Direct chemiluminescence of fluorescent gold nanoclusters with classic oxidants for hydrogen peroxide sensing

2.1 Apparatus

CL measurements and data acquisition were carried out on a model IFFM-D flow injection CL analyser (Xi’an Remex Analytic Instrument Co. Ltd., China), which was equipped with a CR105 photomultiplier tube (Hamamatsu Photonics Co. Ltd., China). CL spectrum was obtained by means of a series of interference filters (Institute Biophysics of Chinese Academy of Science, China), which was set between the reaction cell and the window of the photomultiplier tube. UV–vis absorption spectra were scanned on a UV-1800 spectrophotometer (Shimadzu, Japan) with a slit width of 2 nm by using 1 cm cuvette. Fluorescence spectrum was taken on an F-2700 fluorescence spectrophotometer (Hitachi, Japan). Transmission electron microscopy (TEM) image was measured on a JEM-2100 transmission electron microscope (Japan Electronic Company, Japan) at an accelerating voltage of 200 kV. The X-ray photoelectron spectrum (XPS) was measured on an X-ray photoelectron spectrometer (Ultra DLD, Kratos, Britain) using Al-Kα as the exciting source (1486.6 eV) and binding energy calibration was based on C 1s at 284.8 eV.

2.2 Chemicals

All chemicals were of analytical grade; the water was obtained from an AXLC1805 ultrapure water purification system (Beijing ASTK Technology Development Co., Ltd). Chloroauric acid was purchased from Shanghai Chemical Reagent Company, China. Bovine serum albumin (BSA) was purchased from Beijing Xinjingke Biotechnology Company, China. Ammonium cerium (IV) nitrate and N-bromosuccinimide were obtained from Tianjin Fuchen Chemical Reagents Factory, China. Potassium permanganate, potassium ferricyanide, hydrogen peroxide (30%), and other reagents were purchased from Xi’an Chemical Reagent Company, China.

2.3 Synthesis of BSA-stabilized Au NCs

All glassware was cleaned with aqua regia and then rinsed with ultrapure water. BSA-stabilized Au NCs were prepared according to the literature with minor modification (Xie et al. 2009). Briefly, HAuCl₄ solution (1.0 × 10⁻³ mol L⁻¹, 10 mL, 37 °C) was mixed with BSA solution (5.0 × 10⁻² g mL⁻¹, 10 mL, 37 °C). After stirring for 2 min, NaOH solution (1.0 mol L⁻¹, 1.0 mL) was added into the mixture. The reaction was allowed to perform at 37 °C for 12 h under stirring. During this period, the colour of the solution gradually changed from light yellow to deep brown.

2.4 Batch CL measurement

Into CL reaction cell, 0.5 mL of oxidant solution and 0.5 mL of reaction medium were added. Then 0.5 mL of the as-prepared Au NCs solution in the dilution of 1:20 was injected to initiate the CL reaction. The CL signal was detected with the photomultiplier tube biasing at high voltage of 600 V.

2.5 Flow injection CL measurements

Fig. 1 shows the schematic diagram of CL flow system. The Au NCs solution (50 μL) was injected into carrier H₂O via syringe, which was then combined with the emerged stream of KMnO₄ and diluted H₂SO₄. The CL signal produced in the flow cell was detected by the photomultiplier tube biasing at high voltage of 800 V.

Close

Figure 1

Schematic diagram of CL flow system. P1, P2: peristaltic pump; PMT: photomultiplier tube; HV: high voltage; COM: computer.

Export to PPT

For the detection of H₂O₂, the carrier was replaced with a series of H₂O₂ standards with the concentration range from 1.0 × 10⁻⁶ to 1.0 × 10⁻⁴ mol L⁻¹. The CL intensity in the absence of H₂O₂ was recorded as I₀ and in the presence of H₂O₂ as I. The decreased CL intensity (ΔI) was calculated by subtracting I from I₀. The calibration curve for H₂O₂ was constructed by plotting ΔI to the concentration of H₂O₂.

