Introduction

Tetracyclines are an important class of antibiotics that consist of a broad-spectrum of members, including doxycycline, methacycline, minocycline, oxytetracycline, chlortetracycline, and tetracycline (TC) [1, 2]. They exhibit activity against a wide variety of microorganisms (e.g., gram-positive and gram-negative bacteria, atypical organisms such as rickettsiae, mycoplasmas, and chlamydiae, and protozoan parasites) [3]. They are extensively used as medicine for deterrence of bacterial contagions in humans, animals, and plants, and as growth promoters in the feeds of the poultry, animals and fishes [4]. Due to their wide usage, teracyclines residues have been found in soil, groundwater, surface water, and drinking water [5], and due to their high aqueous solubility and their long environmental half-life, the residues of tetracyclines may develop bacteria resistant strains. The resulting health problems (e.g., allergy, liver toxicity, central nervous system and gastrointestinal system, nephrotoxicity, and adverse reactions of blood system) are serious issues [6, 7], that demand more attention to conform to environmental safety requirements.

Several studies have been done to develop the detection of tetracycline antibiotic by using different methods, such as, high-performance liquid chromatography (HPLC) [8], enzyme-linked immunosorbent assay (ELISA) [9], chemiluminescence (CL) [10], microbiological methods (MM) [11], capillary electrophoresis (CE) [12], antibody-based electrochemical biosensors [13], fluorescence-based assays [14], and immunoassay (IA) [15]. These methods have many drawbacks, such as their complexity, cost, and laborious nature [16].

Plasmonic nanoparticles (NPs) such as Ag NPs and Au NPs have shown new possibilities for the development of the antibiotic detection methods because of its unique optical properties such as localized surface plasmon resonance (LSPR) [17,18,19,20]. The LSPR is a coherent oscillation of free electrons on the surface of plasmonic NPs excited by the incident wave light trapped within plasmonic NPs that are smaller than the incident light wavelength [21].

The localized electric field intensity of the LSPR of metallic NPs can be enhanced by increasing the density of hot spots, which play a major role in the surface-enhanced Raman scattering (SERS) [22]. The Au NPs aggregates more favorable choices for developing SERS because they have higher density of hot spots. The Au NPs can be synthesized on the surface of macroporous silicon (macroPSi) with a higher density of hot spot regions owing to high density of nucleation sites of the macroPSi layer [23, 24]. In this study, new route was innovated for developing high efficient nano biochemical sensor for sensing ultra-low concentration of TC antibiotics, based on the enhanced plasmonic properties of Au NPs deposited on the macroPSi substrate.

Experimental section

Chemicals

Chloroauric acid (HAuCl4, 99.99%), tetracycline hydrochloride (C22H24N2O8 HCl, 96%), and ethanol (C2H5OH, 99.98%) were purchased from Sigma-Aldrich, Germany. Hydrofluoric acid (HF, 40%) was purchased from HiMedia, India.

Preparation of macroPSi substrates

The MacroPSi substrates were obtained by a photo-electrochemical etching process at room temperature. A 100-single-polished silicon wafer with resistivity of 10 Ω cm was divided into square pieces with an area of 2.25 cm2. These pieces were cleaned for several minutes using a solution of HF and ethanol mixed in a volume ratio of (1:10) to remove the native oxide which found on the Si piece surfaces, and then washed with ethanol. The piece of Si wafer was placed in a teflon cell to form an anode electrode, and a platinum ring was immersed in a mixture HF:ethanol = 1:1 to form a cathode electrode. An area of 0.6 cm2 of the Si surface was illuminated with light from a CW diode laser of 30 mW/cm2 power density and 630-nm wavelength. The etching process was performed for 20 min with a current density of 50 mA/cm2. The macroPSi substrates were prepared with a high surface density of dangling bonds (Si-H) by controlling the etching conditions.

