Introduction

In recent years there has been growing interest among chemists in the preparation of new materials that are ordered on length scales that extend beyond the molecule (Mann et al. 2000). Particularly, one-dimensional (1D) assemblies of nanoparticles (NPs), i.e. chains, can significantly help in the understanding of some processes in living organisms (Walker et al. 2002; Alivisatos 2004) and fundamental quantum mechanics of nanometre-scale systems (Maier et al. 2001, 2003). A directional organization process is critical for the preparation of 1D NP assemblies. The use of linear templates (organic polyelectrolyte (Minko et al. 2002; Kiriy et al. 2002; Yu and Cölfen 2004; Cölfen and Qi 2001; Yu et al. 2002, 2003; Zhang et al. 2001) and biomolecules (Warner and Hutchinson 2003; Mann et al. 2000; Lui and Lu 2003; Si et al. 2007a, b; Walsh et al. 2003; Lee et al. 2002), inorganic wires (Ajayan and Lijima 1993; Yin et al. 2004; Han and Zettl 2003; Wang et al. 2004) and nanotubes (Ugarte et al. 1998), pores (Martin 1996) and step-edges (Fort et al. 2003)) is a vital and feasible approach for the preparation of 1D assemblies of NPs (Tang and Kotov 2005).

On the other hand, among the various MNPs, gold (Au) and silver (Ag) NPs can be easily used as optical probes, since they present an intense surface Plasmon resonance band (SPR) in the visible region (ca. 520 nm for Au and ca. 400 nm for Ag). However, the localization of the SPR band depends of some parameters, as size, shape and refractive index of the environment. Depending of the conditions the NPs could present different colours. The NPs colour could also change upon the coupling of SPR of adjacent Nps (Chechik and Crooks 1999; Lowe et al. 2002; Si et al. 2007a, b). Furthermore, AuNPs and AgNPs are known to dramatically change the optical properties of nearby fluorophores (Hong and Kang 2006; Aslan et al. 2007; Cheng and Xu 2007). Studies involving AuNPs and AgNPs coated with fluorophores on surface, such as fluorescein, often show an emission fluorescence quenching (Dulkeith et al. 2005; Schneider and Decher 2006) or enhancement (Aslan et al. 2004, 2006; Lakowicz et al. 2002; Zhang and Lakowicz 2006; Xu et al. 2011), which highly depended on fluorophore–metal distance (Cheng and Xu 2007; De and Kundu, 2011; Saha et al. 2009).

Having in mind this aspect, and following our research project in new fluorescent and/or colorimetric chemosensors (Lodeiro, et al. 2010; Lodeiro and Pina 2009), herein we describe the design of colorimetric probes based on fluorescein-directed assembly of gold and silver NPs, and their use for sensitive and selective detection and quantification of mercury metal ion by the formation of 1D chains without templates. For this study, a novel compound L with formula C26H24O5S2 was synthesized and fully characterized by reaction of fluorescein (free acid) as precursor with 3-chloro-1-propanethiol, yielding 95 % of an orange powder (L) (Scheme 1).

Scheme 1
scheme 1

Schematic synthesis of chemosensor L

Full size image

Experimental details

Physical measurements

Elemental analyses were carried out with Fisons Instruments EA1108 microanalyzer at the University of Vigo (CACTI), Spain. Infrared spectra were recorded in KBr windows using a JASCO FT/IR-410 spectrophotometer. 1H and 13C NMR were carried out in a BrukerAvance III400 at an operating frequency of 400 MHz for 1H NMR and 100.6 MHz for 13C NMR using the solvent peak as an internal reference at 25 ºC.

The MALDI-MS analyses have been performed with a MALDI-TOF-TOF-MS model Ultraflex II (Bruker, Germany) equipped with nitrogen from the BIOSCOPE group. Each spectrum represents accumulations of 5 × 50 laser shots. The reflection mode was used. The ion source and flight tube pressure were less than 1.80 × 10−7 and 5.60 × 10−8 Torr, respectively. The MALDI mass spectra of the soluble samples (1 or 2 μg/μL) were recorded using the conventional sample preparation method for MALDI-MS. One microlitre was put on the sample holder on which the ligand had been previously spotted. The sample holder was inserted in the ion source.

