Introduction

Attributed to the toxicity and World Health Organization (WHO) restriction on permitted level of Cr(VI) ions than those of Cr(III) ions, the development of diverse probes for its selective and sensitive detection is still in need [1, 2]. In a similar trend, due to prominence role over several nervous functions, determination of Dopamine (DA; a catecholamine derivative) is also become vital [3]. The presence of toxic Cr(VI) ions and abnormal or insufficient levels of DA might led to severe disease, such as DNA damage, Alzheimer’s disease, depression, schizophrenia, and Parkinson’s disease [4,5,6]. Therefore, many analytical techniques have emerged for Cr(VI) and DA quantifications. For example, Cr(VI) ions quantifications can be done by gas chromatography, inductively coupled plasma mass spectrometry, atomic absorption spectrometry, and electrochemical studies [7, 8]. On the other hand, DA can be detected through voltametric analysis, chromatographic techniques, and immunoassays [9, 10]. Similar to those conventional methods, the design of fluorescent probes for Cr(VI) and DA detection are still an analytical challenge due to the orthodox sample preparation, high sensitivity and imaging applicability [11].

Luminescent nanomaterials, such as carbon dots (C-dots), quantum dots (QDs), and metallic nanoclusters (MNCs), display the inspiring detection of Cr(VI) ions and DA in biological and environmental samples [12,13,14,15,16,17]. For instance, Mutuyimana et al., described the yellow emissive C-dots for the detection of Cr(VI) ions in fish and shrimps [18]. Similarly, Cui et al., presented the fluorescent iridium nanoclusters (IrNCs) for the effective determination of Cr(VI) ions in real samples [19]. On the other hand, a fluorometric based assay of DA in serum and cell imaging studies was reported by aptamer-functionalized Molybdenum disulfide (MoS2) QDs and nanosheets [20]. However, there is no report so far on bimetallic-nanoclusters (bi-MNCs) based simultaneous detections of Cr(VI) ions and DA, which provides us an initiative to proceed in that direction.

Bi-M NCs usually can be developed by following methods: (1) galvanic exchange; (2) anti-galvanic reduction; (3) thiol-etching of bi-metallic nanoparticles; (4) re-organization of bi-metallic species; (5) potential deposition [21]. Among them, anti-galvanic reduction tactic is quite impressive and provides highly emissive bi-M NCs [22]. For example, Sun and co-workers suggested the one-pot synthesize of 11-mercaptoundecanoicacid (11-MUA) capped Ag/Au bimetallic nanoclusters with 6.81% Φ value via anti-galvanic reduction and employed as pH sensors [23]. Similar to 11-MUA, cysteamine (CA) can also act as an effective stabilizer of M NCs or bi-M NCs through M-S bond and afford free amine (-NH2) for particular analyte detection [24]. Therefore, by adapting the anti-galvanic reduction methodology and utilizing cysteamine as ligand, it is possible to develop cysteamine capped gold-copper nanoclusters (CA-AuCu NCs) as a chemosensory probe.

In this report, CA-AuCu NCs are developed with strong blue emission (Φ = 18%) via one pot anti-galvanic reduction. Upon exposure to metal ions and aminoacids, the CA-AuCu NCs reveal great selectivity to Cr(VI) ions and DA through aggregation mediated static quenching mechanism. In this study, the sensory responses of the CA-AuCu NCs are well authenticated in real, spiked urine and cellular imaging investigations.

Experimental

The general information on stock solutions, Instruments used, sample preparation procedures for TEM, FTIR, XPS studies, MTT assay and cell imaging protocols of CA-AuCu NCs, CA-AuCu NCs + Cr(VI) and CA-AuCu NCs + DA with corresponding data are delivered in Electronic Supporting Material (ESM).