Deposition process of Au NPs

The dipping process of macroPSi substrates into aqueous solution of 10−3 M HAuCl4 and 3 M HF was used to prepare the aggregate of Au NPs. The dipping process was done at room temperature with a 3 min dipping time. The Au NPs were formed on the macroPSi surface through the Au ion reduction process by the Si-H dangling bonds of the surface of macroPSi as follows [25, 26]:

$$ \mathrm{Si}+6\mathrm{HF}\to {\mathrm{H}}_2\mathrm{Si}{\mathrm{F}}_6+4{\mathrm{H}}^{+}+4{\mathrm{e}}^{-} $$
(1)
$$ \kern.1em {\mathrm{Au}}^{+3}+3{\mathrm{e}}^{-}\to \mathrm{A} $$
(2)

Because the prepared macroPSi surface had a high density of dangling bonds (Si-H), the Au NPs was deposited on the surface of macroPSi in aggregated forms with a high density of hot spot regions, leading to synthesis of the Au NPS/macroPSi hot spot SERS sensor.

A standard stock solution of 0.1 mol/L TC was first prepared by dissolving 0.480 g of tetracycline dissolved in 10 mL ethanol, and then the solution was diluted with deionized water prepare TC aqueous solutions with low concentrations of 1 × 10−2, 1 × 10−3, 1 × 10−5, 1 × 10−7, and 1 × 10−9 mol/L. The solutions of TC with different pH values were obtained by adding NaOH or HCl to the 0.01 mol/L TC solution.

Characterization

The TESCAN MIRA3 field-emission scanning electron microscope (FE-SEM) was used for characterizing the morphologies of the bare macroPSi substrates and the deposited Au NPs on the macroPSi surfaces. The SERS efficiency was determined using aqueous solution of TC adsorbed on Au NPs/macroPSi hot spot SERS sensor. The sensor was immersed into these solutions at different low concentrations and various pH values for 10 h. The usual Raman and SERS spectra were measured by the APUS TESCAN Raman microscope, using a 532 nm DPSS ND: YAG (CW) laser for excitation with 10 mW laser excitation power and 3 s integration time.

Results and discussion

Morphological properties

The porosity, thickness layer, pore size, and the pore shape of the macroPSi substrate have very important roles in controlling the morphological properties of Au NPs deposited on the macroPSi surface. Thus, the optimum morphology of Au NPs can be achieved by choosing appropriate etching conditions. One can see in Fig. 1a, b, c, that the bare macroPSi substrate has been prepared with cylindrical pore shapes with sizes ranging from 0.25 to 2.25 μm and estimated porosity and thickness of 51.2% and 2.15 μm, respectively. Figure 1d, e, f, which displays the FE-SEM of the Au NPs deposited on the surface of the prepared macroPSi, shows that the Au NPs were deposited on the pore walls (Fig. 1d). The high resolution image in Fig. 1e confirms that the morphology of the Au NPs deposited on the macroPSi consisted of irregular aggregates. This irregular form yields a high density of efficient hot spot SERS regions among the Au NPs. The size of the aggregated nanoparticles ranges from 25 to 225 nm, while the gaps among them (hot spot regions) ranges from 2.5 to 12.5 nm as shown in Fig. 1g, h, respectively. The high population (percentage) of small gaps of about 2.5 nm is considered to result in efficient hot spot regions due to the fact that the enhanced local electric field originating from these regions increased as the dimensions of the gaps decreased compared with the low percentage of Au NP sizes of about 225 nm. This can be expressed mathematically in the following equation [27], where, EL represents the amplitude of the enhanced local electric field; d and D represent the sizes of hot spot regions and Au NPs, respectively.

Fig. 1
figure 1

a The FE-SEM of the bare macroPSi surface morphology, (b) cross-sectional FE-SEM image of macroPSi surface, (c) the statistical distribution of pore sizes of macroPSi substrate, (d, e) FE-SEM image of the Au NPs/macroPSi hot spot sensor at magnification of 1 μm and 200 nm, respectively, (f) cross-sectional FE-SEM image of Au NPs/ macroPSi hot spot sensor, (g) the statistical distribution of particle sizes and (h) the distance among them, and (i) EDS analysis of this sample

Full size image
$$ {E}_L={E}_i\left(\left(d+D\right)/d\right) $$
(3)

The mechanism of the effect of hot spot regions on the local electric field is shown in schematic diagram Fig. 2. Figure 1i displays the energy dispersive X-ray spectroscopy (EDS) of the Au NPs/macroPSi hot spot sensor. This figure confirms the deposited of Au on the macroPSi surface and the presence of two other elements of O and Si.