The nanoparticles size distributions were measured using a dynamic light scattering system, Malvern Nano ZS instrument with a 533-nm laser diode from the Faculty of Science at Ourense Campus, University of Vigo. To perform the transmission electron microscopy (TEM) images the samples were prepared dropping 1 μL of the colloidal suspension onto a copper grid coated with a continuous carbon film and allowing the solvent to evaporate. TEM images were obtained through a JEOL JEM 1010F TEM operating at 100 kV. To perform the Fourier transformations we used the Digital Micrograph (Gatan) software (Sánchez-Iglesias et al. 2006).

UV–Vis absorption spectra (220–900 nm) were performed using a JASCO-650 UV–Vis spectrophotometer and fluorescence spectra on a HORIBA JOVIN-IBON Spectramax 4. The linearity of the fluorescence emission vs. concentration was checked in the concentration range used (10−5–10−6 M). A correction for the absorbed light was performed when necessary. The spectrophotometric characterizations and titrations were performed as follows: the stock solutions of the compounds (ca. 10−3 M) were prepared by dissolving an appropriate amount of the compounds in a 10-mL volumetric flask and diluting them to the mark with absolute ethanol. The solutions were prepared by appropriate dilution of the stock solutions, which were still 10−5–10−6 M. All of the measurements were performed at 298 K.

Luminescence quantum yield was measured using a solution of fluorescein in sodium hydroxide solution (0.1 M) as a standard ([Φ] = 0.79) (Turkevich et al. 1951; Enüstün and Turkevich 1963; Turkevich 1985a, b) for L. All values were corrected for the refraction index of the solvents.

Synthesis of chemosensor L

A solution of 3-chloro-1-propanethiol (0.333 g, 3.01 mmol) in acetone (50 mL) was added dropwise to a solution of fluorescein (free acid) (0.5 g, 1.505 mmol) in acetone (100 mL). The resulting solution was gently maintained with magnetic stirring for ca. 48 h at room temperature. After that, the solution was gently refluxed with magnetic stirring for ca. 2 h. The solvent was evaporated under reduced pressure, yielding an orange oil. The oil was washed with hexane, leading to an orange powder that was characterized as chemosensor L. L was soluble in acetone, methanol, ethanol and DMSO and insoluble in chloroform, dichloromethane, ethyl acetate and water.

L: Colour: orange powder. (95 %): Anal. Calc. for C26H26O6S2: C, 62.6; H, 5.2; S, 12.8. Found: C, 62.1; H, 4.7; S, 13.0 %. IR (KBr windows) cm−1: ν(C=C) 1,595, ν(S–H) 2,560. NMR spectra: δ H (DMSO-d 6, 400 MHz) ppm: 1.96 (t, J = 7.0 Hz, 2H, H2), 2.06 (t, J = 6.8 Hz, 2H, H1), 2.19 (s, 2H, SH24), 2.70 (t, J = 7.0 Hz, 2H, H2′), 2.82 (t, J = 6.8 Hz 2H, H1′), 3.73 (m, J = 6.5, Hz 4H, H3), 6.54 (s, 2H, H9 and H15), 6.68 (s, 2H, H5 and H13), 7.25 (s, 2H, H6 and H12), 7.70–7.72 (m, 2H, H20 and H18); 7.76–7.80 (m, 1H, H19); 7.99 (d, 1H, H21); δ C (DMSO-d 6, 400 MHz) ppm: 26.61 (C1), 31.32 (C2), 34.40 (C3), 83.05 (C10), 102.14 (C9, C15), 109.49 (C7, C11), 112.51 (C5,C13), 124.49 (C18), 126.08 (C22), 128.92 (C21, C20), 129.95 (C6, C12), 135.47 (C19), 151.76, 152.37 (C4, C14), 159.39 (C8, C16, C17), 168.59 (C23). MALDI-TOF/MS (m/z): 481.60 [L+H]+. The value of the luminescence quantum yield of L was 0.6989. This value was measured using a solution of fluorescein in sodium hydroxide solution (0.1 M) as a standard ([Φ] = 0.79).

