DNA stabilized Ag–Au alloy nanoclusters and their application as sensing probes for mercury ions

Abstract

Metal nanoclusters (NCs) have attracted plenty of attention because of their unique properties and great application potentials. In this work, DNA scaffold Ag–Au alloy nanoclusters (Ag–Au ANCs) were fabricated by a one pot wet-chemical strategy and characterized by various techniques, including TEM, XPS and mass spectrometery (MS). The results indicate that owing to the strong interaction between DNA and Ag⁺, the silver NCs were formed first, then bundled with Au shells. In the Ag–Au ANCs, some of the Au is in an oxidized state as Au(I), which can largely modify the optical properties of the silver NCs. The Ag–Au ANCs demonstrate tunable emissions from green to red with highly improved stability. The fluorescence of Ag–Au ANCs was explored to detect Hg²⁺ in contrast to Ag NCs. The detection using Ag–Au ANCs demonstrated highly improved and excellent linearity and selectivity, which could effectively avoid the disturbance of Cu²⁺ and was promising for applications.

---

Introduction

Over the past 10 years, luminescent noble metal NCs have received enormous attention as highly fluorescent probes with utility for in vitro labelling, imaging, and sensing applications.^1–6 These NCs inherently possess discrete and size-dependent electronic states, and show unique molecule-like optical properties because the size is comparable to the Fermi wavelength of conduction electrons of the metal (∼1 nm).^4,7 Most noble metal NCs can be prepared in an attractive way by using polymers, bio-molecule dendrimers,⁸ proteins^9,10 and nucleic acids¹¹ as soft templates. Among them, DNA is a popular scaffold for Ag NCs because of the abundant base sequence and tunable chain length which can tune the emission wavelength of Ag NCs.^12–14 What's more important is the strong interaction between silver ions (Ag⁺) and DNA which can form the C–Ag⁺–C complex (C represents the cytosine).¹⁵ DNA stabilized Ag NCs, was first prepared and described by Dickson et al. more than 10 years ago and had been extensively studied for their unique structural, optical, and electronic properties.^12–14 Also, Ag NCs with tunable emissions from blue to near infrared regions were realized and applied for many biological applications.²

However, the poor stability is a common question of Ag NCs for the concern of bio-applications,^16,17 whatever kind of templates, resulting in possible ambiguity in the relationships between the detected substance and the change of itself. The neutral silver atom can be easily oxidized by oxygen no matter in colloid solution or solid powders, which seriously restricts the research and applications. Compared to Ag NCs, Au NCs shows better stability but suffers from lower quantum efficiency.¹⁸ For such kind of “homogeneous” NCs, considerable efforts have been devoted to exploration of the improved stability and quantum efficiency, but with little success.

Recently, bimetal nanoclusters or alloy nanoclusters (ANCs) attracted more and more attentions because of its cooperative properties in electronic, optical, and catalytic performance.^19–21 For example, Shi et al. found “silver effect” in gold catalysis, which greatly enhanced the catalytic activities of the bimetallic Ag–Au nanoparticles relative to those of homogenous Au nanoparticles.²² Also, highly luminescent Ag–Au ANCs were synthesized by Zhu et al. using metal exchange method in the thiolate system, with atomic precision.²³ Apart from the methods of synthesis mentioned above, other strategies were also employed for preparing Ag–Au ANCs. Wang's group realized the small enhancement in fluorescence of bovine serum albumin (BSA) stabilized bimetallic core–shell Au–Ag NCs because of “silver effect”.²⁴ Yong and co-workers prepared the Ag_xAu_25−x NCs using glutathione (GSH) templates and obtained the improved rigidity of the NCs which lead to strong fluorescence, because the central Ag atom could stabilize the charges on lowest unoccupied molecular orbital (LUMO).²⁵ Pal et al. reported the synthesis of highly fluorescent “giant” Au–Ag NCs protected by GSH ligands and further applied in Hg²⁺ sensing based on luminescent quenching mechanism.²⁶ Several studies were also carried out for the synthesis of NCs with protein and GSH molecule templates, and succeeded in Hg²⁺ sensing.^{3,5,6,27–29}

Despite numbers of studies on the electronic structures and luminescent properties of ligand stabilized Au–Ag alloy NCs, the photoemission mechanism is still not fully understood due to the complicated structure. In this work, we use a facile one-pot strategy to synthesize the water-soluble, stable and light emitting Au–Ag ANCs stabilized by DNA template. The objective is to gain insight into the composition structures and optical properties of Au–Ag ANCs protected by C rich DNA. To our knowledge, this is the first study about the structure and optical properties of Au–Ag ANCs in the DNA system. Besides the intrinsic optical feature, the Au–Ag ANCs has been used to develop high-performance sensors for Hg²⁺ detection based on fluorescence quenching mechanism.

Experimental

Chemicals

DNA (C₁₂, C₁₆, C₂₀, C₂₄, HPLC grade) were purchased from Invitrogen™|Life Technologies. Other metals salts such as HAuCl₄·3H₂O, AgNO₃, used in this work with analytical grade were obtained from variety of commercial sources. Both DNA template and metal salts were directly used without further purification.

