Introduction

17β-Estradiol (E2), as a typical endogenous estrogen, can not only affect people and animal sex differentiation, and normal metabolic function, but also disturb the normal growth, development, and reproduction of the organism [1]. Fatally, E2 in the environment can be harmful to organisms even at very low concentrations (ng L−1) [2]. Hence, it is critically significant to establish an effective, rapid assay for E2 in the environmental and biological analysis field.

Currently, numerous endeavors have focused on the development of effective E2 assay methods, including spectrometry [3], colorimetry [4], immunoassay [5], fluorescence [6], and electrochemistry (EC) [7]. Although these analytical methods have achieved satisfied results, they still present several obstacles in practical applications, such as expensive device, time-consuming process, technical complexity, or limited sensitivity [8]. Therefore, it still remains a grand challenge to develop a facile, sensitive, and accurate analytical method for E2 assay.

As an emerging technology, photoelectrochemical (PEC) sensors have aroused intense interest among researchers due to its easy miniaturization, low background signal, and fast response [9,10,11,12,13]. Nevertheless, low sensitivity and selectivity of PEC sensors have hindered its further application in environmental analysis [14,15,16,17]. Design of efficient photoelectroactive materials is a pivotal issue in the construction of PEC sensor with excellent sensitivity. Cadmium sulfide (CdS), as significant semiconductor, has commonly used in photocatalysis, PEC sensors, and photodegradation, and demonstrated outstanding photocatalytic performance under visible light irradiation [18,19,20]. Unfortunately, CdS-based photoelectroactive materials suffer from several inherent disadvantages of easy aggregation and photocorrosion, which substantially hamper their extensive application [20]. Integration carbon layer and CdS is one of the most effective surface modification method to solve the above drawbacks of CdS. Introduction of carbon layer can stabilize the surface of CdS and suppress its photocorrosion behavior. Importantly, the photoelectrochemical properties of CdS after modification of coated carbon have been markedly improved and served as potential photoelectric materials. For instance, Hu et al. synthesized carbon-coated CdS with superior photostability, revealing remarkable photocatalytic activity during pollutant degradation process [19].

Biological recognition process is also of great importance for the PEC biosensing system to acquire exceptional selectivity. Aptamer can be commonly regarded as a new molecular identification element to selectively recognize various targets due to its high affinity and specificity [21, 22]. Furthermore, aptamers have demonstrated to be more promising in sensors compared with antibodies, owing to their stability, easy synthesis, facile modification, and relatively low cost [21]. Combined with the readily chemical modification characteristic of aptamer, devising signal amplification strategies is another critical to improve the sensitivity of PEC sensors. Among diverse signal amplification strategies, enzyme-mediate signal amplification gained broad attention due to its advantages of exceptional efficiency, and exquisite selectivity. For example, Ai’s group reported a novel PEC biosensor with high sensitivity for protein kinase A activity detection based on signal amplification of alkaline phosphatase (ALP) [23].

In this work, an ultrasensitive “signal-off” PEC aptasensor for assay of E2 was constructed based on carbon-coated cadmium sulfide rose-like nanostructures (CdS@C NRs), assisted with ALP-mediated signal amplification strategy. As depicted in Scheme 1, the single-stranded DNA was firstly modified on the electrode surface, while the ALP-modified E2 aptamer (ALP-apta) was bonded to the ssDNA via base complementary pairing. In the absence of E2, the enhanced photocurrent of CdS@C NRs was achieved due to the in situ generation of electron donor (ascorbic acid, AA), which was triggered by ALP-mediated hydrolysis reaction. With the addition of E2, ALP-apta were released from the electrode surface after the specific recognition with E2, causing the negative effect on PEC response. Accordingly, this protocol indicated the “signal-off” photocurrent response with increases in concentrations of E2. The developed PEC aptasensor reveals high selectivity and excellent sensitivity for E2 assay in the real samples.

Scheme 1
scheme 1

Schematic diagram of sensitive PEC biosensor for assay of E2

Full size image

Experimental

Materials and instruments

Experimental materials and characterization instruments are provided in detail in Electronic Supplementary Material (ESM).