Synthesis of gold and silver nanoparticles in aqueous solution

Preparation of AuNPs@L

Au@citrate NPs were prepared in aqueous solution following the Turkevich method (Turkevich et al. 1951; Enüstün and Turkevich, 1963; Turkevich 1985a, b). An aqueous solution (125 mL) of hydrogen tetrachloroaurate (0.125 mmol, 49.5 mg) was added rapidly to a solution of 1 % (P/V) sodium citrate (12.5 mL) that was heated under reflux. Heating under reflux was continued for an additional 5 min, during which time the colour changed to deep red. An ethanolic solution (10 mL) of chemosensor L (0.096 mmol, 4.8 mg) was added to 50 mL of AuNPs@citrate and stirred for 24 h (Fig. 1). After that, the resulting mixture was collected by centrifugation three times at 9,000 rpm for 30 min and the supernatant was carefully removed up. The pellet containing the AuNPs@L were re-suspended in 20 mL of milli-Q water. The formation of the AuNPs@L can be observed because of the appearance of the gold plasmonic resonance band (PRB) at 500–600 nm, that is a characteristic of well-dispersed NPs with sizes in the range of 20 ± 5 nm. The absorbance of bound fluorescein at ca. 500 nm leads to a broadening of Au plasmon band (Fig. S1).

Fig. 1
figure 1

Schematic synthesis of the fluorescein-functionalized gold (AuNPs@L) and silver (AgNPs@L) nanoparticles

Full size image

Preparation of AgNPs@L

An aqueous solution (20 mL) of silver nitrate (0.008 mmol, 1.37 mg) was mixed with a solution of 0.3 % (P/V) polyvinylpyrrolidone (PVP) (90 mg) in milli-Q water (5 mL). The mixture was vigorously stirred and a fresh aqueous solution (10 mL) of sodium borohydride (0.050 mmol, 1.891 mg) was added quickly while stirring, during which time the colour changed to yellow.

After that, the resulting mixture was collected by centrifugation three times at 9,000 rpm for 30 min and the supernatant was carefully removed up to eliminate the NaBH4 and PVP in excess. The pellet containing the AgNPs@PVP was re-suspended in 30 mL of milli-Q water.

The formation of the NPs can be observed because of the appearance of the silver SPR at ca. 400 nm, that is a characteristic of well-dispersed NPs with sizes in the range of 10 ± 5 nm (Figure S2).

The formation of AgNPs@L from AgNPs@PVP in aqueous solution was followed by UV–Vis absorption and emission spectroscopy (Fig. 2).

Fig. 2
figure 2

Spectrophotometric (a) and spectrofluorometric (b) titrations of AgNPs@PVP with the addition of L in aqueous phase ([L] = 3.00 × 10−3, room temperature, λ exc = 400 nm). The inset shows the absorption at 400 nm

Full size image

Results and discussion

Small stable gold and silver NPs stabilized with chemosensor L were synthesized to explore their potential application for the detection of silent metal ions (Hg2+, Hg+, Cd2+, Zn2+, Pb2+ and Ag+) as well as Cu2+ in aqueous solution.

First, the formation of AuNPs@L was made by exchanging of the citrate molecules located in the NPs surface of AuNPs@citrate (Turkevich 1985a) for chemosensor L (Fig. 1). After 24 h, the Zeta potential changed from −33.0 mV/cm (AuNPs@citrate) to −35.0 mV/cm (AuNPs@L), introducing more stability in the system. The formation of AuNPs@L with sizes in the range of 20 ± 5 nm was confirmed by the presence of the SPR band between 500 and 600 nm in the UV–Vis spectrum (Figure S1).

The fluorescence study of the composite particle solution of AuNPs@L shows that the luminescence of the fluorescein was totally quenched by the gold metal core. This quenching occurs when the emission spectrum overlaps with the gold SPR band (Sapsford et al. 2006; Gersten and Nitzan 1981; Oh et al. 2005).