Apparatus

Transmission electron microscopy (TEM) images were recorded on a FEI Tecnai G2 S-Twin microscope operated with an acceleration voltage of 200 kV, UV-vis spectra were recorded on a Shimadzu 3100 spectrometer. Fluorescence measurements were carried out using a fluorescence spectrophotometer (SENS-9000). Nanosecond fluorescence lifetime experiments were performed by the time correlated single-photon counting (TCSPC) system (HORIBA Scientific iHR 320). A 482 nm nano-LED light source (<200 ps) was used to excite the samples. Mass spectra were carried out on Bruker autoflex III smartbeam matrix-assisted laser desorption/ionization time of flight/time of flight mass spectrometer (MALDI-TOF/TOF-MS) (Germany), the laser system was smartbeam laser with 355 nm wavelength and the scan mode was reflective/linear, the ion mode was positive under an acceleration voltage of 20.00 kV, the ion extraction time was 0 ns, the solvent was chloroform and the matrix was dihydroxy-benzoic acid. XPS measurements were recorded on an XPS system (Shimadzu AXIS-165) using a monochromatic Al Kα X-ray source (1486.6 eV). Binding energies (BEs) were referenced to the C 1s BEs at 284.8 eV.

Synthesis of DNA Ag NCs and Au–Ag ANCs

The DNA Ag NCs and Au–Ag ANCs were synthesized by similar way according to previous literature.^14,19 Briefly, for the preparation of Au–Ag ANCs, 13.6 nmole DNA was dissolved in 630 μL phosphate buffer solution with phosphate group of 10 mM and pH value of 7.0, after that 68 μL AgNO₃ (2 mM) solution was added into the above solution, keeping at 4 °C for 20 minutes. Then 135 μL HAuCl₄ (2 mM) was added to above mixture, after two minutes, 68 μL fresh coolly prepared NaBH₄ (2 mM) was added to reduce the mixture with rigorously shaking and then kept for 24 hours in dark to obtain reproducible results. The final concentration proportion (molar) of the reagent was DNA : Ag⁺ : Au³⁺ : BH⁻ = 1 : 10 : 20 : 10. The Ag NCs was prepared by using the same starting materials with the same proportions and conditions as Au–Ag ANCs, except addition of Au salts. The prepared rude products were dialyzed (cut-off 1000 Da) with water for 24 hours, then were stored at 4 °C.

Fluorescence detection of Hg²⁺ ions

For the detection of Hg²⁺ ions, 20 μL prepared different concentration of Hg²⁺ ions was added to 380 μL as synthesized Ag NCs and Au–Ag ANCs and reacted 3 minutes to ensure the complete reaction and then was recorded by the associated fluorescence spectra accordingly at room temperature. The selectivity against other metal ions was examined under the identical conditions. All the measurements was recorded once for the Ag NCs and three times for Ag–Au ANCs, respectively.

Results and discussion

Characterization of DNA Ag NCs and DNA Au–Ag ANCs

The representative TEM images of Ag NCs and Ag–Au ANCs was shown in Fig. 1a and b, respectively. Fig. 1c showed the size distribution of the Ag NCs and the mean size was estimated to be 3.0 ± 0.9 nm. The Ag–Au ANCs showed narrower size distribution than the Ag NCs and the average size was 1.9 ± 0.5 nm, as shown in Fig. 1d. The larger mean size and size distribution of Ag NCs can be attributed to the aggregation or oxidization in the preparation, after doping the gold atom, this phenomenon was improved in Ag–Au ANCs.

Fig. 1 Representative TEM graph of prepared aqueous DNA Ag NCs (a) and DNA Ag–Au NCs (b), and the corresponding size distribution is shown in the (c) and (d), respectively.

The absorption spectra of the aqueous DNA Ag NCs (Ag NCs) and Au–Ag ANCs were characterized at room temperature after the reduction of 3 hours. As shown in Fig. 2, there exist three absorption peaks in Ag NCs (corresponding to the line marked 1 : 0), locating around 395, 440 and 535 nm, respectively. With the addition of gold, the peaks at 395 nm and 440 nm gradually decrease and are broadened. When the proportion of Ag : Au reaches to 1 : 1.5, a combined band peaking at 425 nm appears. The peak around 535 nm gradually vanishes and a new wider band around 525 nm gradually occurs with the increase of Au component, which was caused by the strong LSPR absorption of some formed larger gold nanoparticles.^7,30 This can be confirmed by the TEM images of the products prepared when the metal salts proportion (Ag : Au) is 1 : 0, as shown in Fig. S1,† the results indicated that larger aggregated nanoparticles were formed without adding silver salts.

Fig. 2 The absorption spectra of the Ag–Au ANCs with different proportion between silver and gold components.

