Al3 þ can be effectively

3.3. Fluorescence detection of Fe3 þ with the TA-Cu NCs Based on the Fe3 þ -induced quenching of the TA-Cu NCs, the capability of the TA-Cu NCs to allow the highly selective and sensitive detection of Fe3 þ was investigated. To obtain the optimal analytical conditions for the proposed method, we explored the effect of pH in the range of 2.0 11.0 on the FL intensities of the TA-Cu NCs in the absence and presence of Fe3 þ and the kinetic behavior of the reaction. As shown in Fig. 2A, in the absence of Fe3 þ , the FL intensity of the TA-Cu NCs kept almost stable over a wide pH range from 2.0 to 9.0. However, the quenched FL intensity of the TA-Cu NCs is pH-dependent in the presence of Fe3 þ . In acidic media (pHr6.0), the quenching efficiencies increased with pH from 2.0 to 4.0, then remained stable from 4.0 to 6.0, and finally decreased with pH 6.0 to 9.0 (Fig. 2A). Finally, the weakly acid media (pH 6.0) was chosen for the sensitive detection of Fe3 þ . The kinetic behaviors of reactions between the TA-Cu NCs and Fe3 þ , as well as between the TA-Cu NCs and Fe2 þ , were studied by

2.0

2+

1.5

(I0-I)/I0

1.0

1.0

0.5

0.5

4

TA-Cu NCs+Fe +H2O2; 3+

1.5

(I0-I)/I0 value

2

TA-Cu NCs; TA-Cu NCs+Fe 3+ TA-Cu NCs+Fe ; TA-Cu NCs+H2O2

1.2

TA-Cu NCs 3+ TA-Cu NCs+Fe

6 pH

8

10

12

Fluorescence Intensity(a.u.)

Fluorescence Intensity(a.u.)

2.0

0.0

193

0.0

2+

TA-Cu NCs+Fe +H2O2

0.9

0.6

0.3

0

5

10

15 20 Time(min)

25

30

Fig. 2. (A) FL responses of 3 μM TA-Cu NCs at 430 nm in the absence (black triangle) and presence (black square) of 10 μM iron ions at different pH values. Relative FL intensities [(I0  I)/I0] at 430 nm of solutions of TA-Cu NCs in the presence (blue) of 10 μM Fe3 þ ions at different pH values. (B) Kinetics of reactions between the TA-Cu NCs and Fe3 þ or Fe2 þ in the absence and presence of H2O2. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fluorescence Intensity(a.u.)

Fluorescence Intensity/(a.u.)

0.08

600

0.06

500

0.04

400

0.02

300

0.00

200

0.02

0.04 0.06 0.08 0.10 Fe concentration(μM)

100 0

600

0.6

500

0.4

400

0.2

300

0.0

200

0

2 4 6 8 10 Fe concentration(μM)

100

400

450

500 Wavelength(nm)

550

600

0

400

450

500

550

600

Wavelength(nm)

Fig. 3. FL response of 3 μM TA-Cu NCs upon addition of various concentrations of Fe3 þ ions solution. The arrows indicate the signal changes as increases in analyte concentrations (A, 0.01, 0.02, 0.04, 0.06, and 0.1 μM; (I0  I)/I0 ¼ 0.5702 [Fe3 þ ] þ0.0224, r2 ¼ 0.9774; B, 0.1, 0.2, 0.6, 0.8, 1, 2, 4, 6, 8, and 10 μM; (I0  I)/I0 ¼ 0.0528 [Fe3 þ ] þ 0.088, r̄ 2 ¼ 0.995). Inset. Plots of the values of (I0  I)/I0 at 430 nm versus the concentrations of Fe3 þ . The error bars represent standard deviations based on three independent measurements.

194

H. Cao et al. / Biosensors and Bioelectronics 62 (2014) 189–195

120

0.4

100

0.3

80

Cell Viability (%)

(I0-I)/I0

0.5

0.2 0.1 0.0

+

+

+

2+

2+

+

+

3+

-

-

+

+

2+

+

-

-

-

3 2 2 3 2 2 2 2 22+ l 2r Fe Fe Mn Co 2+ Zn Cd Pb Hg Cr PO 4+ F 3+ - B K Na Ca Mg O 4 C O 3 + AcCO 3 - SO 3 3 + In F 3 N S Ni ClO NH 4 Al