Synthesize of CA- AuCu NCs [23]

4 mL of 10 mM Tetrachloroauric acid (HAuCl4) and 40 μL of 10 mM copper nitrate (CuNO3) were added into 31.96 mL of DI water solution (final volume 40 mL) and stirred at 550 C for 30 min to obtain the homogeneous mixture. 4 mL of 1 M cysteamine (CA) was dropped into above homogeneous solution within 5 min and then cooled down to room temperature. Subsequently, 12 mL of 1 M sodium hydroxide (NaOH) was added and stirred at room temperature for 3 h. After that, the mixture was stirred at 700 C for 0 to 72 h. Finally, to rid out the excess HAuCl4, CA and NAOH, the luminescent CA-AuCu NCs solution was subjected to centrifugation and several water washing at 3500 rpm for 1 h. The CA-AuCu NCs were re-dispersed in DI water (25 mg/100 mL) and stored in refrigerator at 4 °C for further use.

Protocol for determination of Cr(VI) and DA

Deionized water (DI), laboratory tap water, NCTU Lake and Pacific Ocean Sea (located in Hsinchu, Taiwan) water samples were collected and filtered over a 0.2 μm membrane. From standard solutions (10, 100 and 1000 μM) of Cr(VI) ions, different volumes were spiked in to 200 μL of DI/lake/tap/sea water samples and subjected to PL based assay. Those samples were added to the solution containing 25 μg of CA-AuCu NCs dispersed in DI water and 400 μL of HEPES (pH -7) buffer. The diverse concentrations of Cr(VI) ions were made-up in respective DI/real water samples and spiked to the above CA-AuCu NCs in HEPES (pH -7) buffer. Next, the PL spectra were recorded after 15 min at room temperature. The results of above conventional sensing method were compared with ICP-MS analysis. For DA assay, similar procedure was followed by accumulating various concentrations of DA standard solutions (1 and 10 mM) in respective DI/real water samples to CA-AuCu NCs in HEPES (pH -7), incubated for 2 h at room temperature and the PL spectra were recorded. For both Cr(VI) and DA assay, the excitation wavelength was set at 350 nm; excitation and emission slits were fixed as 5 nm and the PL recorded between 400 to 600 nm ranges.

Spiked human urine analysis

For practical applications, the detection of Cr(VI)/DA were done in spiked human urine sample inquiry by standard addition method agreed with related reports [25, 26]. Before the analysis, urine samples were diluted 150 times with distilled water to decrease the matrix effect. No other pre-treatment was executed. Thereafter, above declared procedure was followed for urine based detection of Cr(VI) ions or DA, which was further investigated by PL spectra. The excitation wavelength was at 350 nm with excitation and emission slits fixed as 5 nm and the PL was recorded between 400 to 600 nm ranges.

Results and discussion

Choice of materials

Cysteamine (CA) was chosen as a stabilizer consist of free thiol (-SH) that can functionalize over metal surface through M-S bond and remnants free amine (-NH2) group to coordinate with specific analyte. Based on galvanic series, noble Au and Cu atoms were selected at 100:1 (Au/Cu) molar ratio to produce highly emissive CA-AuCu NCs. To induce effective anti-galvanic reduction, 1 M NaOH was used as a basic medium, which improved nanostructure affordability and galvanic exchange. Furthermore, the comparable low toxicity, stability and electronic structure of CA-AuCu NCs than those of well-known C-dots, semiconductor QDs and noble NPs make them more favourable for practical use. Note that, contrary to colorimetric responses of noble NPs, CA-AuCu NCs can be applied in fluorometric assays.

Synthesis and characterization of CA-AuCu NCs

As shown in Fig. 1, the CA-AuCu NCs are synthesized via one-pot anti-galvanic reduction [23]. But, CA-AuCu NCs with high fluorescent quantum yield (Φ) are attained through optimization of HAuCl4, CuNO3, Cysteamine (CA), and NaOH concentrations, as well as reaction time. Overall, the best emissive results of CA-AuCu NCs were accomplished at (1) 436 nm under excitation wavelength at 350 nm (see Fig. S1a and Table S1, ESM); (2) 1:0.1:100:300 equivalents of HAuCl4: CuNO3: Cysteamine (CA): NaOH, respectively (see Table S2, ESM); (3) after 48 h reaction time (see Fig. S1b and Table S2, ESM). Compared to Riboflavin monophosphate (Φ = 26.8% in water), the CA-AuCu NCs possess the Φ value of 18%. Increase of excitation wavelengths from 310 to 380 nm induced red shift of PL peaks because the electrons stimulated to the lower energy level relaxed and then produced emission at longer wavelengths.