Actually, the optical properties of the DNA-modified Ag NCs depend strongly on the chain length of DNA and due to the species of Ag NCs are various, it is difficult to distinguish the detail origins of these absorption peaks. However, the Ag–Au ANCs prepared with different chain lengths of DNA demonstrate similar optical properties (see Fig. S2†). The results suggests that after the addition of gold, the similar alloy structure might be formed. Further work should be performed to identify the accurate structure of the Ag–Au ANCs. Moreover, all the absorption spectra of prepared Ag NCs with different DNA showed same changes after 5 days in ambitious environments (Fig. S3†). The freshly prepared Ag NCs showed several peaks, after 5 days, all the NCs showed only one main peaks locating at 395 nm, which may come from the oxidized Ag NCs.

For confirming these DNA Ag NCs and Ag–Au ANCs species, the matrix-assisted laser desorption/ionization time-of-flight/time-of-flight mass spectrometry (MALDI-TOF/TOF-MS) was performed to identify the composition and number of silver atom of Ag NCs, as shown in Fig. 3. The results indicated that for the Ag NCs, there were several Ag_n NCs species with n = 3–13 and the main species of Ag NCs was Ag₁₁ NCs. The MALDI-TOF/TOF-MS was also performed to identify the Ag–Au ANCs species, it is a pity that no signal was got although several matrix was used in the measurements.

Fig. 3 The MALDI-TOF-MS results of the Ag NCs, the m/z value shows the molecular weight of the different Ag NCs species, the molecular weight of used DNA template is 5722. The number of silver atom can be calculated by n = (m/z − 5722)/108.

To validate the construction and valence state of elements in the Au–Ag ANCs, the XPS measurements were performed. From Fig. 4a, the binding energy of Ag 3d_5/2 peak slightly shifted to lower side relative to the pure Ag NCs as gold content increased from 1 : 0 to 1 : 0.5, indicating some Ag atoms were positively charged.^21,29 When the proportion of gold was further increased from 0.5 to 2, the peaks of Ag 3d_5/2 showed no obvious shift. The Au 4f_7/2 peaks (see Fig. 4b) were observed to shift to high energy side relative to pure Au NCs or bulk Au (83.8 eV) but far lower than the binding energy of Au³⁺ 4f_7/2 (87.3 eV), implying the existence of Au(I) and Au(0). In addition, the peaks value of the Au 4f shifted to high energy gradually with the increase of gold content, showing that more Au(I) was formed in the Ag–Au ANCs.

Fig. 4 The XPS spectra of Au–Ag ANCs and Ag NCs. (a) The binding energy Ag 3d orbits and (b) Au 4f orbits with different molar proportion between Ag and Au.

Photoluminescence properties of DNA Ag NCs and DNA Au–Ag ANCs

In order to better understand the modified luminescence properties of DNA Au–Ag ANCs, the luminescent properties of DNA Ag NCs were first studied. As shown in Fig. 5, the Ag NCs demonstrate two main emission bands. An emission band centering at 535 nm can be identified under the excitation of 467 nm, corresponding to the absorption band of Ag NCs around 440 nm. The gap between the absorption and excitation spectra can be attributed to internal conversion.³¹ Another emission band locates at 640 nm, under the excitation of 570 nm, corresponding to the absorption band around 540 nm. Dickson attributed these two emission bands to oxidized and un-oxidized Ag NCs, respectively.¹² According to the Fygenson³² and co-workers results about the poly-cytosine loops stabilize Ag NCs, the red emission linked with the Ag₁₃ NCs and green correlated to Ag₁₁ NCs. It should be noted that the green emission gradually increases with the increasing reaction time, while the red emission decreases, which is consistent with the literature of Dickson (see Fig. S4†).

Fig. 5 The two emission band of Ag NCs and the corresponding excitation spectra.