10 μM

1000 μM

60 40 20 0 0

Fig. 4. Relative fluorescence intensities [(I0  I)/I0] at 430 nm of the TA-Cu NCs solution after addition of 10 μM Fe3 þ and various other ions (10 μM or 1000 μM).

eliminated (Fig. 4). Thus, the proposed method was highly selective for Fe3 þ over the other ions. As indicated previously, the high selectivity of TA-Cu NCs for Fe3 þ is probably due to its relatively high charge density with the stronger electron-withdrawing ability (Li et al., 2013) in comparison with other metal ions. Based on Li et al., the charge density of the tested metal ions follows the order of Al3 þ 4 Fe3 þ 4Cr3 þ »Ni2 þ 4 Co2 þ 4Mn2 þ 4Ca2 þ 4 Hg2 þ 4Na þ 4K þ (Li et al., 2013). Fe3 þ has a high charge density. The outer electronic structure of Fe3 þ is 3d54s0 (Scheme S1), and the five d orbits are half-filled. This leads to its stronger electronwithdrawing ability (Li et al., 2013). Mn2 þ also has a complete set of partially filled outer d orbitals; however, its charge density and standard electrode potential (  1.18 V, Table S3) are both relatively low. Although the standard electrode potential of Hg2 þ /Hg þ (0.91 V) is higher than that of Fe3 þ /Fe2 þ (0.77 V), Hg2 þ has a low charge density and no partially filled outer d orbitals. Note that although Cr3 þ has an incomplete set of partially filled outer d orbitals, its charge density and standard electrode potential (  0.41 V) are both much lower than that of Fe3 þ . Thus, among the investigated transition metal ions, only Fe3 þ can effectively withdraw electrons from the TA-Cu NCs. Therefore, upon addition of Fe3 þ ions, they can be quickly adsorbed onto the surface of TACu NCs through forming the complexes between electron-deficient Fe3 þ and electron-rich TA (Ernst and Menashi, 1963), leading to the static FL quenching of TA-Cu NCs. It should be indicated that Al3 þ has a high charge density, thus showing the effect. However, it does not possess any partially filled outer d orbitals. Furthermore, the standard electrode potential of Al3 þ is very low (  1.66 V, Table S3). The possible reason for Al3 þ -induced effect is that Al3 þ is one of main group metal ions and its electrical nature is quite different from that of transition-metal ions. However, it should be indicated that further research is needed to explore the detail signal-generation mechanism.

-5

10

-4

-3

-2

-1

10 10 10 10 1 TA-Cu NCs Concentration(μM)

10

Fig. 5. The viability of A549 cells after being incubated with TA-Cu NCs in the concentrations range from 0 to 10 μM. The error bars represent standard deviations based on three independent measurements.

from Al3 þ . As can be seen from Table S4, the results obtained by the proposed method agree well with those obtained by atomic absorption spectrometry, showing great potential of the sensor for the routine estimation of ion content in clinical samples. 3.5. Application in cell imaging To demonstrate the feasibility of the TA-Cu NCs fluorescent probe for cell imaging, we performed fluorescence imaging of the selectivity of TA-Cu NCs to Fe3 þ using the A549 cells as the biological test model. The cytotoxicity of TA-Cu NCs was firstly evaluated by CCK-8 assays. As seen in Fig. 5, almost 100% of A549 cells are alive in the presence of TA-Cu NCs. Thus, TA-Cu NCs show very low cytotoxicity and are suitable for imaging studies. For cell imaging, the A549 cells were first incubated with 0.2 mM TA-Cu NCs for 5 h, and then washed with PBS five times. As seen in Fig. 6B, all the cells exhibit the clear cell morphology and light blue fluorescence which is very stable and does not decay under the UV lamp illumination for more than 30 min, suggesting that TA-Cu NCs can be transfected into the living cells and applied for fluorescence imaging. When the cells were incubated with 0.2 mM TA-Cu NCs for 5 h and then stimulated with 0.05 mM Fe3 þ and 0.1 mM Fe3þ for 1 h, a decrease in the intracellular fluorescence was observed (Fig. 6C and D). And the higher the concentration of Fe3 þ , the lower the fluorescence observed. The negative control also exhibit weak blue fluorescence under UV illumination, but the fluorescence intensity is the lowest among these four images (Fig. 6A). The above results demonstrate that the TA-Cu NCs probe can image the cell with good biocompatibility, and positively support the TA-Cu NCs as a promising potential fluorescent probe for cell imaging.