Fig. 1
figure 1

Synthesis of CA-AuCu NCs; a 550 C, Cysteamine (CA), RT, 3 Hrs; b Stir at 700 C, 48 Hrs

Full size image

The FTIR spectra of CA-AuCu NCs (Fig. S2, ESM) confirm the cysteamine functionalization via more diverse peaks than that of cysteamine. Wherein, the –CH and –NH2 bands are broadened between 2950 ~ 3600 cm−1 and the –NH asymmetric stretching occurs at 1670 cm−1 instead of at 1612 cm−1 (as in the case of the CA). Above stretching bands also confirm the possible free –NH2 groups. On the other hand, the free thiol (-SH) band of Cysteamine at 2569 cm−1 is almost vanished for the CA-AuCu NCs, hence confirmed the (Au-Cu)-S bond formation [27]. The HR-TEM image (Figs. S3a, ESM) confirms that the CA-AuCu NCs have a size between 1~5 nm [28]. However, among selected 100 particles, more than 50% have an average size of 3 nm as displayed in Fig. S3b (ESM). Besides, from Figs. S3c and d (ESM), the diffraction value is estimated as 0.218 which lies between the values of Au and Cu and is in good agreement with the 111 pattern of the Au-Cu alloy [29]. Moreover, DLS studies reveal an average size of 4.6 ± 3.3 nm for the CA-AuCu NCs (Fig. S4, ESM), which is fairly similar to the HR-TEM results. Next, zeta potential (ζ) of CA-AuCu NCs is established as −12.76 mV, (Fig. S5, ESM) and hence can be stabilized via –Ve repulsive force between particles.

As seen in Fig. S6 (ESM), the Maldi-Tof of CA-AuCu NCs give m/z values of 1883.47 and 2136.56, which can be related to CA10Au5Cu2 NCs and CA10Au5Cu5 NCs, respectively. However, it is well recognized that the majority of NCs have the molecular structure of CA10Au5Cu2 NCs and it also confirms the cysteamine capping over Au-Cu alloy surface [30]. Results from the XPS of CA-AuCu NCs demonstrate the presence of Au4f, Cu2p, S2p, C1s, and N1 s peaks (Fig. S7a, ESM). The Au4f peaks occur at 84.8 eV (4f7/2) and 88.4 eV (4f5/2) as displayed in Fig. S7b (ESM). Likewise, Cu2p peak at 935.6 eV (2p3/2) can be clearly seen in Fig. S7c (ESM). However, due to deficient amount of Cu atoms, the peak of 2p1/2 was not observed. It is noteworthy that the absence of the satellite peak at 941.9 eV (associated with the Cu+) also confirms its clusterification [31]. The appearance of the S2p, C1s, and N1 s peaks at 164.4, 284.7, and 401.8 eV (Figs. S7d-f, ESM), correspondingly, proves the existence of S, C and N atoms over Au-Cu alloy. Due to the possible formation of (Au-Cu)-S bond, the S2p peak is strongly affected, hence confirms the cysteamine functionalization.