Fig. 6a and b show the emission spectra and the corresponding excitation spectra of Ag–Au ANCs with different proportions of silver and gold under 467 nm excitation, respectively. From Fig. 6b it can be seen that with the increase of Au content, the emission peak at 535 nm gradually decreases, a new broad band peaking at 650 nm appears, and it gradually increases with the increase of Au content under the excitation of 467 nm. The spectral range covers from 490 nm to 800 nm. The emission color evolves from green, yellow to red under the excitation of 467 nm when the gold content increases, as the digital pictures of the corresponding solutions shown in the inset of Fig. 6b. It should be noted that the origin of the 650 nm band in the Ag–Au ANCs quite differ from the red emission band in the Ag NCs. The reasons are as follows: (1) the peak of the red emission in the Ag–Au ANCs is at 650 nm, while that in the Ag NCs locates at 640 nm (see Fig. S5 in detail†). (2) The band width at half height of the red emission in the Ag–Au ANCs is much broader than that in the Ag NCs. (3) The lifetime constant of the red emission in the Ag–Au ANCs is three-orders slower than that in Ag NCs, which will be discussed in detail in the following text. Moreover, the photoemission of Ag–Au ANCs under excitation of 570 nm was also measured when the proportion of Ag and Au was changed (see Fig. S6†), the prepared Ag NCs showed bright red emission, when the content of gold increased to 1 : 0.5 and 1 : 1, the emission at 640 nm was completely quenched, and new weak red emission occurred as the content increased to 1 : 1.5 and 1 : 2. This also indicated that the red emission of Ag–Au ANCs originated from new structures. In order to determine the structure and formation process of Ag–Au ANCs prepared by one pot, the Ag–Au ANCs were also prepared by the two-step method, in which the Ag NCs were first prepared, then different amounts of Au³⁺ was added to the Ag NCs. The corresponding emission spectra of the Ag–Au ANCs by the two-step method were recorded, as shown in Fig. 7a. The results show that the spectra shape and evolution trend of the samples papered by the two-step method are very similar to those prepared by one step method using different Ag and Au proportion. This suggests that in the one pot preparation, the Ag⁺ was reduced firstly to form the Ag NCs, then Au³⁺ quickly reacted with Ag NCs, forming the Ag–Au ANCs (pure DNA Au NCs shows no fluorescence emission, see Fig. S1†). This can be attributed to the combination¹⁵ of DNA with Ag⁺ is far stronger than that with Au³⁺, which makes the formation of Ag NCs preferentially. According to XPS results above, the existence of gold in the Ag–Au ANCs should be Au(I) and Au(0).

Fig. 6 (a) Photo excitation spectra of the Ag–Au ANCs with different proportion of Ag and Au. Two main monitoring emission wavelength (535 nm and 650 nm) was used when the proportion of Ag and Au equaled to 1 : 0.5 and 1 : 1. (b) The photoemission spectra with different proportion between silver and gold salts. The inset shows the digital photo of Au–Ag ANCs when the proportion of Ag : Au is 1 : 0, 1 : 1, and 1 : 2, respectively.

Fig. 7 (a) The photoluminescence change of the Ag NCs after adding different Au³⁺ ions. (b) Adding Au³⁺ to the Ag–Au ANCs, (c) the fluorescence emission spectra of prepared Ag–Au ANCs with adding different BH⁻ and the proportion of Ag : Au was fixed to 1 : 2.

There are two possible structure composition of the Ag–Au ANCs according to previous analysis results, as shown in Scheme 1: A represents the “mixture structure” and B represents “core–shell structure”. For testing the two models, plenty of Au³⁺ ions were added to the prepared Au–Ag ANCs, then the emission spectra were recorded (Fig. 7b). No remarkable change was observed in the spectra, which supported the model B. Because if some Ag(0) atoms were exposed to outside like the model A, they could be oxidized by the Au³⁺ in the model A (mixture structures), resulting in the spectral change of the Au–Ag ANCs.

Scheme 1 The possible structure of the Ag–Au ANCs, for abbreviation, the DNA template is not showed. The golden balls represent the Au(I) and Au(0) and the sky blue balls represent the Ag(0) and Ag(I).

In order to confirm this point further, the proportion of metal precursor and reduction agent NaBH₄ was adjusted to detect the spectral change. Because the initial molar proportion of Ag⁺ : Au³⁺ : BH⁻ we used was 1 : 2 : 1 in the synthesis of the Au–Ag ANCs, which meant that the reducing agent NaBH₄ in the reaction was inadequate and only part of the metal ions was reduced. So we fixed the molar proportion of Ag⁺ : Au³⁺ at 1 : 2, and different amount of BH⁻ was added. The results indicated that when the BH⁻ was less than or equal to the initial proportion, the corresponding Ag–Au ANCs showed red emission under 467 nm excitation. When BH⁻ was higher than the starting proportion, the red emission decreased, leading to the relative increase of the green emission which was in accordance with the green emission of Ag NCs (see Fig. 7c). When the amount of BH⁻ reaches four proportion, theoretically, all the metal ions including Au³⁺ and Ag⁺ can be reduced to the neutral states. This indicates that the red emission of Ag–Au ANCs was probably related to the positive states Ag(I) or Au(I).

In order to validate the points, we performed experiments about the influence of free Ag⁺ on photoluminescence of Ag NCs, because the luminescence of Ag NCs can be easily influenced by free Ag⁺ which can form ligand metal charge transfer states (LMCTs) as previous reported.³³ Fig. 8a shows the emission spectra of the Ag NCs when different amount of free Ag⁺ was added. It can be seen that the green emission of pure Ag NCs gradually red shifts to 605 nm, which can be attributed to the formation of new LMCTs. In contrast, the free Ag⁺ was also added into Ag–Au ANCs prepared by one pot, and the spectra displayed rarely change (see Fig. 8b). This indicates that the surface of the prepared Ag–Au ANCs is without Ag(I) or Ag(0). As model B described, the Au(I) shell around Ag NCs core can prevent the Ag⁺ combining with the Ag NCs. Therefore, it can be confirmed that the red emission band in the Ag–Au ANCs may originate from Au(I) shell. The Au³⁺ in the model A (mixture structures), resulting in the spectral change of the Au–Ag ANCs.

