Introduction

The monitoring of toxic metal ions in aquatic ecosystems is an important issue because of its adverse effect on human health and the environment [1]. Mercury (II) ion (Hg(II)), one of the most potently toxic metal ions, can cause severe irreversible harm to the human body mainly via the central nervous system, digestive system, and internal organs [2]. Furthermore, the Hg(II), arised from a variety of sources (such as oceanic, volcanic emissions, gold mining, the combustion of solid waste and fuels) is widely distributed in atmosphere, water, and soil. And this prompted the development of new methods for Hg(II) determination with great simplicity, high sensitivity and good selectivity [3]. Up to now, numerous methods have been developed for Hg2+ determination, such as inductively coupled plasma mass spectrometry [4], atomic absorption spectroscopy [5], atomic fluorescence spectroscopy [6], and electrochemical methods [7]. However, most of them require expensive and sophisticated instrumentation, skilled operator, and complex sample pre-processing, and they cannot be used for on-site determination of Hg(II) in the environment.

Aptamers, first introduced by three groups independently in 1990 are the artificial single-stranded DNA or RNA sequences, that can recognize target molecule (small molecules, proteins, and even entire cells) with extremely high specificity [8–10]. Hg(II) can specifically interact with thymine-rich DNA sequences (called Hg(II) aptamers) to form thymine–Hg(II)–thymine (T–Hg(II)–T) complexes [11]. The use of aptamer opens a new field for highly sensitive and simple determination of Hg(II) due to its inherent advantages of high stability, easy storage, easy modification, and high affinity. Nowadays, many biosensors combined with surface plasmon resonance (SPR) spectroscopy [12], colorimetric assays [13], fluorescence detection [14], and electrochemical transducers [15] have been developed for Hg(II) determination. However, the limit of detection (LOD) and analysis time are often cited as problems in the Hg(II) determination. Therefore, chemiluminescence (CL) is considered as a promising alternative to resolve the above-mentioned problems, because of its high sensitivity, wide linear range, simple instrumentation, and no background scattering light interference.

We describe here, a novel CL aptasensor is designed to realize the simple and on-site determination of Hg(II) by incorporating the aptamer recognition elements onto the recently developed microfluidic paper-based analytical devices. So our method can combine the advantages of paper-based tests and conventional lab-on-chip devices [16, 17]. Paper, a three-dimensional cellulose fiber web with high surface area, is an attractive substrate for microfluidic devices. This can be due to its various unique advantages: cheap, abundant, easy to use and disposable, convenient to store, transport, modify, and able to move fluids by capillary action without an external power source. Hydrophilic channels are patterned on paper by hydrophobic walls of photoresist/polymer, inks, wax, and plasma treatment, laser treatment or by cutting method [18–21]. N,N’-carbonyldiimidazole (CDI), containing two acylimidazole leaving groups, was used as a novel simple crosslinker strategy to covalently immobilize the aptamer on the paper test strip without any spacer molecules. CDI can react with hydroxyl groups to create a reactive intermediate-imidazolyl carbamate. Then the reactive intermediate can be coupled to amine-containing molecules with the result of a one-carbon spacer, forming stable urethane (N-alkyl carbamate) linkages. Thus, CDI was employed as viable alternatives to traditional crosslinker (such as periodate and chitosan) in our previous work [22] due to its simple fabrication procedures. The covalent coupling of aptamers on a paper device display high binding stability for regeneration and reusability for the further development of low-cost application. Therefore, a novel CL system based on the CDI modified paper and phenylene-ethynylene derivative (4,4′-(2,5-dimethoxy-1,4-phenylene)bis (ethyne-2,1-diyl) dibenzoic acid) (P-acid)-potassium permanganate (KMnO4) system was designed. And the figure with the chemical formula of P-acid was added in Fig. 1. We describe here, the P-acid was selected as the CL signal labeling for the first time.