3.4. Practical application in serum samples To evaluate the utility of the Fe3 þ -responsive TA-Cu NCs FL probes, the analysis of the serum iron in clinical samples has been tested. Trichloroacetic acid was used to split off iron from proteins and hydrogen peroxide was used to oxidize Fe2þ into Fe3þ (Kyaw, 1976). Then the pretreated serum sample was mixed with the Cu NCs and the fluorescence intensity was measured after 10 min incubation. Ten serum samples from the Ninth People's Hospital of Chongqing were detected and the results are shown in Table S4. Among them, four serum samples were from the uremic patients on hemodialysis with enhanced aluminum levels. The concentration of aluminum in healthy person is less than 1 μM, whereas it ranges from 2.04 to 4.19 μM in these patients measured by flameless atomic absorption spectrophotometer (Table S4). Therefore, the 0.2 mM NaF was added and used as masking reagent to eliminate the interference derived

4. Conclusion In summary, we have developed a simple and facile strategy for the synthesis of Cu NCs using tannic acid as a stabilizing reagent. These Cu NCs are water soluble, extremely stable, and highly quantum efficient. The luminescence of as-prepared Cu NCs can be used to construct a highly sensitive and selective fluorescence sensor for detecting ferric ions. The results from the serum samples suggest the applicability of Cu NCs for the analysis of iron contents in real clinical samples. Furthermore, the Cu NCs show very low cytotoxicity and have been successfully used for imaging ferric ions in living cells, showing great potential of fluorescent Cu NCs probe for cell imaging. The novel synthetic strategy and the new metal ion sensing protocol described herein

H. Cao et al. / Biosensors and Bioelectronics 62 (2014) 189–195

195

Fig. 6. Fluorescence images of A549 cells for the identifications of Fe3 þ . (A) Untreated control A549 cells, (B) A549 cells incubated with TA-Cu NCs, (C) A549 cells incubated with TA-Cu NCs in the presence of 0.05 mM Fe3 þ , and (D) A549 cells incubated with TA-Cu NCs in the presence of 0.1 mM Fe3 þ . All the concentrations are the final values. Scale bar¼12 μm.

may open a new window of interest in developing a nanoscale platform for diverse biological analytical applications.

Acknowledgments Financial support from the National Natural Science Foundation of China (Grant 21277111) and the Fundamental Research Funds for the Central Universities (XDJK2013D001 and XDJK2013A022) is gratefully acknowledged. We thank Ms. Ping Teng (Southwest University) for her help with the experimental measurements.

Appendix A. Supplementary information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.bios.2014.06.049.

References Adhikari, B., Banerjee, A., 2010. Chem. Mater. 22, 4364–4371. Ajlec, R., Stupar, J., 1989. Analyst 114, 137–142. Allen, L.H., 2002. J. Nutr. 132, 813s–819s. Bowie, A.R., Achterberg, E.P., Sedwick, P.N., Ussher, S., Worsfold, P.J., 2002. Environ. Sci. Technol. 36, 4600–4607. Brege, J.J., Hamilton, C.E., Crouse, C.A., Barron, A.R., 2009. Nano Lett. 9, 2239–2242. Bricks, J.L., Kovalchuk, A., Trieflinger, C., Nofz, M., Büschel, M., Tolmachev, A.I., Daub, J., Rurack, K., 2005. J. Am. Chem. Soc. 127, 13522–13529. Bao, Y., Yeh, H.C., Zhong, C., Ivanov, S.A., Sharma, J.K., Neidig, M.L., Vu, D.M., Shreve, A.P., Dyer, R.B., Werner, J.H., Martinez, J.S., 2010. J. Phys. Chem. C 114, 15879–15882. Chen, J., Liu, J., Fang, Z., Zeng, L., 2012. Chem. Commun. 48, 1057–1059. Dwivedi, A.K., Saikia, G., Iyer, P.K., 2011. J. Mater. Chem. 21, 2502–2507. Ernst, Z.L., Menashi, J., 1963. Trans. Faraday Soc. 59, 2838–2844. Fan, X.B., Peng, W.C., Li, Y., Li, X.Y., Wang, S.L., Zhang, G.L., Zhang, F.B., 2008. Adv. Mater. 20, 4490–4493. Goswami, N., Giri, A., Bootharaju, M.S., Xavier, P.L., Pradeep, T., Pal, S.K., 2011. Anal. Chem. 83, 9676–9680.