Synthesis of CdS@C NRs

CdS@C NRs was prepared using a facile solvothermal process. Typically, 0.6851 g of CdCl2·2.5H2O, 0.4476 g of D-penicillamine, 0.4280 g of glucose, and 0.3330 g of polyvinylpyrrolidone (PVP; MW ~ 58 K) were sequentially dissolved in 30 mL of ethylene glycol (EG) under sonication condition. The resulting homogeneous solution was then transferred into a 50-mL Teflon-lined stainless-steel autoclave and kept for 12 h at 160 °C. After cooling down to room temperature, the samples were collected by centrifugation, and sufficiently washed with anhydrous ethanol and distilled water, subsequently dried at 60 °C for further use.

Construction of the aptasensor

For the construction of PEC aptasensor, 20 μL of 2.0 mg mL−1 CdS@C NRs was modified on indium tin oxide (ITO) surface (diameter = 5.0 mm). To ensure that the -NH2-modified ssDNA could be linked on the surface of the above electrode, chitosan (CS) and glutaraldehyde (GLD) were progressively coated on the electrode. Twenty microliters of 0.02 wt% chitosan (CS) solution was dropped and then dried at 40 °C for 1 h. Twenty microliters of 5.0 wt% glutaraldehyde (GLD) solution was introduced and incubated for 1 h in the dark, followed by incubation with NH2 modified ssDNA (1.0 μM, 20 μL) at 4 °C overnight. The electrode was subsequently cultured with 10 μL BSA solution (1.0 wt%) at 37 °C for 1 h to block non-specific binding sites. Afterwards, 20 μL of ALP-modified aptamer (2.0 μM) was added to the above electrode and reacted with the ssDNA at 37 °C for 2 h to form double-strand DNA by hybridization. Notably, during the preparing progress, the electrode was extensively rinsed with phosphate buffer solution (PBS, pH 7.4, 10 mM) after each modification step to remove the residual reagent. At last, the prepared electrode was stored at 4 °C for further use.

PEC assay

For E2 analysis, different concentrations of E2 (20 μL) were introduced and reacted with the developed aptasensor at 37 °C for 40 min. Then, the aptasensor was immersed into 2.5 mM L-ascorbic acid 2-phosphate solution (AAP, 4 mL) for 1 h at 37 °C. Eventually, the PEC measurements were performed on the PEC measurement system, which consist of a CHI 660A electrochemical workstation and a PEC light source system at a constant bias of 0 V, by using PBS (pH 8.0) as the electrolyte.

Results and discussion

Characterization of as-synthesized samples

In this study, CdS rose-like nanostructures coated with a thin carbon layer (CdS@C NRs) were prepared using a facile solvothermal method and served as the platform of the aptasensor. The scanning electron microscope (SEM) images show the rose-like nanostructure of the samples with uniform size of about 586 nm (Fig. 1A–B). These rose-like nanostructures are mainly composed of multiple petal-like units with polygonal shape, which afford the large surface area for the immobilization of biomolecules. As shown in Fig. 1C, rose-like nanostructure assembled by irregular polygonal shape is further elucidated by transmission electron microscopy (TEM), which is consistent with the SEM results. As observed from high-resolution TEM (HR-TEM), the calculated lattice spacing of different regions are 0.322 nm and 0.362 nm, corresponding to the (101) and (100) crystal planes of CdS particle [24], respectively. Additionally, an amorphous carbon layer with thickness of 4.77 nm is observed outside of CdS surface (Fig. 1D). The polycrystalline structure of the samples is demonstrated by the selected area electron diffraction (SAED, Fig. 1E).

Fig. 1
figure 1

A–B SEM images. C-D Low- and high-resolution TEM images. E SAED image. F XRD pattern of CdS@C NRs. Inset in A shows the particle-size distribution

Full size image

The crystal structure of CdS@C NRs was identified by X-ray diffraction (XRD) analysis. As depicted in Fig. 1F, there are several obvious characteristic diffraction peaks at 24.8°, 26.5°, 28.2°, 36.6°, 43.7°, 47.8°, and 51.8°, corresponding to (100), (002), (101), (102), (110), (103), and (112) planes of the CdS (JCPDS No. 41–1049), respectively [24]. No peak corresponding to the carbon layer is observed for the as-synthesized carbon coated samples, which may be due to the existence of ultrathin or amorphous carbon layer, and the possible peaks overlap for carbon materials and CdS at 26° [18].