Fig. 1
figure 1

Schematic representation of the fabrication process of paper-based CL device. a paper working zone; b the paper was modified with CDI and aptamer; c after immobilized with NPS@P-acid and capturing with Hg2+; d the picture with chemical formula of P-acid

Full size image

Recently, nanoporous silver (NPS) with controllable three-dimensional (3D) structure, high surface-to-volume ratio, good stability, and nice biocompatibility has roused great interest [23, 24]. Due to its advantages, NPS was used to conjugate with P-acid (formed NPS@P-acid composite) for signal amplification strategy. In the presence of Hg(II), the aptamer probe folds into T–Hg(II)–T mediated hairpin structure, and the NPS@P-acid labeled on aptamer can provide a readout signal for quantitative determination of Hg(II). The system using NPS@P-acid composites exhibit 8-fold enhancement of CL intensity as compared to that based on pure P-acid, revealing a more sensitive sensor for Hg(II) determination. Herein, the LOD as low as 0.01 nM for Hg(II) was obtained under optimal experimental conditions. Furthermore, the system was further applied to the determination of Hg(II) in river water samples with satisfactory results. This paper-based CL aptasensor will be a promising candidate for on-the-spot detection of Hg(II) in real environmental samples.

Materials and methods

Reagents and materials

Tween-20, N,N’-carbonyldiimidazole (CDI), N-(3-dimethylaminopropyl)-N’- ethylcarbodiimidehydro-chloride (EDC), NaOH, ethylenediamine (EDA) were obtained from Sigma (St. Louis, MO, USA, http://www.sigmaaldrich.com). KMnO4, ethanol and ethanolamine were obtained from Alfa Aesar China Ltd (http://alfa.com/china). Ag23Al77(at. %) alloy precursors were fabricated by melt-spinning, which generated thin foils with thickness of about 50 μm. Whatman chromatography paper 1 (200.0 mm × 200.0 mm) (pure cellulose paper) was purchased from GE Healthcare World-wide (Pudong Shanghai, China, http://www3.gehealthcare.com/en/Global_Gateway). This type of Whatman paper was chosen because of its uniform composition (relative to other types of paper) and lack of additives that affect flow rate and CL reaction. The DNA oligonucleotide sequences shown below were purchased from Shanghai Linc-Bio Science Co. LTD (Shanghai, China, http://lincbio.cn.china.cn/).

(S1) 5′-NH2-(CH2)6-CAGTTTGGAC-3′

(S2) 5′-NH2-GTCCTTTCTG-3′

0.05 % Tween-20 was spiked into 0.1 M pH 7.4 PHOSPHATE BUFFER as washing buffer. Ultrapure water obtained from a Millipore water purification system (≥18 MΩ⋅cm, Milli-Q, Millipore) was used in all assays and solutions.

The scanning electron microscopy (SEM) images were obtained from a QUANTA FEG 250 thermal field emission SEM (FEI Co., USA, http://www.fei.com/). CL emission was measured using a computerized ultraweak luminescence analyzer (Type RFL-200, manufactured at Xi’an Remex Analysis Instrument Co, Ltd, Xi’an, China, http://ccn.mofcom.gov.cn/1096057). The FT-IR spectra were performed on Thermo Scientific Nicolet 380 FT-IR spectrometer (Shanghai Keduo Co, Ltd, Shanghai, http://194127.ccen.net/). UV–Vis experiments were performed with a UV-3900 spectrophotometer (Hitachi, Japan, http://www.hitachi.com/). The contact angle tests were performed on contact angle measurement (Model OCA40, Dataphysics, http://www.dataphysics.com/zh).

Preparation of NPS@P-acid conjugated aptamer probe

According to the reported method [23], the NPS was prepared by dealloying the Ag23Al77 alloy foils in 1 M NaOH solutions for 72 h at 30 °C, followed by rinsing thoroughly with pure water. P-acid, first demonstrated the CL behavior with KMnO4, was employed here and it was prepared according to the published literature [25].