Analytical responses of CA-AuCu NCs

Initially, 300 μM of numerous metal ions [Na(I), Ni(II), Fe(III), Cd(II), Ca(II), Ga(III), Cr(III), Y(III), Cu(III), Fe(II), Mg(II) Ba(II), Au(III), Ag(I), Co(II), Zn(II), Pb(II), Sn(II), Al(III), Cr(VI) and Hg(II)] or 3 mM of other bio-analytes (Glucose, Cysteine, Homo-Cysteine, Lysine, Melamine, Glutamic Acid, Histidine, Tyrosine, Sucrose, Ornithine, Arginine, Ascorbic Acid, Glycine, Urea; Dopamine and Glutathione) were added to the CA-AuCu NCs solution in DI water and the respective fluorescence changes were recorded subsequently. The above investigations reveal excellent fluorescence quenching to Cr(VI) ions and DA than that of other interferences as shown in Fig. 2a and b. Under UV-lamp irradiation at 365 nm, the visual changes are clearly demonstrated by the photo images [Figs. 2a, b (insets), S8 and S9 (ESM)]. During the discovery of DA, a dark brown color is observed visually as shown in Fig. S9a (ESM). By studying different combinations of pH, concentrations of probe/analyte, and incubation time [Figs. S10-S12 (ESM)], the best sensory result was identified at (1) pH -7 HEPES buffer; (2) 25 μg/mL of CA-AuCu NCs concentration; (3) 300 μM and 1 mM of Cr(VI) and DA, individually (4) 15 min and 2 h incubation time for Cr(VI) and DA detection, respectively.

Fig. 2
figure 2

a Fluorescence responses of CA-AuCu NCs towards metal ions (λex/em = 350/436 nm); Inset: PL change by Cr(VI) ions (under UV lamp, λex = 365 nm); b Fluorescence responses of CA-AuCu NCs towards dopamine (DA) and other interferences (λex/em = 350/436 nm); Inset: PL change by DA (under UV lamp, λex = 365 nm)

Full size image

Next, the exceptional sensitivity of CA-AuCu NCs to Cr(VI) and DA is demonstrated by single and dual analyte studies in HEPES buffer (pH 7). In contrast to the remaining interferences, the CA-AuCu NCs show remarkable selectivity towards Cr(VI) and DA with 12 and 25 folds of PL quenching, respectively, as seen in Figs. S13a and b (ESM). At 1:5 ratio [Cr(VI) to other interferences], except the mild effect Al3+ and Cr3+ ions (7 folds quenching), none of other metal ions reveal any interfering effect on Cr(VI) selectivity as displayed in Fig. S14 (ESM). On the other hand, in the presence of all other metal ions, the PL quenching is enhanced between 20 ̴ 120 folds.

In a similar fashion, at 1:3 ratio (DA to other interferences), except Ascorbic acid (AA; 3.9 folds quenching) and Tyrosine (Tyr; 3.7 folds quenching), none of other interferences affect the selectivity of DA as shown in Fig. S15 (ESM). However, at 1:1 mixing ratio of AA or Tyr, the selectivity of CA-AuCu NCs is not strongly affected and still displays 16 folds in PL quenching. This may be attributed to the structural similarity or redox potential similarity of both AA and Tyr to DA, and hence interfere the selective quenching. In particular, with excess amount of AA and Tyr, the -OH of Tyr and two similar –OH groups of AA may exceptionally interfere the electron transfer between catechol of DA to CA-AuCu NCs involved in the sensing mechanism. Therefore, to attain the reliable results with CA-AuCu NCs, complex samples containing excess of AA and Tyr should be avoided. From Figs. S16a and b (ESM), the Stern-Volmer Quenching Constant (KSV) values [32] of CA-AuCu NCs + Cr(VI) and CA-AuCu NCs + DA are calculated to be 8.26 × 104 and 2.426 × 104 M−1, correspondingly. Moreover, linearity in fluorescent responses of CA-AuCu NCs by Cr(VI) and DA also demonstrates the feasible static quenching. As noted in Tables S3 and S4 (ESM), the KSV values of residual metal ions and bio-interferences oscillate between 102 ~ 103 M−1, hence proves the selective discrimination of Cr(VI) and DA.

Upon the accumulation of Cr(VI) ions (0 ~ 300 μM) or DA (0 ~ 1 mM) to CA-AuCu NCs in HEPES buffer (pH 7), the fluorescent peak at 436 nm is linearly quenched as depicted in Figs. 3a and b. In accordance with the above PL changes, the Φ values of CA-AuCu NCs + Cr(VI) and CA-AuCu NCs + DA are also reduced to 0.43 and 1.5%, respectively. From these titrations, linear regressions of Cr(VI and DA are adopted as 0.2 ̴ 100 and 0.4 ̴ 250 μM, correspondingly, as seen in Figs. 3c and d. By following the 3σ (signal to noise) and the linear regression equations of y = 0.00788x + 2.53566 × 10−4 (R2 = 0.98873, n = 4) and y = 0.01806x + 5.81193 × 10−4 (R2 = 0.9867, n = 4), the LODs of Cr(VI) ions and DA are calculated as 80 and 135 nM, respectively, which is below the prescribed limits by WHO.

