3.1 Characterization of Fe-HHTP@HZIF-8@AuNCs
Hybrid nanozyme Fe-HHTP@HZIF-8@AuNCs was obtained through three mild and easy steps as shown in Scheme. 1A [18, 19]. Firstly, Fe-HHTP was synthesized from one-step self-assembly between Fe3+ and HHTP at room temperature. Then, by using Fe-HHTP as the core, ZIF-8 shell was formed through the coordination between Zn2+ and 2-methylimidazole. Lastly, based on cation exchange between Au3+ and Zn2+, Au element was embellished on the surface of Fe-HHTP@HZIF-8 in the format of AuNCs, forming Fe-HHTP@[email protected] demonstrate successful synthesis of Fe-HHTP@HZIF-8@AuNCs, a battery of integral characterization experiments were executed including TEM, HRTEM, XRD, EDS and XPS. TEM images of Fe-HHTP, Fe-HHTP@ZIF-8 and Fe-HHTP@HZIF-8@AuNCs were showed in Fig. 1A-C. From the comparison of these three graphs, Fe-HHTP displayed a larger lamellar structure, while Fe-HHTP@ZIF-8 presented typical rhombic dodecahedron structure, which was almost the same as that of ZIF-8 [22]. Fe-HHTP@HZIF-8@AuNCs generally maintained the rhombic dodecahedron structure of ZIF-8. However, some obvious hollow nanoparticles arisen from Kirkendall effect could be easily observed in Fig. 1C and Fig. S1. From HRTEM image in Fig. 1D, clear streaks on the surface of AuNCs were showed with lattice spacing of 0.215 nm [23]. From particle size distribution chart in Fig. 1E, the particle size range of AuNCs on the surface of Fe-HHTP@HZIF-8 was between 1.1–2.3 nm and was mainly distributed in the range of 1.3–1.9 nm, accounting for 77.8%. Their average diameter were measured to be 1.60 nm,which was much lower than AuNPs distributed on the surface of other MOFs [24, 25]. From XRD images in Fig. 1F, the diffraction peaks of Fe-HHTP@HZIF-8@AuNCs were basically consistent with that of ZIF-8, indicating its certain crystallinity resulting from MOF structure [18]. EDS element mapping in Fig. 1G showed the uniform distribution of six elements (C, N, O, Fe, Zn and Au) on the obtained nanoparticles, indicating its successful synthesis.
Moreover, to confirm the chemical composition of Fe-HHTP@HZIF-8@AuNCs, high-resolution XPS was conducted with overall spectrum shown in Fig. 1H. Six peaks at 88.1, 284.8, 399.5, 531.4, 723.9 and 1021.9 eV were ascribed to the characteristic peaks of Au 4f, C 1s, N 1s, O 1s, Fe 2p and Zn 2p, respectively. In detail, XPS spectra of Au element in Fig. 1I showed four distinct peaks. Two peaks at 87.7 and 91.3 eV corresponded to Au (III) ions, while the other two peaks at 84.8 and 88.6 eV indicated the presence of Au (0) nanoclusters, which might be produced from a partial reduction reaction between Au3+ and N atoms in ZIF-8 [26–28]. From high-resolution Fe 2p spectrum in Fig. 1J, four peaks at 710.8, 712.4, 723.9 and 727.1 eV corresponded to FeII 2p3/2, FeIII 2p3/2, FeII 2p1/2 and FeIII 2p1/2, respectively. In addition, two small satellite peaks was observed at 718.1 and 731.3 eV. These peaks illustrated the coexistence of Fe2+ and Fe3+ in Fe-HHTP@HZIF-8@AuNCs [20, 27, 29, 30]. Additionally, high-resolution XPS spectra of Zn 2p, C 1s, N 1s, O 1s was presented and analyzed in Fig. S2.
3.2 Oxidase-like activity of Fe-HHTP@HZIF-8@AuNCs
TMB was selected as a chromogenic substrate in the study about oxidase-like activity of Fe-HHTP@HZIF-8@AuNCs. As shown in Fig. 2A, TMB was oxidized to generate dark blue oxTMB product by Fe-HHTP@HZIF-8@AuNCs under air-saturated condition, which accompanied by an enhancement of the ultraviolet absorption peak at 652 nm. On the contrary, TMB oxidation reaction was dramatically inhibited in N2-saturated solution [20]. This phenomenon indicated that O2 in solution played a considerable role in the catalytic reaction, manifesting great oxidase-like activity of Fe-HHTP@HZIF-8@AuNCs. Additionally, the catalytic activities of different nanoparticles, including Fe-HHTP, ZIF-8, Fe-HHTP@ZIF-8 and Fe-HHTP@HZIF-8@AuNCs, were comparatively studied in Fig. 2B. Fe-HHTP@HZIF-8@AuNCs showed the highest oxidase-like catalytic activity, indicating important role of AuNCs in the improvement of oxidase-like activity of hybrid material. Also, the formed hollow structure could accelerate substrate molecules transfer and improve collision efficiency between Fe-HHTP core and TMB.