3.1 CL of BSA-stabilized Au NCs with classic oxidants

The BSA-stabilized Au NCs were prepared following the previous work of Xie et al. (2009). The as-prepared Au NCs exhibit a red emission peak at 610 nm upon excitation at 490 nm (Fig. S1). The size of Au NCs is less than 2 nm proved by TEM image (Fig. S2). When Au NCs solution was injected into mixed solution of oxidant with reaction medium, strong CL signal was recorded (Fig. 2A). The CL intensities were in the orders of KMnO₄ > NBS > K₃Fe(CN)₆ > H₂O₂ > Ce(IV). Acidic KMnO₄–Au NCs CL reaction was selected to further investigate for its producing highest CL signal. The acidic KMnO₄–Au NCs CL reaction is a fast process; the CL intensity reached a maximum after ∼1.25 s upon injection of Au NCs. To rule out the possible CL from the reaction between acidic KMnO₄ and the reactants used for preparation of Au NCs, control experiments were conducted. As shown in Fig. 2B, no measureable CL signal was recorded for the reaction of acidic KMnO₄ with Au(III). And for the reaction between acidic KMnO₄ and BSA, only a very weak CL was observed which was much lower than that of acidic KMnO₄ with Au NCs. Accordingly, the CL emission was attributed to the reaction between acidic KMnO₄ and Au NCs not from other concomitants.

Close

Figure 2

(A) CL profiles of Au NCs with different oxidants. 1:20 Au NCs. The reaction conditions are 5.0 × 10⁻⁴ mol L⁻¹ KMnO₄ in 0.05 mol L⁻¹ H₂SO₄, 1.0 × 10⁻³ mol L⁻¹ Ce(IV) in 0.1 mol L⁻¹ H₂SO₄, 0.1 mol L⁻¹ H₂O₂ in 0.1 mol L⁻¹ NaOH, 0.01 mol L⁻¹ NBS in 0.5 mol L⁻¹ NaOH, and 0.01 mol L⁻¹ K₃Fe(CN)₆ in 0.5 mol L⁻¹ NaOH. (B) CL of acidic KMnO₄ with Au NCs, BSA, and Au(III), respectively. 1:20 Au NCs. 2.2 × 10⁻⁴ mol L⁻¹ Au(III) and 1.09 × 10⁻³ g mL⁻¹ BSA which were the same amounts to Au NCs in the dilution of 1:20.

Export to PPT

3.2 CL mechanism of BSA-stabilized Au NCs with acidic KMnO₄ reaction

The identification of luminophor is of significant importance for a new CL reaction. The CL spectrum of the reaction was acquired by using a series of filters. The CL reaction had one emission band locating at around 640 nm (Fig. 3). Obviously, Au NCs were not the luminophor of the reaction since the CL spectrum of acidic KMnO₄–Au NCs reaction was remarkably different with the fluorescence spectrum of Au NCs. Barnett et al. (2002) and Hindson et al. (2010) have thoroughly investigated the mechanism of acidic permanganate CL system and suggested the excited state Mn(II)^∗ as the possible luminophor which emits emission around 610–660 nm, assigning to the phosphorescence of Mn(II) (from 4T1 to 6A1 transition). The CL spectrum of this reaction matched well with that of the excited state Mn(II)^∗. Thus, the excited state Mn(II)^∗ was suggested as the potential luminophor for this system.

Close

Figure 3

The CL spectrum for the reaction of Au NCs with acidic KMnO₄. 5.0 × 10⁻⁴ mol L⁻¹ KMnO₄ in 0.05 mol L⁻¹ H₂SO₄, 1:20 Au NCs. The error bars represent standard deviations of three replicate measurements.

Export to PPT

When the Au NCs solution was mixed with acidic KMnO₄ solution, the characteristic absorption band of KMnO₄ around 530 nm disappeared (Fig. 4), implying that KMnO₄ was reduced by Au NCs during reaction. Meanwhile, the purple of KMnO₄ solution faded and the colour of the mixture turned into light yellow matching well with the colour of Au(III) solution. Thus, it was deduced that during the reaction Au NCs were oxidized to Au(III), while Mn(VII) was readily reduced to Mn(II) in acid medium. To further identify the oxidation state of gold and manganese after the reaction, XPS experiments were performed. Prior to XPS measurement, the mixture of Au NCs with KMnO₄ in 4.0 mol L⁻¹ HNO₃ (v/v = 3:1), KMnO₄, MnSO₄ powder was treated via vacuum-drying for 2 h and the Au NCs were treated by freeze-drying for two days (Zhang et al. 2006). Fig. 5A showed that the Au 4f7/2 band taken from Au NCs located at 84.5 eV, confirming the presence of Au(0) core in Au NCs (Jiang et al. 2014). After the reaction, the Au 4f7/2 band emerged at higher core binding energy (86.3 eV) in the range of Au(III) species (Odio et al. 2014), confirming the formation of Au(III) after the reaction. And for the mixture of KMnO₄ with Au NCs, the peak of Mn2p binding energy assigned to Mn(VII) species (642.6 eV) was absent (Fig. 5B), while the band of Mn2p binding energy taken from the mixture of KMnO₄ with Au NCs was in good agreement with that of MnSO₄ standard, indicating that Mn(VII) species was reduced to Mn(II) after the reaction.