In order to keep the emission property of fluorescein units, silver NPs with PVP instead of citrate-like stabilizer were synthesized (Fig. S2). One of the principal advantages of PVP is that it presents higher molecular weight, resulting in the formation of a complete network in solution and is a biocompatible molecule that allows higher biological applications (Zielinska et al. 2009). Although the Zeta potential of AgNPs@PVP was −12.6 mV/cm, the system appears to be very stable by sterical stabilization.

The formation of AgNPs@L with sizes in the range of 11 ± 5 nm was made by exchanging of the PVP molecules located in the NPs surface of AgNPs@PVP for chemosensor L (Fig. 2). After the addition of 5 μL of L to the system AgNPs@PVP, the exchange was totally completed. The absorption spectra shows the SPR band of Ag near 400 nm showing the absorbance of bound fluorescein at ca. 500 nm as a hump near Ag plasmon band (Fig. 2a).

In this acid medium, the fluorescein ring closes and the protonation of the thiol groups takes place. This medium supports the approach of the protonated thiol groups towards the surface of the AgNPs@PVP and the superficial exchange of PVP for L occurs. A fluorescence enhancement is observed due to increased excitation rate of fluorophore by local field enhancement (Kuhn et al. 2006) (Fig. 2b).

With the aim to confirm the influence of the pH value in the absorption behaviour, a pH titration of the AgNPs@L was carried on Fig. 3. TEM images of AgNPs@L in aqueous phase to different pH values were also shown. The initial pH of the AgNPs@L in aqueous phase was 4.2. At this initial value the AgNPs are located very close and only the SPR band of AgNPs was observed at ca. 400 nm. At it is shown in the absorption spectrum of Fig. 3, at pH > 7, an increase in the band centred at ca. 510 nm, assigned to the isobenzofuran-1(3H)-one group in the open form, was observed. This form of the fluorescein in the opened form promoted a negative charge on the surface of the NPs that produces the separation between them and consequently an increase in their stability. Completely opposite behaviour was observed at more acidic conditions (pH 2.01) because at this pH value the isobenzofuran-1(3H)-one group in the closed form promoted the aggregation of the AgNPs.

Fig. 3
figure 3

Influence of the pH in the absorption (a) and emission (b) spectra of AgNPs@L in aqueous phase ([L] = 3 × 10−3 M, room temperature). TEM images of AgNPs@L in aqueous phase to different pH

Full size image

The sensing of Hg2+ in aqueous solution using the fluorescein-functionalized AuNPs@L and AgNPs@L was carried out. Figure 4a shows the changes in the absorption spectra of the colloidal system AuNPs@L with the addition of increasing amounts of Hg(NO3)2. As a result of this interaction, a new absorption band centred at ca. 635 nm was observed due to the formation of chains in which Hg2+ ions played the role of connectors between the AuNPs@L (see TEM images). The colour changes from red to blue, and indicates the presence of this silent heavy-metal ion in solution (Fig. 4a).

Fig. 4
figure 4

a Spectrophotometric titration of AuNPs@L with the addition of increasing amounts of Hg(NO3)2, naked eye detection and TEM images of AuNPs@L with the addition of 200 μL of [Hg(NO3)2] = 1.00 × 10−3 M. Modifications with time in the absorption spectra and TEM images of an aqueous solution of AuNPs@L with the addition of 25 μL b 50 μL c and 100 μL d of [Hg(NO3)2] = 1.00 × 10−3 M

Full size image

In the final point of the titration of AuNPs@L with Hg2+, the Zeta potential changes from −35 to −27 mV/cm. This means that a destabilization of the system took place, as can be seen in the TEM images (Fig. 4). The formation of the 1D chains between AuNPs@L modulated for the Hg2+ metal ions was possible by chelation of the Hg2+ ions with the carboxylate groups of chemosensor L located in the AuNPs surface (Fig. 5).

Fig. 5
figure 5

Schematic representation of the chelation interaction of the Hg2+ ion with the carboxylate groups located in the NPs surface (the picture is not to scale)

Full size image

In order to evaluate the detection and the quantification limits of this system, for the detection of Hg2+ in aqueous solution, the evaluation of the SPR band was carried out (Fig. 4b–d). Thus, the minimum amount of mercury detectable at 635 nm was 0.68 ppm and the minimum amount quantified was 1.60 ppm.