Fig. 2
figure 2

The mechanism of the effect of the hot spot regions on the local electric field

Full size image

SERS spectra of tetracycline antibiotic

The SERS spectra of the TC are related to the structure of TC. The chemical structure of tetracycline is composed of four rings of D (6a-7-8-9-10-10a-6a), C (5a-6-6a-10a-11-11a-5a), B (4a-5-5a-11a-12-12a-4a), and A (1–2–3-4-4a-12a-1), as shown in Fig. 3. The ring D is planar due to the double bonds of (C7 = C8, C9 = C10, and C6a = C10a). The sites (C9 to C12) which are located at the bottom half of the molecules are planar. The rings C and B are planar because both C6a = C10a-C11 = O and O = C11-C11a = C12 are coplanar with the ring D. The structure of ring A is nearly above the moiety plane (C9-C12) owing to the location of the C12a-OH and C4a-H sites below the plane, and C3 = C2-C1 = O is planar but not in the C9-C12 plane [28].

Fig. 3
figure 3

The chemical structure of TC [28], and the pH-dependent speciation of tetracycline (TC) [29]

Full size image

Figure 4 displays the SERS spectra of the tetracycline aqueous solution at different concentrations (10−3, 10−5, 10−7, and 10−9 mol/L) for bare macroPSi and Au NPs/macroPSi hot spot SERS sensor. It is shown that the Raman signal of bare macroPSi is very weak compared with that of the Au NPs/macroPSi hot spot SERS sensor. And, the major tetracycline aqueous solution can be observed at positions 608, 707, 861, 946, 1074, 1255, 1310, 1459, 1579, 1619, 1776, 1831, 1881, and 1995 cm−1. According to the reference [4], the peaks of the tetracycline hydrochloride powder are located at 455, 497, 605, 709, 859, 945, 1073, 1136, 1285, 1317, 1462, 1632, and 1777 cm−1. So, the peaks are at 608, 707, 861, 946, 1074, 1255, 1310, 1459, 1619, and 1776 cm−1 can be compared with that of the tetracycline hydrochloride powder, and can be selected as the characteristic peaks of SER for the detection of tetracycline aqueous solution using the Au NPs/macroPSi hot spot SERS sensor. The assignments of the SERS characteristic bands shown in Fig. 4 are listed in Table 1. It is found that the Raman signals are very strong for TC antibiotic adsorbed on the Au NPs/macroPSi hot spot sensor. The inset of Fig. 4 displays the relation between the intensity of SERS at the highest peak 1310 cm−1 and concentration of TC aqueous solution, where it can be seen that the relationship between them can be described by the following logarithmic equation: y = 30.917 ln(x) + 1046.7 with a correlation coefficient (R2) of (0.9982). The EF was determined by using the following equation for evaluating the lowest detection limit of the TC antibiotic [22], where, ISERS, IR are the signal intensities of the SERS and Raman, at the TC concentration of CSERS and CR, respectively.

$$ EF=\left({I}_{\mathrm{SERS}}\times {C}_R\right)/\left({I}_R\times {C}_{\mathrm{SERS}}\right) $$
(4)
Fig. 4
figure 4

The SERS spectra of the tetracycline aqueous solution at different concentrations of (10−3–10−9 mol/L) for bare macroPSi and Au NPs/macroPSi hot spot sensor. The inset of this figure: the relation between the SERS intensity at highest peak of 1310 cm−1 and concentration of TC aqueous solution

Full size image
Table 1 The assignments of the SERS characteristic bands of TC antibiotic [2]
Full size table

The relation between EF and the concentration of TC antibiotic is manifested in Fig. 5, the base line is the intensity of the bare PSi at wave number 1310 cm−1. The EF value depended on the contribution of TC in the aqueous solution, which varied from 4.1 × 102 to 2.0 × 108, to a maximum value at a TC concentration of 10−9 mol/L. This means that the TC concentration of 10−9 mol/L is the minimum detection limit of the TC antibiotic. The detection process of TC antibiotic is at the lowest concentration with a higher value of EF (2.0 × 108), due to the plasmonic properties of Au NPs which enhanced as the density of the hot spot SERS increased.

