Chemical nanosensors based on molecularly-imprinted polymers doped with silver nanoparticles for the rapid detection of caffeine in wastewater

Accepted Manuscript Chemical nanosensors based on molecularly-imprinted polymers doped with silver nanoparticles for the rapid detection of caffeine in wastewater Ran Hu, Rui Tang, Jiyang Xu, Feng Lu PII:

S0003-2670(18)30757-8

DOI:

10.1016/j.aca.2018.06.012

Reference:

ACA 236027

To appear in:

Analytica Chimica Acta

Received Date: 24 April 2018 Accepted Date: 5 June 2018

Please cite this article as: R. Hu, R. Tang, J. Xu, F. Lu, Chemical nanosensors based on molecularlyimprinted polymers doped with silver nanoparticles for the rapid detection of caffeine in wastewater, Analytica Chimica Acta (2018), doi: 10.1016/j.aca.2018.06.012. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

Chemical nanosensors based on molecularly-imprinted polymers doped with silver nanoparticles for the rapid

Ran Hu 1,2, Rui Tang 3, Jiyang Xu 2*, Feng Lu 1*

RI PT

detection of caffeine in wastewater

School of Pharmacy, Second Military Medical University, 200433, Shanghai, China;

2

School of Life Science and Technology, China Pharmaceutical University, Nanjing,

SC

1

3

M AN U

211198, Jiangshu, China;

West China School of Pharmacy, Sichuan University, Chengdu, 610041, Sichuan

China. *

Corresponding author: Dr. Feng Lu, Tel ： +86 21 81871260 E-mail:

EP

TE

Additional Corresponding author: Dr. Jiyang Xu, E-mail: [email protected]

AC C

*

D

[email protected]

ACCEPTED MANUSCRIPT

Abstract Caffeine is a common pharmaceutical and personal care product pollutant in wastewater. This work offers rapid and single-step detection of caffeine in an aquatic matrix based on high performance surface-enhanced Raman scattering (SERS). Novel

RI PT

chemical SERS nanosensors were developed employing molecularly-imprinted polymer (MIP) particles loaded with Ag nanoparticles (AgNPs) using precipitation polymerization to form AgNPs@MIP nanocomposites. Theophylline was applied as a dummy template molecule in the synthesis process due to its high structural similarity

SC

with caffeine and greater availability. The nanocomposite was characterized by Fourier transform infrared spectroscopy（FTIR）, X-ray diffraction（XRD）, and

M AN U

ultraviolet-visible (UV-vis) spectroscopy. Static and kinetic adsorption testing demonstrated the specific affinity of AgNPs@MIP nanocomposites for caffeine and a rapid adsorption equilibration rate. Moreover, a simple solid phase extraction cartridge comprising AgNPs@MIP nanocomposites as adsorbents (AgNPs@MISPE), a syringe, and a removable microporous membrane were employed to detect the

D

SERS signal of caffeine. The AgNPs@MISPE was used to detect caffeine with

TE

excellent uniformity (relative standard deviation, RSD=4.8%) and good repeatability (RSD=8.7%). The separation and detection processes were integrated into a single

EP

step, and the overall analysis time was 23 min. The detection limit was 100 ng L-1, which is less than the caffeine content reported in many rivers. The experimental

AC C

results demonstrate that the proposed chemical nanosensors are a low-cost and reliable tool for the rapid screening of caffeine in wastewater or other aquatic matrices.

Keywords: AgNPs@MIP nanocomposites; AgNPs@MISPE-SRES; Caffeine; PPCPs

1. Introduction

ACCEPTED MANUSCRIPT Pharmaceuticals and personal care products (PPCPs) are emerging chemical pollutants and organic, inorganic, biodegradable, and non-biodegradable forms that threaten aquatic environments and human health. A recent review of environmental contamination in different areas of China has concluded that the level of PPCP

RI PT

contamination is very serious in aquatic and natural environments and potentially harmful to humans. PPCPs persist in wastewater, groundwater, and drinking-water sources and have pharmacological effects on living organisms [1].

Caffeine is a PPCP that is released into the environment mainly via

SC

pharmaceutical wastewater, energy drinks, colas, tea, coffee beans, and drugs [3, 4]. An investigation of PPCP concentrations in the effluent-dominated coastal sea waters

M AN U

of the Red Sea demonstrated that caffeine was one of the most abundant PPCPs with a maximum concentration greater than 3 g L-1 [5]. This common occurrence of caffeine with other PPCPs indicates that caffeine is a reliable indicator of PPCPs in wastewater [6-8].