To obtain the NPS@P-acid composites, 1 mg NPS nanoparticles and 1 mL EDA were dispersed in 3 mL ethanol by sonication for 10 min. After washing with 3 mL phosphate buffer (pH 7.4) and being centrifugal separated, the amino group functionalized NPS nanoparticles were added into 3 mL anhydrous solution of P-acid (4 mM) containing 5 mM EDC by sonication for 10 min. EDC, as the linking agent, can react with carboxy group to form an active acylimidazole group capable of coupling with amine-containing molecules [26]. After rinsing with ultrapure water thoroughly, the formed NPS@P-acid composites were immediately redispersed into 200 μL of 5 μM thiolated DNA aptamer (S2) solution. Then the mix was sonicated, and incubated for 30 min at room temperature to form the NPS@P-acid conjugated aptamer probe (S2). It has been proven that amino group and thiol group can be bound strongly to Ag [27]. The resulting aptamer probe were washed with phosphate buffer (pH 7.4) and then dispersed in phosphate buffer and stored at 4 °C before use.

Fabrication of the paper-based CL aptasensor

Fabrication details of this chemiluminescence (CL) paper test strip can be found in the supplemental information. As shown in Fig. 1, the fabrication process of CL aptasensor was constructed as follows: First, the paper zone was activated by adding 5 μL 0.3 M CDI for 5 min, and then washed sequentially with anhydrous ethanol, ice-cold water and phosphate buffer (pH 7.4) to remove excess CDI and reaction by-products. Then 10 μL of 5 μM aptamer probe (S1) solution was added on the paper zone, and reacted at room temperature for 10 min. Subsequently, excess aptamer probe (S1) was washed out with phosphate buffer containing 0.05 v/v Tween-20. And 5 μL 1 M ethanolamine was added and incubated for 5 min at room temperature to quench the remaining active groups. After washing with phosphate buffer containing 0.05 v/v Tween-20, the resulting CL aptasensor was obtained and stored at 4 °C before use.

Determination of Hg2+

The procedures of Hg2+ determination on paper-based CL aptasensor were similar to our previous work [28], and a detailed procedure is described below. The determination of Hg2+ was performed by first adding 5 μL of Hg2+ solution with different concentration to the paper zone, then added 10 μL NPS@P-acid conjugated aptamer probe (S2), incubating for 15 min. After washing with phosphate buffer (pH 7.4), CL experiments were carried out through sequential injection of the p-acid-KMnO4 system by a pipette according to the protocol reported in our previous work [22]. The CL signals related to the Hg2+ concentrations can be measured.

Results and discussion

Characterization of NPS, NPS@P-acid

The morphology of the NPS and the NPS@P-acid composites were characterized by the SEM images. As shown in Fig. 2a, it was observed that the ligament and nanopore channels were uniformly distributed across the entire sample. And the average size of the smooth silver ligament was about 30 nm. The 3D nanoporous material provided a flexible substrate for further functionalization. In addition, a slightly coarsened ligament size of 50 nm in Fig. 2b demonstrated the fact that P-acid can be immobilized on the NPS through ethanediamine cross-linking.

Fig. 2
figure 2

Representative SEM images of a NPS; b NPS@P-acid composites

Full size image

Characterization of NPS@P-acid conjugated aptamer probe

Figure 3a illustrates the formation of the NPS@P-acid composites and the NPS@P-acid conjugated aptamer probe. It can be seen from Fig. 3a that there were two obvious absorption peaks of P-acid, while no absorption peak was observed for the NPS. After NPS conjugated with P-acid, two distinct adsorption peaks were observed on the UV–Vis spectra of NPS@P-acid composites, which attributed to the adsorption peaks of P-acid. In addition, the adsorption peak from DNA aptamer was about at 260 nm, and the obtained NPS@P-acid conjugated aptamer probe showed another obvious adsorption peak. And this confirmed the formation of the NPS@P-acid conjugated aptamer probe.