Fig. 8 PL change of Ag NCs (a) and Au–Ag ANCs (b) after adding different amount of free Ag⁺, all the emission spectra was recorded under the excitation wavelength at 467 nm.

As previous reports, Au(I)-thiolate complex shell also can form the LMCTs in the core–shell structure Au NCs which prepared by the GSH template.^34,35 such LMCTs generally was demonstrated long decay time constants ranging of several microseconds.^33,36,37 We guess that Au(I) may also form the LMCTs in the prepared Ag–Au ANCs. In order to prove this point, the fluorescence dynamics of Ag NCs and Ag–Au ANCs were investigated, as shown in Fig. 9. The decay dynamics of Ag NCs demonstrate a bi-exponential function, for both the green and red emissions, which can be expressed as I = I₁ exp(−t/τ₁) + I₂ exp(−t/τ₂), where I₁ and I₂ represent the faster and slower decay time constants, respectively, τ₁ and τ₂ are corresponding constants. The average decay constants for the green (at 535 nm) and red (at 640 nm) emissions of Ag NCs were determined to be 3.73 ns and 2.43 ns (Fig. 9a), respectively, which was similar to previous reports.¹⁴ Interestingly, the dynamics of the red emission of Au–Ag ANCs (at 650 nm) demonstrated a four exponential decay, which can be written as I = I₁ exp(−t/τ₁) + I₂ exp(−t/τ₂) + I₃ exp(−t/τ₃) + I₄ exp(−t/τ₄), where τ₁, τ₂, τ₃ and τ₄ represent the different decay time constants, respectively. By fitting, τ₁, τ₂, τ₃ and τ₄ was 4.04 ns (7.26%), 69.03 ns (1.61%), 522.2 ns (9.95%) and 3.38 μs (81.17%) (Fig. 9b), respectively. This indicated that the microsecond predominance of lifetime components and the average decay time was 2.79 μs, such long lifetimes for Au–Ag ANCs at 650 nm are quite different from the emissions of pure Ag NCs, which are usually observed in emissions of LMCTs, such radiative relaxation was most likely via a metal-centered triplet state.³⁴ It is suggested that in the Ag–Au ANCs Au(I) formed the LMCTs in the gap of the HOMO and the LUMO offered by Ag NCs. As shown in the Scheme 2, after pumped from the HOMO (S₁) to LUMO (S₀), the electrons can relax to the LMCTs states.

Fig. 9 Photoluminescence decay of Ag NCs at 535 nm and 640 nm (a), and the prepared Ag–Au ANCs at 650 nm (b), respectively.

Scheme 2 Illustration of the photoluminescence mechanism of (a) green emission of Ag NCs and (b) the emission of new LMCTs of Ag–Au ANCs.

For investigating universality of alloy structures, DNA templates with different chain length from 12 to 24 bases was employed. Quite different absorption and emission behavior can be observed in Fig. 10a, due to different sizes of Ag NCs species caused by template. The same experiments were carried on the Au–Ag ANCs with gold adoption (base : Ag : Au = 2 : 1 : 2 for all the templates), while both absorption and the emission of Au–Ag ANCs centered at 650 nm shows similar position as DNA template varies (see Fig. 10b). This result indicated that the core shell Ag–Au ANCs can formed in different templates and showed good universality.

Fig. 10 (a) Emission spectra of pure Ag NCs and Au–Ag ANCs (b) with different chain length of DNA template under excitation at 467 nm.

Detection of Hg²⁺

The homogenous Ag NCs was explored to detect the Hg²⁺ in the past several years, but further application was restricted by its bad stability.^3-6,27–29 Moreover, the cooper ions (Cu²⁺) showed serious disturbance in many previous works, because of the anti-galvanic reduction (AGR),^21,24,38,39 namely, both Hg²⁺ and Cu²⁺ ions can oxidize Ag(0) because of the enhanced metal activity in the Ag NCs. Also, some previous works pointed out the d¹⁰–d¹⁰ interaction (also called metalophilic interaction) between Hg²⁺ and Ag NCs would induce the charge transfer and cause fluorescence quenching.^24,29,38 In this work, we investigated the sensing properties of the Au–Ag ANCs as the fluorescence probe for detecting the Hg²⁺. As a control, the Ag NCs was also performed to detect the Hg²⁺ in the same conditions.

Firstly, the fluorescent stability of Ag–Au ANCs and Ag NCs in the ambitious environments was investigated, as shown in Fig. 11a. It can be seen that the red emission of Ag NCs decreased rapidly and almost quenched completely within three days. On the contrary, the green emission of Ag NCs showed gradually enhanced fluorescence intensity, and nearly reserved as a constant after that. As discussed above, the red emission originated from the unstable Ag NCs species, which could be easily oxidized by the oxygen in the air, while the green emission originated from the stable Ag₁₁ NCs species. Differing from Ag NCs, the fluorescence of Ag–Au ANCs kept good stability all the time because of protection of Au(I) shell. Fig. 11b shows the luminescent intensity change of Ag NCs and Ag–Au ANCs with ten times measurements. The emission intensity of Ag NCs at 640 nm decreased quickly because of the accelerated oxidization process under continuous radiation. The green emission showed increase initially and approached at a maximum at 5 times, then gradually decreased. The reason is that the initial radiation might be helpful of forming green Ag NCs species, but further irradiation also could bleach the Ag NCs. The prepared Ag–Au ANCs showed a slight decrease about 5% after irradiation of ten times, indicating better photo-stability.