Current approaches to measure caffeine include high-performance liquid

D

chromatography (HPLC) [9], reversed-phase HPLC (RP-HPLC) [10], and liquid

TE

chromatography/mass spectrometry (LC-MS/MS) [11]. These are labor-intensive and time-consuming and require skilled operators and large amounts of environmentally

EP

unfriendly organic solvents. Most studies have mainly focused on the determination of caffeine residues in tablets, plasma, tea leaves, and coffee, but few studies have tested the concentration of caffeine in aquatic matrices. In addition, the relatively high

AC C

limit of detection (LOD) of these methods make them ill-suited for evaluating the occurrence of caffeine in trace concentrations ranging from ng/L to g/L in complex water samples [4, 6].

Recently, some new detection methods have been used to detect trace substances

including surface-enhanced Raman scattering (SERS), UV/Vis spectroscopy [12], graphene quantum dots [13], optical sensors [14, 15]. Of these, surface-enhanced Raman scattering (SERS) is a molecular spectroscopic analysis technique that provides high sensitivity and rapid analysis with wide application regimes. However, one of the main problems associated with SERS analysis is its low affinity and poor

ACCEPTED MANUSCRIPT selectivity for target molecules [16]. Moreover, the presence of interferents in complex aquatic matrices has hindered the application of SERS in these samples [17]. Therefore, the accurate separation and enrichment of the analyte in complex samples is a major challenge for SERS applications.

RI PT

The efficient enrichment and separation of analytes in complex matrices has been pursued using molecularly-imprinted polymers (MIPs) that employ a target analyte as a template to realize specific recognition. Thus, this technique forms a biomimetic material based on an antigen-antibody principle. Versus other molecular recognition

SC

systems, MIPs offer prefabricated structures, specific recognition, and universal applications [18, 19]. MIP technology has recently been coupled with SERS

M AN U

(MIP-SERS) as an emerging technique for the detection of specific molecules in complex samples. This coupling combines the specific recognition properties of MIPs and the signal amplification properties of SERS to yield exceptional selectivity and sensitivity.

MIP-SERS sensors usually employ either two-step or single-step detection

D

strategies. In two-step MIP-SERS detection, an MIP component is employed as an

TE

adsorbent substrate for solid-phase extraction. The first step involves loading, washing, and eluting. The eluted analyte molecules must then be subsequently mixed

EP

with the Ag or Au nanoparticles (i.e., Ag NPs or Au NPs) prior to SERS detection in the second step of the process. This results in complicated operation and long analysis In contrast, single-step detection nanosensors introduce the Ag NPs

AC C

times [20-26].

or Au NPs directly into the MIPs to form AgNPs@MIP or AuNPs@MIP nanocomposites for MIP-SERS nanosensors that integrate the separation and detection procedures into a single device [27-36]. Single-step detection nanosensors offer significant advantages, and many recent

studies have applied these nanosensors for the detection of chemical hazards. However, only two approaches have been reported for embedding Ag NPs directly into MIPs for the development of single-step MIP-SERS nanosensors. The first involves core-shell MIPs developed by coating a thin MIP film on Ag NPs or Ag thin films [18, 29, 31, 35, 37]. Unfortunately, the core-shell structure is based on surface

ACCEPTED MANUSCRIPT molecular imprinting that involves synthesis conditions that are difficult to control and reproduce. The other approach employs MIPs that are doped with Ag NPs by bulk polymerization for SERS sensing [28, 33, 38], but this is time-consuming, difficult to elute, and has a low imprinting efficiency. Therefore, the synthesis methods reported

RI PT

for single-step AgNPs@MIP or AuNPs@MIP nanocomposite MIP-SERS nanosensors have some substantial drawbacks.

To rectify these drawbacks, this work uses an in situ reduction method to load Ag NPs into MIP materials via precipitation polymerization. To the best of our

SC

knowledge, the novel AgNPs@MIP-based MIP-SERS nanosensors developed here are the first to provide single-step separation and detection of caffeine in wastewater.

M AN U

In addition, a simple SPE cartridge was constructed using the AgNPs@MIP nanocomposites as adsorbents for the selective separation and enrichment of caffeine. This eliminated the elution step and the final introduction of SERS substrates to reduce the analysis time (Fig. 1). The results demonstrate that the proposed SERS analysis approach employing AgNPs@MIP-based SPE (AgNPs@MISPE-SERS) is a

D

low-cost and reliable point-of-care tool for the rapid screening of caffeine in

TE

wastewater or other aquatic matrices.

2. Experimental section

EP

2.1 Materials

Methacrylic acid (MAA; 99% purity), ethylene glycol dimethacrylate (EGDMA;

AC C

98% purity), and 2,2’-azobis (isobutyronitrile) (AIBN; 99% purity) were purchased from Aladdin-Reagent Co., Ltd. (Shanghai, China); MAA and AIBN were purified via recrystallization prior to use. Methanol (analytical grade), acetonitrile (HPLC grade), ethanol (analytical grade), and acetic acid (analytical grade) were purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Silver nitrate, theophylline, and caffeine were provided from the National Institutes for Food and Drug Control of China (Beijing, China).