Fig. 3
figure 3

a UV–Vis absorption of NPS@P-acid conjugated aptamer; b FTIRS of (curve A) original paper; (curve B) NPS@P-acid composites; (curve C) NPS@P-acid conjugated aptamer

Full size image

Characterization of the paper-based aptasensor device

Fourier Transform Infrared Spectroscopy (FTIRS) shown in Fig. 3b was used to confirm the immobilization of the NPS@P-acid conjugated aptamer probe on the paper zone. The absorption in curve B at 1,760 and 1,660 cm−1 can be assigned to the imidazole ester νC=O, and the C = C and C = N stretching modes of the imidazole heterocycle of CDI. Due to the C = O stretching and NH deformation, the peaks at 1,740 and 1,530 cm−1 were observed in curve C with the disappearance of the absorptions at 1,760 and 1,660 cm−1. So we can conclude the successful immobilization of the amino functionalized NPS@P-acid conjugated aptamer probe on CDI activated paper zone.

A pure cellulose paper was used as the substrate for the paper-based CL aptasensor device, and the morphology of the porous structures of paper electrodes were shown by SEM images in Fig. 4. The porous structure and microfibres of pure cellulose paper zone were shown in Fig. 4a and b. Due to the 3D porous structure of paper, when the wax screen-printed paper were baked, the melted wax would permeate into the paper, and hydrophobic walls were then fabricated (Fig. 4c and d). After CDI was modified on the wax-patterned paper, the surface of the paper was smoother than the wax penetrated one and the fiber was thicker than the pure cellulose paper (Fig. 4e and f). The contact angle on the front and back side wax-penetrated surface is 126° and 118°, respectively. Wax-screen-printing did not leave any contamination on hydrophilic paper zones. And after the curing process, the unprinted area maintained good hydrophilicity, flexibility, 3D porous structure and their original property as well. As shown in Fig. 4g and h, the surface of the paper was coated with some granular NPS after reaction, and 3D porous structure of the paper was not changed.

Fig. 4
figure 4

Representative SEM image of a, b pure cellulose paper; c, b wax-printed paper; e, f CDI modified wax-penetrated paper

Full size image

Kinetic characteristics of CL response on paper-based aptasensor device

Compared to the P-acid labeled aptamer probe, the CL response of the as-prepared paper-based NPS@P-acid tagged aptasensor was investigated to confirm the feasibility of CL reaction on wax-printed paper zone. As shown in Fig. 5a, the pure aptamer exhibited weak CL response (curve 1), and the signal intensity of the NPS@P-acid (curve 3) was greatly increased compared with the P-acid (curve 2) under the same conditions, which indicated the enhanced sensitivity. The improved CL performance for our method can be due to a higher number of P-acid molecules loaded on the NPS to label the aptamer probe to achieve the signal amplification. In addition, the obtained signal increased with the increasing concentration of the analytes. Therefore, our paper-based aptasensor exhibited a good analytical performance for the Hg(II) determination and can be used in real sample.

Fig. 5
figure 5

a Kinetic characteristics of CL response on (1) pure aptamer; (2) P-acid; (3) NPS@P-acid composites under the same conditions; b Specificity with divalent metal ions (such as Pb2+, Ca2+, Mg2+, Pd2+, Cu2+, Co2+, Zn2+, Ni2+, Fe2+, Ag+, Au3+). Error bars represent the standard deviations of measurements taken from at least five independent experiments; c The calibration curves of the paper-based CL aptasensor for Hg2+ determination in a low concentrations from 0.02 to 1 nM; d The calibration curves of the paper-based CL aptasensor for Hg2+ determination in the wide range from 0.02 to 500 nM. Error bars showed the standard deviations of six measurements

Full size image

The effect on CL sensor performance with different size of NPS

According the reported method [23], the NPS with different pore diameter can be prepared by dealloying the Ag23Al77 alloy foils in 1 M NaOH solutions for different time. To obtain optimum pore diameter, the dealloying time varied from 36 to 96 h, and the different CL sensor performance were shown in Figure S2A. From Figure S2, when the dealloying time varied from 36 to 96 h, the optimum CL performance was obtained at the dealloying time of 72 h. The pore diameter distribution for NPS at the dealloying time of 72 h was obtained according to the reported method [29], and the result with most pore diameter of 50 nm was shown in Figure S2B.