X-ray photoelectron spectroscopy (XPS) was used to study the chemical state and chemical composition of CdS@C NRs. Survey XPS spectrum of the sample confirmed the existence of Cd, S, C, and O elements (Fig. 2A). The high-resolution Cd 3d XPS spectrum shows two obvious characteristic peaks at 411.4 eV and 404.7 eV attributing to Cd 3d3/2 and Cd 3d5/2, respectively, which prove the presence of Cd2+ species (Fig. 2B) [25]. There were two characteristic peaks at 161.0 eV and 162.3 eV of S 2p assigning to S 2p3/2 and S 2p1/2, indicating the existence of S2– species (Fig. 2C) [26]. As seen in Fig. 2D, the C 1 s peak can also be divided into two separate peaks. The C 1 s peak at 284.7 eV and 287.5 eV belong to sp2-carbon material (C = C) and oxygen-containing carbon material (C = O), respectively [27]. These XPS results further attest the co-existence of carbon layer and CdS in the prepared samples.

Fig. 2
figure 2

A Survey XPS spectrum. B–D High-resolution Cd 3d, S 2p, and C 1 s XPS spectra. E UV–vis diffuse reflectance spectra. F The band gap energy of CdS@C NRs

Full size image

UV–vis diffuse reflection spectrum (DRS) analysis was performed to investigate the optical properties of CdS@C NRs (Fig. 2E). The CdS@C NRs display a wide UV–vis absorption in the range of 300 ~ 550 nm. The band gap of CdS@C NRs can be calculated to be 2.07 eV based on the UV–vis DRS (Fig. 2F) [28], which is narrower in contrast with pure CdS nanoparticles (2.4 eV) due to the modification of carbon layer [20].

The possible electron transfer mechanism was proposed on the basis the above results, shown in Scheme 2. Under illumination, amorphous surface carbon of CdS@C NRs can absorb visible light, and then the excited electrons of CdS transfer from valence band to conduction band, producing electron–hole (e−/h+) pairs. More importantly, the coated carbon species of CdS@C NRs can effectively promote the electron transfer and inhibit recombination of photoexcited e−/h+ pairs [29]. Simultaneously, in situ generated AA can be considered as effective electron donor to capture the photo-generated h+ for preventing recombination of e−/h+ pairs, obtaining enhanced and high stable photocurrent response.

Scheme 2
scheme 2

Possible charge-transfer mechanism in CdS@C NRs under light illumination (CB, conduction band; VB, valence band)

Full size image

Optimization of experimental conditions

To achieve excellent analytical performance of PEC aptamer sensor, several main experimental parameters were optimized (Fig. S1, ESM). The experimental results show that the optimal CdS@C NR-loaded concentration is 2.0 mg mL–1; the optimal E2 incubation time is 40 min; and the optimal pH of PBS is 8.0. Then, the above parameters were chosen for the subsequent experiments. The specific experimental results are provided in ESM.

Analytical performance of the developed PEC aptasensor

To demonstrate the feasibility, the construction process of the aptasensor was studied through the PEC response and electrochemical impedance spectroscopy (EIS). As shown in Fig. 3A, an obvious photocurrent response can be observed after the introduction of CdS@C NRs (curve a). When CS (curve b), ssDNA (curve c), BSA (curve d), and ALP-apta (curve e) are successively modified on the electrode surface, the photocurrent will decrease gradually owing to the introduction of non-conductive substances, such as aptamers, proteins, and polymer. With the addition of analyte E2 (curve f), the photocurrent could return to a certain extent, which may be caused by the release of the formed E2-apta conjugate from the electrode surface.

Fig. 3
figure 3

A–B Photocurrent response, and the corresponding EIS spectra of the CdS@C NRs/ITO (curve a), CS/CdS@C NRs/ITO (curve b), ssDNA/CS/ CdS@C NRs/ITO (curve c), BSA/ssDNA/CS/CdS@C NRs/ITO (curve d), ALP-apta/BSA/ssDNA/CS/CdS@C NRs/ITO (curve e), and E2/ALP-apta/BSA/ssDNA/CS/CdS@C NRs/ITO (curve f). C Photocurrents of the aptasensor with various concentrations of E2 from 0 to 250 nM. (D) The corresponding calibration plots. E Specificity of the developed aptasensor against 100 nM E2 (f), 100-fold bisphenol A (a), nonyl phenol (b), ethinylestradiol (c), atrazine (d), ciprofloxacin (e), and a mixture (containing all the analytes, g). F Stability of the proposed aptasensor, CE2 = 100 nM (error bar = standard deviation, n = 3)