Fig. 3
figure 3

a PL changes of CA-AuCu NCs in HEPES buffer (pH -7) with Cr(VI) ions (0 to 300 μM; λex/em = 350/436 nm); b PL changes of CA-AuCu NCs in HEPES buffer (pH -7) with dopamine (DA; 0 to 1 mM; λex/em = 350/436 nm); c, d detection limits (LODs) of Cr(VI) ions and DA calculated via standard deviation and fittings (n = 4; λex/em = 350/436 nm)

Full size image

Mechanism involved in Cr(VI) and DA detections

The presence of Cr(VI) ions and DA induced aggregation of CA-AuCu NCs are visualized in Figs. S17a and b (ESM). Similarly, for CA-AuCu NCs + Cr(VI) and CA-AuCu NCs + DA systems [Figs. S18a and b (ESM)], particle sizes of 129.9 ± 34.8 and 179.7 ± 51.1 nm determined by DLS are 28 and 39 folds larger than that of origin, respectively. Therefore, aggregation induced fluorescent quenching phenomena may possibly involve in Cr(VI) and DA detection [33], which can be further enhanced by manipulating the binding forces via FTIR interrogations. During the Cr(VI) recognition, free –NH2 band of CA-AuCu NCs at 3000 ̴ 3500 cm−1 (shown in Fig. S19a, ESM) is broadened along with the affected symmetric or asymmetric stretching bands. Hence, it indicates that coordination possibly takes place between free –NH2 groups and Cr(VI) ions. In contrast, FTIR spectrum of CA-AuCu NCs + DA is completely broadened with mild shifting of –NH symmetric /asymmetric bands (Fig. S19b, ESM), which suggests feasible diverse mechanism. As reported earlier, the CA-AuCu NCs may attract the two acidic catechol protons of DA to form the stable phenoxide-enolate equilibrium [34] and lead to –Ve charge over aromatic ring and charge transfer between CA-AuCu NCs and DA to induce fluorescent quenching or ACQ. The aforementioned mechanisms are authenticated by zeta potential (ζ) data. For CA-AuCu NCs + Cr(VI), ζ value is shifted to +14.12 mV (Fig. S20, ESM) compared to that of −12.76 mV for CA-AuCu NCs, which supports the feasible coordination of free –NH2 with Cr(VI) ions. Likewise, due to the diverse mechanism on DA determination, +Ve ζ value (+19.00 mV) is observed as shown in Fig. S21 (ESM). Finally, based on the TEM, DLS, FTIR, and zeta potential studies, these mechanisms are schematically illustrated in Figs. S22, and S23 (ESM).

Determination of Cr(VI) and DA

The distinct quantifications of Cr(VI) and DA can be done as follows. (1) At 200 μM concentration, Cr(VI) ions induce rapid fluorescent quenching than that of DA, which only prompt colorimetric response (colorless to dark brown); (2) Contrast to Cr(VI), higher concentration of DA (3 mM) is required to achieve immediate fluorescence quenching; (3) Consecutive addition of EDTA to CA-AuCu NCs + Cr(VI), the fluorescence recovered and extended up to four cycles (Figs. S24a-c, ESM). Above reversibility is effective at mixed states [Cr(VI) ions(200 μM) and DA (200 μM) with CA-AuCu NCs] as demonstrated in Figs. S25a and b (ESM).