Selectivity, stability and reproducibility of the paper-based aptasensor device

Furthermore, the selectivity of the present paper-based aptasensor was tested for 10 nM Hg(II) over other metal ions such as Pb2+, Ca2+, Mg2+, Pd2+, Cu2+, Co2+, Zn2+, Ni2+, Fe2+, Ag+ and Au3+, at 4 mM concentration under the optimum conditions, respectively. As shown in Fig. 5b, the resulting paper-based aptasensor exhibited remarkable CL intensity only in the presence of Hg(II) rather than other metal ions, indicating good selectivity for the Hg(II) detection. Likewise, the device’s CL response to Hg(II) was unaffected by the presence of mixture of metal ions.

When this paper-based aptasensor device was stored at 4 °C in a dry environment for over 1 month and it was used to measure the same Hg(II) concentration. The CL intensity of the paper-based aptasensor device was decreased to 91 % of its initial response, indicating the acceptable stability and suitability for storage or long-distance transport to remote regions and developing countries.

Reproducibility is also an extremely important feature for the paper-based aptasensor. The reproducibility was investigated by repetitive measuring of 10 nM Hg(II) in 11 individual strategies. And the relative standard deviation (RSD) of the reproducible CL response was 6.2 %, showing the good reproducibility of the protocol.

Performance of the paper-based aptasensor for Hg2+ determination

To evaluate the sensitivity of our method, the CL intensity of the novel paper-based aptasensor device was measured with various concentrations of Hg2+ ions. And the CL intensity was very sensitive to the change of the Hg2+ concentration. The enhanced CL intensity was linear in the low Hg2+ concentrations from 0.02 to 1 nM, and the equation of the calibration curve was I CL = 4,525.56 + 10,504.88 [Hg2+] ([Hg2+]/nM)(R = 0.9895). And CL signal was logarithmically proportional to the Hg2+ concentration in the wide range from 0.02 to 500 nM, and the equation of the calibration curve was I CL = 13,063.72 + 6,348.75 lg[Hg2+] ([Hg2+]/nM)(R = 0.9985). In addition, the detection limit for Hg2+ was 0.01 nM at a signal-to-noise ratio of 3, indicating that the method can be used for the ultrasensitive determination of Hg2+. The characteristics of various method for the determination of Hg2+ are summarized in Table 1. Compared with the other methods, the method has a wide linear range and a lower detection limit.

Table 1 Comparison of major characteristics of different methods
Full size table

Application of the paper-based aptasensor in real samples

To further explore the practical applicability of the paper-based aptasensor device, the Hg(II) detection was conducted in water samples from the Yellow River, a realistically complex sample containing a variety of interferences. After filtering with 0.2 mm membrane to remove the insoluble substances, the samples were spiked with target Hg(II) over the range of 0.0–10 nM. As shown in Table 2, the recovery was between 96.9 and 102.4 %, which indicated that the system allowed the accurate quantification of Hg(II) in real samples without any interference from other potentially coexisting metal ions.

Table 2 Application of Hg2+ in real samples
Full size table

Conclusion

In summary, an ultrasensitive paper-based CL aptasensor was developed for Hg(II) determination based on the simplicity and low-cost of paper test strip and the selectivity of the aptamer. This assay allows us to determine Hg(II) down to 0.1 nM. The main advantages are as follows: First, the aptamer was covalently immobilized on the paper zone through CDI covalent crosslinking with simple fabrication procedures. Second, P-acid, first demonstrated a well-defined CL response with KMnO4, was selected as the CL reagent we describe here. The as-prepared NPS@P-acid composites were used as signal label with signal amplication technique. This low-cost, sensitive, stable, rapid biosensor can detect Hg(II) in real samples, which are well suited for facile on-site and real-time determination of Hg(II) in biological and environmental monitoring. It supplies a promising platform for detecting other metal ions by replacing the other metal-dependent aptamer.