Close

Figure 4

UV–Vis spectra before and after the reaction. (a) 1.0 × 10⁻⁴ mol L⁻¹ KMnO₄ in 0.01 mol L⁻¹ H₂SO₄, (b) 1:50 Au NCs, and (c) the mixture of a and b. The inset is the photographs of acidic KMnO₄, Au NCs and the mixture.

Export to PPT

Close

Figure 5

XPS for the (A) Au4f binding energy before and after the CL reaction and (B) Mn2p binding energy in KMnO₄, MnSO₄ and the mixture.

Export to PPT

Based on the discussion above, the CL mechanism of the reaction may be attributed, therefore, to the following reactions. KMnO₄ was rapidly reduced by Au NCs in an acid medium to produce the excited state Mn(II)^∗, which returns to the ground state accompanying with CL emission around 640 nm, as illustrated in Scheme 1.

Close

Scheme 1

Schematic illustration of CL mechanism for acidic KMnO₄–Au NCs reaction.

Export to PPT

3.3 Effect of experimental variables

To achieve the maximum CL response, the effect of experimental variables, including Au NCs concentration, KMnO₄ concentration, and H₂SO₄ concentration was investigated. The CL intensity was strongly dependent on the concentration of the Au NCs employed (Fig. S3A). The higher the Au NCs concentrations the higher the CL signals. The CL intensity increased with KMnO₄ concentration up to 5.0 × 10⁻⁴ mol L⁻¹ and then decreased with further increase in KMnO₄ concentration (Fig. S3B). This observation was reported in many acidic KMnO₄-based CL system. Most KMnO₄ CL systems occur in acidic media and H₂SO₄ is a popular one. The effect of H₂SO₄ concentration was examined in the range 1.0 × 10⁻³–0.15 mol L⁻¹ (Fig. S3C). The CL intensity had a maximum at 0.05 mol L⁻¹ H₂SO₄. Higher or lower concentration of H₂SO₄ caused a decrease in the CL intensity.

3.4 Response of the system to hydrogen peroxide

The detection of hydrogen peroxide (H₂O₂) attracts a great deal of attention due to its involvement in many chemical, biological, and environmental processes. Particularly, H₂O₂ is one of the products of enzymatic reactions, thus enabling quantitative assays of enzyme substrates as well as enzyme activity. We here found that the CL signal of acidic KMnO₄–Au NCs system was significantly decreased in the presence of H₂O₂ (Fig. 6A). A linear relationship between the decreased CL intensity and H₂O₂ concentration was achieved in the range of 1.0 × 10⁻⁶–1.0 × 10⁻⁴ mol L⁻¹ (Fig. 6B). The detection limit, three times standard deviation of blank, was calculated to be 5 × 10⁻⁷ mol L⁻¹. The method exhibited good precision with a relative standard deviation of 1.4% for 1.0 × 10⁻⁵ mol L⁻¹ hydrogen peroxide in 11 replicated measurements.

Close

Figure 6

(A) Response to different concentration of H₂O₂ and (B) the linear relationship between the decreased CL intensity and the concentration of H₂O₂. 5.0 × 10⁻⁴ mol L⁻¹ KMnO₄, 0.05 mol L⁻¹ H₂SO₄, and 1:20 Au NCs were employed. The error bars represent standard deviations of three replicate measurements.

Export to PPT

To evaluate the potential application of this method for real sample analysis, it was applied to the determination of H₂O₂ in water samples. Rainwater sample was collected from the city of Xi’an and drinking water samples were purchased from the local market. The water samples were filtered using a membrane filter (pore size: 0.45 μm) before analysis. The results are shown in Table 1. The recovery studies were carried out at the same time by adding H₂O₂ standard to water samples. As can be seen from Table 1, the recoveries of added H₂O₂ can be quantitative, demonstrating the method can be used to determine H₂O₂ in water samples.