To maximize the oxidase-like catalytic properties of Fe-HHTP@HZIF-8@AuNCs, influence of pH, temperature and incubating time was fully investigated. As shown in Fig. 2C, Fe-HHTP@HZIF-8@AuNCs exhibited the best catalytic activity at pH 4.0 while its catalytic activity dropped sharply under strong acid or near neutral conditions. This tendency could be explained by the effect of pH value on the surface charge of materials and dissociation state [31–33]. As depicted in Fig. 2D, the catalytic activity of Fe-HHTP@HZIF-8@AuNCs gradually increased at 25 ~ 37℃, reached the highest peak at 37℃, and gradually decreased at 37 ~ 70℃. The influence of temperature on its oxidase like activity were similar to that of some other nanozymes [21, 34, 35]. Then, from the influence of incubation time presented in Fig. 2E, UV signal of oxTMB product showed an obvious upward trend during 1–5 min, and basically remained stable after 5 min. Therefore, optimal conditions for color reaction mediated by Fe-HHTP@HZIF-8@AuNCs were pH 4.0, 37℃ and 5 min’s incubation. To confirm the oxidase like activity of Fe-HHTP@HZIF-8@AuNCs from the view of radical, EPR spectroscopy was utilized to quantify the types of transient free intermediates. As shown in Fig. 2F, by using DMPO as a trapping agent of ·O2− and ·OH while TEMPO as a trapping agent of 1O2, the presence of ·O2−, ·OH and 1O2 was demonstrated in Fe-HHTP@HZIF-8@AuNCs mediated catalytic system. The content of ·O2 − was obviously higher than that of ·OH and 1O2. The production of reactive oxygen species (ROS) could promote the oxidation of TMB to the dark blue TMB complex [21]. Moreover, three types of free radicals enhanced gradually from 0 to 20 min ( Fig. S3).
The catalytic kinetic study of Fe-HHTP@HZIF-8@AuNCs was conducted by varying TMB concentrations in the range of 0.1–0.9 mM. According to reaction kinetic curve in Fig. 3A-B and Michaelis-Menten equation, Km and Vmax of Fe-HHTP@HZIF-8@AuNCs towards TMB were 0.21 mM and 1.74×10− 6 M·s− 1, respectively. This Km value was much lower than that of other nanozymes (Table S1), indicating that Fe-HHTP@HZIF-8@AuNCs has much higher affinity towards TMB and resulted higher catalytic activity. The resulted superior catalytic performance of Fe-HHTP@HZIF-8@AuNCs mighty be attributed to well dispersed AuNCs on the surface and the formed hollow ZIF-8 structure through Kirkendall effect.
3.3 Colorimetric method for nitrite detection
Based on the excellent oxidase-like activity of Fe-HHTP@HZIF-8@AuNCs, a ratiometric colorimetric method was established for the detection of nitrite in food. As shown in Fig. 4A and Scheme 1B-C, Fe-HHTP@HZIF-8@AuNCs catalyzed colorless TMB to produce corresponding blue oxide oxTMB, which subsequently converted to yellowish-brown diazotized oxTMB in the presence of nitrite. Thereby, on the basis of the color changes of reacted solution, highly sensitive detection of nitrite could be realized by measuring of UV signal changes at the wavelength of 652 nm and 445 nm. As illustrated in Fig. 4B, with the increase of nitrite concentration in the range of 0 ~ 700.0 µM, the color of reacted solution mediated by Fe-HHTP@HZIF-8@AuNCs gradually changed from blue to yellowish-brown, accompanied by decreased UV absorbance at 652 nm and increased UV absorbance at 445 nm.
Then, to establish ratiometric colorimetric method for nitrite, the linear relationship between UV absorbance ratio (A652/A445) and nitrite concentration was explored in the linear range of 2.0 ~ 400.0 µM. As presented in Fig. 4C, the calibration curve equation was A652/A445 = 7.049–1.951 log([NO2−]) (R2 = 0.993), where logarithm nitrite concentration (log([NO2−])) was the independent variable while ratio absorbance (A652/A445, A652 and A445 were the UV absorbance of the standard sample at 652 nm and 445 nm, respectively) was the dependent variable. The limit of detection (LOD) was calculated to be 0.12 µM based on S/N = 3. Compared with other reported nitrite sensing detection methods in Table S2, this ratiometric colorimetric method has excellent advantages of low LOD and wide linear range.