Real and human fluid analysis

As presented in Table 1, the results of CA-AuCu NCs towards Cr(VI) in real and urine samples are in good agreement with ICP-MS data with less than 2.17% relative error. The recovery values of Cr(VI) ions in tap, lake, sea, and urine water fall between 110.42~116.01%, 109.03~116.41%, 106.22~114.03%, and 110.29~118.02%, correspondingly. Similarly, the PL based recovery of DA shows a relative error of less than 3.42% as shown in Table 2. The recovery values of DA in Tap, Lake, Sea, and Urine water lie between 99.78~100.33%, 98.45~101.45%, 98.73~99.77%, and 100.81~102.63%, individually. Separate titrations of CA-AuCu NCs with Cr(VI) ions and DA in Tap, Lake, Sea, and spiked urine water also experience the fluorescence quenching as displayed in Figs. S26 and S27 (ESM). However, the DA detection in sea water is strongly affected than the others, which is attributed to the salts presented in the medium. The linear regressions of Cr(VI) ions in tab/lake/sea/urine samples are found to be 0.3 ̴ 160, 0.4 ̴ 160, 0.5 ̴ 100 and 0.2 ̴ 120 μM, individually. Likewise, linear regressions of DA in tab/lake/sea/urine samples are established as 1 ̴ 300, 1 ̴ 500, 50 ̴ 600 and 0.4 ̴ 700 μM, independently. From standard deviation and linear fittings and 3σ [signal to noise, n = 4; Figs. S28 and S29 (ESM)], the LODs of Cr(VI) ions and DA in those real samples are estimated. The LODs of Cr(VI) in tap, lake, sea, and urine samples are determined to be 99, 106, 114, and 85 nM, correspondingly. The LODs of DA in tap, lake, sea, and urine samples are calculated as 171 nM, 223 nM, 1.465 μM, and 206 nM, respectively. Table S5 (ESM) summarizes the linear ranges and LODs of Cr(VI) and DA which prove their successful detection in real and urine samples. Photographs of time and concentration dependent fluorescent and visual changes of CA-AuCu NCs to Cr(VI) and DA in real/urine samples are displayed in Figs. S30-S37 (ESM).

Table 1 The recovery values of spiked Cr(VI) ions in real/urine samples and the relative standard deviation (RSD) and ICP-MS values
Full size table
Table 2 The recovery values of spiked DA in real/urine samples and the relative standard deviation (RSD) values
Full size table

Cellular imaging of Cr(VI) and DA

Before moving on to cell imaging, the cytotoxicity of CA-AuCu NCs is established in Fig. S38a (ESM). In which, CA-AuCu NCs demonstrate the viability of 80% even at a concentration of 60 μg/mL, and hence prove low in toxicity [35]. The IC50 value of CA-AuCu NCs is predicted as 149 μg/mL as displayed in Fig. S38b (ESM). Upon incubation of those low toxic CA-AuCu NCs with Raw 264.7 cells, bright blue luminescent cell lines are visualized as shown in Fig. 4. These luminescent cells are then stained with Cr(VI) or DA and incubated for either 30 or 60 min. Once stained and incubated, luminescence in cells begins to quench. Therefore, we conclude that CA-AuCu NCs-based Cr(VI) and DA detection can be done in bio-samples. Moreover, in compared to earlier diverse methods, Tables S6 and S7 (ESM) show that Cr(VI) and DA discriminations are equally effective in terms of detection methods, linear ranges, LODs and applicability.

Fig. 4
figure 4

Cellular imaging of CA-AuCu NCs in presence of Cr(VI) ions and dopamine (DA) in Raw 264.7 cell lines (scale: 0–25 μm)

Full size image

Conclusion

Highly emissive low toxic cysteamine capped gold-copper nanoclusters (CA-AuCu NCs) were synthesized via one-pot anti-galvanic reduction approach and employed in the detections of Cr(VI) ions and dopamine (DA) in real/urine samples. Notably, Cr(VI) ions and DA quantifications can be distinguished by their visual and fluorescent responses. However, DA detection is still limited to complex samples consist of higher volume of tyrosine or ascorbic acid as interferences. Demonstrated cellular imaging utilities and diverse aggregation/electron transfer mechanisms on Cr(VI) and DA assays may contribute to design such unique analytical probes.