Fig. 11 (a) The luminescence stability of Ag–Au ANCs and Ag NCs in different days storing at the ambitious environments. (b) The normalized luminescence intensity change with continuous measurements, all the intensity measured at the first time was set to one.

Then, we performed the detection Hg²⁺ using the prepared Ag NCs and Ag–Au ANCs. Fig. 12a–c show the dependence of the luminescence intensities of Ag NCs green emission, red emission and Ag–Au ANCs red emission on the concentration of Hg²⁺ ions, respectively. All the luminescence intensity quenches linearly with the concentration of Hg²⁺ ions and can be fitted by the Stern–Volmer equation:⁴⁰ I₀/I = C + K_SV[Q], where I₀ and I represent the fluorescence intensity of the Au–Ag ANCs with blank and reacts with different ions concentration, respectively. K_SV is the quenching coefficient and [Q] is the Hg²⁺ concentration, C is a constant. Both the green and red emission intensity change of Ag NCs was measured only once because of the poor stability of itself. For evaluating the measurements error, the freshly prepared samples was divided into ten samples and the standard error of initial measurements of these samples was used as the measurements error. As for Ag–Au ANCs, all the measurements results was recorded three times, and the average value and standard error was used as the response value for the Hg²⁺ concentration and measurements error, respectively. The emission at 535 nm of Ag NCs showed bad response both in the larger scale (0–10 μM) and in smaller regime (0–0.9 μM), as shown in Fig. 12a. The red emission of Ag NCs showed good linear relationship from 0 to 8000 nM, but from 0 to 800 nM, the response showed great measurements error, see Fig. 12b. The Ag–Au ANCs showed excellent linear response for Hg²⁺ concentration, no matter in the larger scale (0–10 μM) or smaller range (0–1 μM), as shown in Fig. 12c. The limit of the detection (LOD) was estimated by the 3σ/s, where σ was the standard deviation of ten times measurements about the blank Ag NCs or Ag–Au ANCs solutions, and s represented the sensitivity. The LOD for Ag NCs at green and red emission was estimated with 35.1 nM and 29.2 nM, respectively. While for the Ag–Au ANCs, the LOD was as lower as 1.64 nM.

Fig. 12 Left column: Stern–Volmer plots representing the quenching effect of Hg²⁺ on the fluorescence emission of the Ag NCs at 535 nm (a) and 640 nm (b) and emission of the Ag–Au ANCs at 650 nm (c), respectively. All the inset: the quenching effect of Hg²⁺ at concentrations from 1 nM to 900 nM. Right column: the effect of different metal ions on the fluorescence emission of the Ag nanoclusters at 535 nm (d) and 640 nm (e) and the Ag–Au ANCs at 650 nm (f), respectively.

Fig. 12d and e showed the relative change of green and red emission intensity of Ag NCs after adding metal ions (the last concentration in the mixture solution for Hg²⁺, Cu²⁺ and other metal ions was 10 μM, 10 μM and 100 μM, respectively) to the divided samples. For green emission, most of the metal ions showed the slight difference from the blank intensity, the Cu²⁺ enhanced the luminescence largely but Hg²⁺ quenched the luminescence greatly in comparison to other metal ions. When the red emission was used as the detection fluorescence probe, both the Cu²⁺ and Hg²⁺ quenched the red emission of Ag NCs seriously. The different response of Cu²⁺ to green and red emission of Ag NCs can be explained by the AGR effects.⁴¹

Because of the enhanced activity for Ag NCs, the Cu²⁺ can oxidize the Ag NCs lead to the green emission increase and red emission decrease. So the new metal activity can be ranked as: Ag NCs > Cu > Hg > Ag.²³ According to such results, the AGR can also happen between Ag NCs and Hg²⁺, so the green emission should be enhanced but the truth was the emission was quenched. The results can be attributed to the strong metalophilic interaction between Hg²⁺ and Ag NCs, which become dominant in the detection of Hg²⁺ using the Ag NCs, inducing charge transfer and luminescence quenching. For the Ag–Au ANCs, the selectivity for Hg²⁺ was evaluated by a control experiments. The results was recorded after adding the metal ions to the Ag–Au ANCs solutions, the last concentration of Hg²⁺ and other metal ions in the detected solution was 10 μM and 100 μM, respectively. As shown in Fig. 12f, only Hg²⁺ quenched the luminescence of Ag–Au ANCs heavily and other metal ions showed slight effect in comparison to the blank solution. Compared to the Ag NCs red emission probe, Cu²⁺ showed negligible influence to the detection results. As discussed before, the AGR may be the main reason in the reaction of Cu²⁺ and Ag NCs, but in the Ag–Au ANCs, the red emission at 650 nm originated from the Au(I) LMCTs, and the Cu²⁺ cannot oxidize the Au(I) to Au³⁺, so the Cu²⁺ showed slightly influence to the luminescence of Ag–Au ANCs.