Additionally, the reproducibility, accuracy and selectivity of this nitrite visual sensor was validated to demonstrate its feasibility. The relative standard deviation (RSD) of six repeated detection signal of nitrite standard sample (50.0 µM) was 1.8%, indicating good reproducibility. To verify the accuracy of constructed ratiometric colorimetric sensing system, different concentrations of nitrite (10.0, 50.0, 300.0 µM) were added to different real samples of sausage, and then quantitative detection was carried out. As shown in Table 1, the recovery rate of nitrite in all actual sample ranged from 95.11 to 102.14%, indicating qualified accuracy of this colorimetric biosensor. Considering complex composition of food, it was necessary to study the influence of potential interfere on nitrite detection and evaluate the selectivity of this method. Some potential interferes coexisting with nitrite in commercially available sausages, such as Na+, K+, Ca2+, Ba2+, Mg2+, Mn2+, Fe3+, SO42−, CO32−, PO42−, NO3− and glucose (Glu) were analyzed by this colorimetric method. The concentrations of interfering substances were set at 400.0 µM, which was the upper limit of linear range of nitrite detection. Figure 4D showed that only nitrite presented the lowest ratio absorbance while all other interfering substances presented much higher ratio absorbance under the same conditions, showing nice selectivity of this ratiometric colorimetric biosensor for nitrite detection.
Table 1
Nitrite content in real samples detected by ratiometric colorimetric method and commercial kits
Samples | Spiked concentration (µM) | Detected by this method (µM) | Recovery (%) | RSD (%, n = 3) | Detected by commercial kit (µM) | Recovery (%) | RSD (%, n = 3) |
1 | 0 | 4.14 | NA | 1.25 | 4.35 | NA | 1.83 |
10 | 13.78 | 97.44 | 3.09 | 14.75 | 102.79 | 2.36 |
50 | 53.85 | 99.45 | 1.35 | 53.88 | 99.14 | 1.05 |
300 | 301.69 | 99.19 | 2.44 | 306.91 | 100.84 | 1.57 |
2 | 0 | 2.73 | NA | 0.76 | 2.53 | NA | 0.89 |
10 | 12.92 | 101.55 | 1.32 | 12.46 | 98.64 | 0.76 |
50 | 51.46 | 97.60 | 1.83 | 53.15 | 101.18 | 2.13 |
300 | 298.14 | 98.49 | 2.89 | 311.74 | 103.04 | 1.15 |
3 | 0 | 3.70 | NA | 1.46 | 3.87 | NA | 2.39 |
10 | 13.39 | 97.74 | 1.22 | 13.53 | 97.55 | 1.28 |
50 | 54.85 | 102.14 | 3.41 | 54.56 | 101.28 | 2.64 |
300 | 292.69 | 96.38 | 3.17 | 312.92 | 102.98 | 3.55 |
4 | 0 | ND | NA | 1.34 | ND | NA | 0.96 |
10 | 9.60 | 96.01 | 1.53 | 10.16 | 101.60 | 1.67 |
50 | 49.45 | 98.89 | 2.05 | 48.76 | 97.52 | 2.02 |
300 | 285.34 | 95.11 | 3.24 | 295.51 | 98.50 | 2.58 |
NA = not applicable; ND = not detected. |
3.4 Detection of nitrite in real samples
By using this ratiometric colorimetric method, nitrite content in three types of sausages and tap water were analyzed. From the results in Table 1, nitrite concentrations in sausage 1–3 were 4.14, 2.73 and 3.70 µM, respectively. Thus, the nitrite content of sausage 1 was the highest, followed by sausage 3, and the nitrite content of sausage 2 was relatively low. These detected concentrations were all far below the limited concentration of nitrite in sausages (30.0 mg/kg, 652.1 µM) by China food and drug administration (GB 2760 − 2014). In addition, nitrite in tap water could not be detected because its concentration was lower than LOD of this method. Moreover, nitrite content in three types of sausages and tap water (Fig. S4) were also detected by commercial kits (GB 5009.33–2016 from CFDA). The obtained detection results were highly consistent with the detection result of nitrite from ratiometric colorimetric method based on Fe-HHTP@HZIF-8@AuNCs. That is to say, this method has good application value for detecting nitrite in food and water environment.