Full size image

EIS was further carried out to characterize the electrochemical properties of the aptasensor during stepwise reactions (Fig. 3B), which were measured at the applied potential of 0.241 V in the frequency range from 0.1 to 100 kHz in 0.1 M KCl solution containing 5.0 mM [Fe(CN)6]3–/4–. With the immobilization of CS (curve b), ssDNA (curve c), BSA (curve d), and ALP-apta (curve e) on the CdS@C NRs/ITO (curve a) surface, the charge transfer resistance (Rct) gradually increases causing by the non-conductivity of introduced biomolecules. Following the addition of E2, the Rct declines (curve f), because a E2-ALP-apta conjugate is formed and dislodged from the electrode surface. In addition, the gel electrophoresis experiment was carried to further verify the construction principle of our proposed aptasensor (Fig. S2, ESM). These above results demonstrate the successful construction of the aptasensor.

To evaluate the analytical performance, the proposed PEC aptasensor was explored against diverse concentrations of E2 under the optimized conditions. Figure 3C displays that the photocurrents gradually decrease with the E2 concentrations from 0 to 250 nM, because of the decrease of in situ produced AA by the ALP enzymatic catalysis. Meanwhile, a good linear relationship was achieved between the photocurrent change value ΔI (ΔI = I—I0) and the concentration of E2 from 1.0 to 250 nM with the detection limit of 0.37 nM (Fig. 3D), and the linear regression equation is ΔI (μA) =  − 0.00386 CE2—0.1932, with a correlation coefficient (R2) of 0.9985. The obtained results exhibited comparable or even better analytical performance for E2 determination in contrast with the obvious reported work (Table S1, ESM) [6, 30,31,32,33,34,35].

Specificity and stability of the fabricated sensor

To investigate the selectivity of the PEC aptasensor, the PEC signals were recorded against the 100 nM E2, and several 10 μM interfering substances, including bisphenol A, nonyl phenol, ethinylestradiol, atrazine, ciprofloxacin, and mixture (Fig. 3E). The concentration of interferents is 100-fold than that of the target (E2), so the non-specific is inevitable. And the chemical structure of ethinylestradiol and ciprofloxacin among these interferents are more similar with E2; hence, the interference signal is relatively large. However, the photoelectrochemical response signal of the mixture (include 100 nM analyte and 10 μM all the interferents, column f in Fig. 3E) is almost identical in comparison with 100 nM pure analyte (column g in Fig. 3E), which further demonstrates the high selectivity of the prepared aptasensor for E2 detection. Then, the photocurrent at 100 nM E2 was detected to assess the stability of aptasensor within 210 s (Fig. 3F). There is no obvious photocurrent variation with the relative standard deviation (RSD) of 4.7% during the testing process, indicating the excellent stability of the constructed aptasensor.

Precision was vital for the proposed PEC aptasensor, which can be evaluated via its reproducibility. The reproducibility was measured by five independent aptasensor at different E2 concentrations (50 nM, 100 nM, and 150 nM E2) with the RSD of 0.91%, 0.92%, and 1.63%, respectively. The result confirms the acceptable repeatability of the PEC aptasensor.

Real sample assay

Utility and reliability of the PEC aptasensor was conducted through standard addition method in the real samples’ analysis. A total of 50 nM, 100 nM, and 150 nM E2 were separately added to the tap water and human serum samples to calculate the recoveries of the spiked amount of E2. The tap water and human serum samples were respectively obtained from laboratory water pipes and Jinhua Municipal Central Hospital Medical Group. Table 1 exhibits the recoveries of E2 analysis in the real samples. The recoveries are in the range of 99.10 ~ 102.43% with the RSDs of 0.42 ~ 2.01% (n = 3), revealing the feasibility and potential utility of the developed aptasensor in environmental samples analysis.

Table 1 Detection of E2 in tap water at different concentrations (n = 3)
Full size table

Conclusions

In this work, an ultrasensitive signal-off PEC aptasensor for the E2 detection was developed by integrating the prepared CdS@C NRs as photoactive substance and enzyme-mediated signal amplification strategy. The developed PEC aptasensor has a relatively wide linear range of 1.0 ~ 250 nM and low LOD of 0.37 nM. Although the present work has achieved satisfactory results and expanded the possibilities of sensitive analysis of other environmental pollutants, the development of photoinductive sensors still faces some challenges, such as the design of ideal photoactive materials, lower detection limits, and higher specificity.