Fig. 5
figure 5

The relation between EF and concentration of TC antibiotic

Full size image

To study the reproducibility of SERS spectra with the Au NPs/macroPSi hot spot SERS sensor for the detection process of TC antibiotic, the SERS spectra were examined at a number of points on the Au NPs/macroPSi hot spot SERS sensor surface at the lowest concentration of 10−9 mol/L. The reproducibility of these spectra for TC antibiotic is shown in Fig. 6. From this figure and the lowest value of the relative standard deviation for the SERS intensities of 2%, it was concluded that the Au NPs/macroPSi hot spot SERS sensor has excellent reproducibility.

Fig. 6
figure 6

The reproducibility of the SERS spectra for TC antibiotic

Full size image

The effect of pH value on the TC adsorbed on Au NPs/macroPSi hot spot SERS sensor

Figure 7 presents the SERS curves of the TC aqueous solutions at 0.02 mol/L before (natural state) and after the additional process of HCl or NaOH, respectively, revealing the relation between the pH of TC solution and behavior of the SERS curves. It is clearly shown that the behaviors of the curves change after the additional process of pH, and they are varied with the pH values, except that when the pH values are equal 5 and 6, the two curves behave in the same way. Also, the Raman signal intensity decreases after additional process and with increasing of pH values. The results demonstrated that the pH value of the TC solution is one of the most significant parameters affecting the adsorption process of the TC solution on the Au NPs/macroPSi hot spot SERS sensor. This can be attributed to the finding of three ionizable groups of the TC antibiotic, including, carboxymethyl at pKa of 3.3, phenolic diketone at pKa of 7.68, and dimethylamine cation at pKa of 9.69. Thus, the changing of the pH values leads to the existence of TC in the solution as positively charged species (TCH3+) at the value of pH < 3.3 and as a zwitterion (TCH20) at 3.3 < pH value < 7.7, while it exists in the solution as negatively charged species (TCH− or TC2−) when the pH value > 7.7 [29]. In addition, when 3.3 > pH value > 7.7, the TC is unstable and can form reversible epimers [30]. Therefore, the behavior of the SERS curves at the pH values of 9 and 12 changes, and the intensity of Raman signal is very small. Because the TC solution is predominantly neutral at 3.3 < pH value < 7.7, the SERS curves at the pH values of 5 and 6 have the same behavior and a higher Raman signal intensity. And, in spite of the behavior of these two curves differing compared with that at the natural state, they have several SERS bands, and can be compared with those at the natural state, including 614, 672, 974, 1091, 1255, 1320, and 1464 cm−1. Thus, the pH values of 5 and 6 are the best for the detection process of TC antibiotic adsorbed on the Au NPs/macroPSi hot spot SERS sensor based on SERS as compared with the other values. Wali et al. [26] reported that the study of the pH influence on the SERS intensity can be satisfied by calculating the EF of SERS intensity for analyte under acidic and alkaline conditions, which is denoted as EFpH by using the following equation:

$$ {EF}_{\mathrm{pH}}={I}_{\mathrm{SERS}}/{I}_R $$
(5)
Fig. 7
figure 7

The SERS curves of the TC aqueous solutions at 0.02 mol/L before (natural state) and after additional process of HCl or NaOH. The inset of figure display the natural state

Full size image

Figure 8 shows the various EFpH and pH values, it was found that the relationship between the EFpH and the pH values is defined by the polynomial equation y = 1.2643x2 − 27.233x + 148.26 with R2 of 0.9997, so the highest and lowest EFpH were obtained at the pH values of 5 and 12, respectively.

Fig. 8
figure 8

The EFpH versus pH values

Full size image

Conclusions

In summary, an efficient and highly rapid detection process has been innovated for detecting the residual TC antibiotic in water. The Au NPs/macroPSi hot spot sensor was developed through deposition process of Au NPs on the macroPSi surface with a high density of hot spot regions. The results elucidated that the Au NPs/macroPSi hot spot SERS sensor successfully provided chemical information for diagnosing the TC antibiotic with a high EF and reproducibility at ultra-low concentration of 10−9 mol/L. It was observed that the pH values of 5 and 6 are the best for the detection process of TC antibiotic adsorbed on the Au NPs/macroPSi hot spot SERS sensor based on SERS as compared with the other pH values of 9 and 12.The results proved that the Au NPs/macroPSi hot spot sensor is an excellent detection tool that can be used for qualitative and quantitative detection of TC residues in wastewater, and then for safety control.