Ho, J.A., Chang, H.C., Su, W.T., 2012. Anal. Chem. 84, 3246–3253. Huang, K.W., Yu, C.J., Tseng, W.L., 2010. Biosens. Bioelectron. 25, 984–989. Iijima, M., Kamiya, H., 2009. Kona Powder Part. J. 27, 119–129. Jin, R.C., Jureller, J.E., Kim, H.Y., Scherer, N.F., 2005. J. Am. Chem. Soc. 127, 12482– 12483. Kawasaki, H., Kosaka, Y., Myoujin, Y., Narushima, T., Yonezawa, T., Arakawa, R., 2011. Chem. Commun. 47, 7740–7742. Kyaw, A., 1976. Clin. Chim. Acta 69, 351–354. Li, D., Kaner, R.B., 2006. J. Am. Chem. Soc. 128, 968–975. Li, X., Liao, Y., Huang, M., Strong, V., Kaner, R.B., 2013. Chem. Sci. 4, 1970–1978. Luan, F., Burgos, W.D., 2012. Environ. Sci. Technol. 46, 11995–12002. Lu, Y.Z., Chen, W., 2012. Chem. Soc. Rev. 41, 3594–3623. Ma, J., Zhu, Z.L., Chen, B., Yang, M.X., Zhou, H.M., Li, C., Yu, F., Chen, J.H., 2013. J. Mater. Chem. A 1, 4662–4666. Mott, D., Galkowski, J., Wang, L.Y., Luo, J., Zhong, C.J., 2007. Langmuir 23, 5740–5745. Omara, F.O., Blakley, B.R., 1993. J. Nutr. 123, 1649–1655. Que, E.L., Domaille, D.W., Chang, C.J., 2008. Chem. Rev. 108, 1517–1549. Rao, T.U.B., Pradeep, T., 2010. Angew. Chem. Int. Ed. 49, 3925–3929. Rotaru, A., Dutta, S., Jentzsch, E., Gothelf, K., Mokhir, A., 2010. Angew. Chem. Int. Ed. 49, 5665–5667. Salorinne, K., Chen, X., Troff, R.W., Nissinen, M., Hakkinen, H., 2012. Nanoscale 4, 4095–4098. Salzemann, C., Lisiecki, I., Brioude, A., Urban, J., Pileni, M.P., 2004. J. Phys. Chem. B 108, 13242–13248. Schaeffer, N., Tan, B., Dickinson, C., Rosseinsky, M.J., Laromaine, A., McComb, D.W., Stevens, M.M., Wang, Y., Petit, L., Barentin, C., Spiller, D.G., Cooper, A.I., Levy, R., 2008. Chem. Commun. 34, 3986–3988. Shang, L., Dong, S., Nienhaus, G.U., 2011. Nano Today 6, 401–418. Tsai, C.Y., Lee, D.S., Tsai, Y.H., Chan, B., Luh, T.Y., Chen, P.J., Chen, P.H., 2004. Mater. Lett. 58, 2023–2026. Van den Berg, C.M.G., 2005. Anal. Chem. 78, 156–163. Vilar-Vidal, N., Blanco, M.C., Lopez-Quintela, M.A., Rivas, J., Serra, C., 2010. J. Phys. Chem. C 114, 15924–15930. Wang, Q., Huang, X., Long, Y., Wang, X., Zhang, H., Zhu, R., Liang, L., Teng, P., Zheng, H., 2013. Carbon 59, 192–199. Wei, W.T., Lu, Y.Z., Chen, W., Chen, S.W., 2011. J. Am. Chem. Soc. 133, 2060–2063. Wu, Z., Jin, R., 2010. Nano Lett. 10, 2568–2573. Xiong, J., Wang, Y., Xue, Q.J., Wu, X.D., 2011. Green Chem. 13, 900–904. Yi, Z., Li, X.B., Xu, X.B., Luo, B.C., Luo, J.S., Wu, W.D., Yi, Y.G., Tang, Y.J., 2011. Colloids Surf. A 392, 131–136. Yuan, X., Luo, Z.T., Zhang, Q.B., Zhang, X.H., Zheng, Y.G., Lee, J.Y., Xie, J.P., 2011. ACS Nano 5, 8800–8808. Yu, C., Zhu, Z., Wang, L., Wang, Q., Bao, N., Gu, H., 2014. Biosens. Bioelectron. 53, 142–147. Zhou, Z., Du, Y., Dong, S., 2011. Anal. Chem. 83, 5122–5127.