The interaction Hg²⁺ and the Ag–Au ANCs can also be explained by the strong metalophilic interaction because both the Au(I) and Hg²⁺ have the same 5d¹⁰ electron configuration.⁴² Further verification was performed by the XPS measurements, as shown in Fig. 13a, there are four obvious peaks in the mixture of Hg²⁺ and Ag NCs around the binding energy of Hg 4f orbits which comes from the Hg²⁺ and Hg⁰, respectively. In the Ag–Au ANCs, it can also be found one obvious peaks (it is hard to fit and distinguish the valence of Hg for only one peaks) which indicate that Hg²⁺ can combined with the Ag NCs and Ag–Au ANCs by the metalophilic interaction, and obvious results of AGR process can be found when the Hg²⁺ reacted with the Ag NCs. However, as shown in Fig. 13b, there was no peaks around the binding energy of Cu 2p orbits. This implied that the interaction between Cu²⁺ or Cu(0) and the Ag NCs or Ag–Au ANCs is very weak, and the interaction of Cu²⁺ and Ag NCs and Ag–Au ANCs should not be the metalophilic interaction but may be the AGR process. For the Ag–Au ANCs, because of the protection of Au(I), such process was prevented which lead to great selectivity in the detection of Hg²⁺ by using Ag–Au ANCs.

Fig. 13 The binding energy of Hg 4f (a) and Cu 2p (b) in the Ag NCs and Ag–Au ANCs, respectively.

Conclusions

In summary, DNA stabilized Ag–Au alloy ANCs were prepared by a one-pot wet-chemical method, the structure and emission mechanism of them were systemically investigated. The Au–Ag ANCs formation process can be decomposed to two steps: the formation of Ag NCs at the first step, the package and oxidization of Ag NCs by gold at the second step. At this step some Au(III) was reduced to Au(I). Such core–shell Au–Ag ANCs show highly modified optical properties relative to Ag NCs. The Au(I) can form LMCTs in the gap of LUMO and HOMO offered by Ag NCs, which can tune the transition wavelength of Ag NCs from green to red and demonstrate long decay time constant ranging of microsecond, large Stokes shifts and broad line width. On the other hand, the Au shell may also stabilize the structure of Ag NCs core to form stable Ag–Au ANCs. Lastly, the Au–Ag ANCs was used for the Hg²⁺ detection based on the fluorescence quenching induced by metalophilic interaction. This kind of Au–Ag ANCs sensing prober showed excellent linearity, good stability and high selectivity, exhibiting great potential in heavy metal sensing application.

Acknowledgements

This work was supported by the Major State Basic Research Development Program of China (973 Program) (no. 2014CB643506), the National Natural Science Foundation of China (Grant no. 51202019, 21403084), the Program for Chang Jiang Scholars and Innovative Research Team in University (no. IRT13018).