From this UV–Vis spectral study, it was observed that, after the addition of 6.70 ppm of Hg2+ ion, the new SPR band developed at higher wavelength became saturated and the TEM micrograph revealed that the AuNPs were totally assembled into small 1D chains (Fig. 4). Figure 6 shows that a specific interaction between AuNPs@L with Hg2+ was observed in comparison with the other metal ions explored.

Fig. 6
figure 6

Bar chart (a) and visual colour changes (b) observed for an aqueous solution of 1 AuNPs@L after titration with 2 Cu2+ 3 Ag+ 4 Zn2+ 5 Cd2+ 6 Hg+ 7 Hg2+ and 8 Pb2+

Full size image

On the other hand, a selective interaction of AgNPs@L with Hg2+ was also observed in aqueous phase. As a result, a modification of the AgNPs structure, the formation of 1D chains between them and a colour change from yellow to orange of the aqueous solution of AgNPs@L were observed in the presence of mercury(II) (Fig. 7). TEM images show the aggregation process by the interaction of the AgNPs@L with the other explored metal ions, without an orderly disposition in chains to that observed for Hg2+ (Figure S3).

Fig. 7
figure 7

Spectrophotometric (a) and spectrofluorimetric (b) titration of AgNPs@PVP with compound L in absolute ethanol with the addition of Hg2+ in aqueous solution ([L] = 3.00 × 10−4 M, [Hg2+] = 1.00 × 10−3 M, room temperature, λ exc = 400 nm). Insets show absorption at 400 and 487 nm. Naked eye detection and TEM images of AgNPs@L before (yellow color) and after (orange color) the addition of 200 μL of [Hg(NO3)2] = 1.00 × 10−3 M. (Color figure online)

Full size image

Similar chelation interaction of the Hg2+ ions with the carboxylate groups of chemosensor L located in the AgNPs surface was predicted to that shown in Fig. 5, being responsible for the formation of chains between the silver NPs. In this case, the minimum amount of mercury detectable at 400 nm was of 0.34 ppm and the minimum amount quantified was 2.05 ppm.

As it is shown in Fig. 7b, an increase in the emission band centred at ca. 520 nm was also observed for the system AgNPs@L in the presence of Hg2+. This result suggests that the interaction of Hg2+ with the free molecular probe does not take place only by the external carboxyl group of the isobenzofuran-1(3H)-one group, because this effect could produce the quenching of the emission for system L. It is important to mention that probably, a fraction of the Hg2+ metal ions added are located in the inner cavity of the ligand anchored to the silver metal surface for the thiol groups and the internal oxygen atoms exert a cooperative effect in the coordination of the metal ion. Probably this could also take place after the titration of AuNPs@L with Hg2+.

Two main reasons support the existence of this binding mode: first, the negative superficial charge of the silver nanoparticle, and, second, the ionic radius of this metal ion is small enough to favour the accommodation of the metal ion in the inner cavity. This effect was also observed in similar recently reported systems (Fernández-Lodeiro et al. 2013).

Moreover, the interaction of the remaining fraction of the Hg2+ ions with the external carboxylate groups of chemosensor L promoted the assembly of the AgNPs into small 1D chains. The formation of these molecular systems introduces more rigidity in the system that at the same time is in agreement with the increase in the emission intensity observed in the AgNPs@L after the interaction with Hg2+ metal ions.

Conclusions

In conclusion, new fluorescein-functionalized AuNPs@L and AgNPs@L have been employed in aqueous solution as colorimetric probes for the detection of mercury ions by 1D-chain formation. Changes in the SPR bands as well as colours due to the network of 1D assemblies of NPs were observed. The AuNPs@L was totally assembled into 1D chains after the addition of 6.70 ppm of Hg2+ in aqueous solution. The chelation interaction of the Hg2+ ions with the carboxylate groups of chemosensor L located in the AuNPs and AgNPs surface is responsible for the selective formation of chains between the NPs modulated for the Hg2+ metal ions.