Notes and references

1. J. Zheng, P. R. Nicovich and R. M. Dickson, Annu. Rev. Phys. Chem., 2007, 58, 409–431.
2. S. Choi, R. M. Dickson and J. Yu, Chem. Soc. Rev., 2012, 41, 1867–1891.
3. L. Deng, Z. Zhou, J. Li, T. Li and S. Dong, Chem. Commun., 2011, 47, 11065–11067.
4. I. Diez and R. H. A. Ras, Nanoscale, 2011, 3, 1963–1970.
5. C. Guo and J. Irudayaraj, Anal. Chem., 2011, 83, 2883–2889.
6. X. Yuan, T. J. Yeow, Q. Zhang, J. Y. Lee and J. Xie, Nanoscale, 2012, 4, 1968–1971.
7. S. K. Ghosh and T. Pal, Chem. Rev., 2007, 107, 4797–4862.
8. J. Zheng, C. W. Zhang and R. M. Dickson, Phys. Rev. Lett., 2004, 93, 077402.
9. J. Xie, Y. Zheng and J. Y. Ying, J. Am. Chem. Soc., 2009, 131, 888.
10. T. Zhou, Y. Huang, W. Li, Z. Cai, F. Luo, C. J. Yang and X. Chen, Nanoscale, 2012, 4, 5312–5315.
11. J. T. Petty, J. Zheng, N. V. Hud and R. M. Dickson, J. Am. Chem. Soc., 2004, 126, 5207–5212.
12. C. M. Ritchie, K. R. Johnsen, J. R. Kiser, Y. Antoku, R. M. Dickson and J. T. Petty, J. Phys. Chem. C, 2007, 111, 175–181.
13. C. I. Richards, S. Choi, J.-C. Hsiang, Y. Antoku, T. Vosch, A. Bongiorno, Y.-L. Tzeng and R. M. Dickson, J. Am. Chem. Soc., 2008, 130, 5038.
14. S. Choi, J. Yu, S. A. Patel, Y.-L. Tzeng and R. M. Dickson, Photochem. Photobiol. Sci., 2011, 10, 109–115.
15. V. Soto-Verdugo, H. Metiu and E. Gwinn, J. Chem. Phys., 2010, 132, 195102.
16. J. Sharma, R. C. Rocha, M. L. Phipps, H.-C. Yeh, K. A. Balatsky, D. M. Vu, A. P. Shreve, J. H. Werner and J. S. Martinez, Nanoscale, 2012, 4, 4107–4110.
17. Z. Huang, J. Ren, W. Yang and X. Qu, Chem. Commun., 2013, 49, 10856–10858.
18. L. Yan, Y. Cai, B. Zheng, H. Yuan, Y. Guo, D. Xiao and M. M. F. Choi, J. Mater. Chem., 2012, 22, 1000–1005.
19. W.-Y. Chen, G.-Y. Lan and H.-T. Chang, Anal. Chem., 2011, 83, 9450–9455.
20. H. Barron, L. Fernandez-Seivane, H. C. Weissker and X. Lopez-Lozano, J. Phys. Chem. C, 2013, 117, 21450–21459.
21. J. Sun, H. Wu and Y. Jin, Nanoscale, 2014, 6, 5449–5457.
22. Y. Yong, C. Li, X. Li, T. Li, H. Cui and S. Lv, J. Phys. Chem. C, 2015, 119, 7534–7540.
23. S. Wang, Y. Song, S. Jin, X. Liu, J. Zhang, Y. Pei, X. Meng, M. Chen, P. Li and M. Zhu, J. Am. Chem. Soc., 2015, 137, 4018–4021.
24. N. Zhang, Y. Si, Z. Sun, L. Chen, R. Li, Y. Qiao and H. Wang, Anal. Chem., 2014, 86, 11714–11721.
25. R. Kobayashi, Y. Nonoguchi, A. Sasaki and H. Yao, J. Phys. Chem. C, 2014, 118, 15506–15515.
26. M. Ganguly, C. Mondal, J. Pal, A. Pal, Y. Negishi and T. Pal, Dalton Trans., 2014, 43, 11557–11565.
27. J. Liu, X. Ren, X. Meng, Z. Fang and F. Tang, Nanoscale, 2013, 5, 10022–10028.
28. J. L. MacLean, K. Morishita and J. Liu, Biosens. Bioelectron., 2013, 48, 82–86.
29. J. Zhang, Y. Yuan, Y. Wang, F. Sun, G. Liang, Z. Jiang and S.-H. Yu, Nano Res., 2015, 8, 2329–2339.
30. J. M. Luther, P. K. Jain, T. Ewers and A. P. Alivisatos, Nat. Mater., 2011, 10, 361–366.
31. M. Pelton, Y. Tang, O. M. Bakr and F. Stellacci, J. Am. Chem. Soc., 2012, 134, 11856–11859.
32. P. R. O'Neill, L. R. Velazquez, D. G. Dunn, E. G. Gwinn and D. K. Fygenson, J. Phys. Chem. C, 2009, 113, 4229–4233.
33. Y. Chen, T. Yang, H. Pan, Y. Yuan, L. Chen, M. Liu, K. Zhang, S. Zhang, P. Wu and J. Xu, J. Am. Chem. Soc., 2014, 136, 1686–1689.
34. Z. Luo, X. Yuan, Y. Yu, Q. Zhang, D. T. Leong, J. Y. Lee and J. Xie, J. Am. Chem. Soc., 2012, 134, 16662–16670.
35. K. Pyo, V. D. Thanthirige, K. Kwak, P. Pandurangan, G. Ramakrishna and D. Lee, J. Am. Chem. Soc., 2015, 137, 8244–8250.
36. S. H. Yau, N. Abeyasinghe, M. Orr, L. Upton, O. Varnavski, J. H. Werner, H.-C. Yeh, J. Sharma, A. P. Shreve, J. S. Martinez and T. Goodson III, Nanoscale, 2012, 4, 4247–4254.
37. Z. Wu and R. Jin, Nano Lett., 2010, 10, 2568–2573.
38. D.-Q. Feng, G. Liu and W. Wang, J. Mater. Chem. B, 2015, 3, 2083–2088.
39. N. Xia, J. Yang and Z. Wu, Nanoscale, 2015, 7, 10013–10020.
40. L. Shang and S. Dong, J. Mater. Chem., 2008, 18, 4636–4640.
41. Z. Wu, Angew. Chem., Int. Ed., 2012, 51, 2934–2938.
42. J. Xie, Y. Zheng and J. Y. Ying, Chem. Commun., 2010, 46, 961–963.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra07563k

---