2.2 Synthesis of theophylline-templated AgNPs@MIP nanocomposites A schematic illustrating the synthesis of AgNPs@MIP nanocomposites is shown

ACCEPTED MANUSCRIPT in Fig. 1. Precipitate polymerization was used to prepare dummy-template MIP particles embedded with Ag NPs. Theophylline was used as the template based on a reported method with some improvements [33]. Briefly, theophylline (0.26 mmol) was dissolved in anhydrous acetonitrile (40 mL) and MAA (1.04 mmol) and

RI PT

magnetically stirred for 30 min. Then, EGDMA (4.16 mmol), AgNO3 (148 mg), and AIBN (14 mg) were added sequentially, and the solution was saturated with dry nitrogen for 5 min, and then sealed for 12 h. The resulting fine particles were collected by centrifugation at 8000 rpm for 10 min. Prior to extraction to remove the

SC

template, excess NaBH4 was added to the aqueous solution containing the AgNO3@MIP particles to reduce Ag+ into Ag nanoparticles, which were randomly

M AN U

distributed on the imprinted sites of the MIP particles.

UV-vis absorbance spectroscopy was performed before and after the reduction process to confirm the formation of AgNPs@MIP particles. The template was extracted by extensive washing with methanol:acetic acid (8:2, v/v) until no template molecules could be detected at 272 nm in the UV-vis spectrum. Finally, the remaining

D

polymer particles were further washed several times with methanol and dried under

TE

vacuum at 50 °C. A control solution composed of non-imprinted polymer (AgNPs@NIP) particles was prepared similarly except without the template

EP

molecules.

2.3 Absorption capacity tests of AgNPs@MIP nanocomposites

AC C

Adsorption capacity is the standard means of evaluating the binding capacity performance of AgNPs@MIP nanocomposites. To measure the static and kinetic adsorption capacities of the AgNPs@MIP nanocomposites for caffeine, 5 mg of the AgNPs@MIP particles were mixed with various 2 mL caffeine standard solutions with caffeine concentrations ranging from 0.1 to 0.4 mg mL-1. The cleaning effects of different incubation solvents including deionized water, methanol, acetonitrile, and their corresponding aqueous solutions were investigated. For static adsorption tests, the mixed solutions were shaken for 2 h at 25 °C and subsequently centrifuged at 6000 rpm for 1 min. The supernatant was purified by filtration through a 0.22-

ACCEPTED MANUSCRIPT nylon syringe filter membrane, and the remaining caffeine in the solution was analyzed according to the UV-vis spectrum at 272 nm. Kinetic adsorption tests were conducted by mixing 5 mg of the AgNPs@MIP particles with 2 mL of the 0.2 mg mL-1 caffeine standard solution and shaken for

RI PT

different periods (1, 5, 10, 20, 30, 45, 60, 90, 120, and 180 min). The caffeine concentration of the supernatant was determined according to the UV-vis spectrum at 272 nm after the centrifugation and filtration processes for static adsorption testing. The static and kinetic adsorption capacities of the AgNPs@NIP nanocomposites were

SC

similarly determined. The SERS intensities of AgNPs@MIP and AgNPs@NIP nanocomposites in the absence and presence of caffeine were also recorded by Raman

M AN U

spectroscopy at 785 nm to confirm and compare their specific affinities for caffeine.

2.4 Characterization of AgNPs@MIP nanocomposites Fourier transform infrared spectroscopy (FTIR) was conducted using an ALPHA (Bruker, Germany) instrument within a scan range of 4000–500 cm−1. UV-vis spectra were measured on a TU-1901 spectrophotometer (Puxi, China). X-ray diffraction

D

(XRD) patterns were obtained with an AXS D8 ADVANCE (Bruker, Germany)

TE

diffractometer using Cu-K radiation (λ = 1.5418 Å).

2.5 Sample pretreatment

EP

The water matrix samples were obtained from a local river (Qiujiang River, Shanghai, China). Each 2 ml sample was centrifuged at 8000 rpm for 2 min and

AC C

filtrated through a 0.22-

filtrate was

spiked with various caffeine contents for final concentrations of 0 ng L-1, 10 ng L-1, 100 ng L-1, 1 g L-1, 10 g L-1, 100 g L-1 and 1 mg L-1.

2.6 AgNPs@MIP nanocomposite SPE We evaluated the SPE of the proposed nanocomposites by mixing 5 mg of the AgNPs@MIP particles with 2 ml of the pretreated water matrix samples presented in Subsection 2.5 with different caffeine concentrations. After drawing the mixture with a syringe and horizontally shaking for 20 min, the syringe and filter membrane were assembled into a simple SPE cartridge (Fig. 1). After removing the excess solution,

ACCEPTED MANUSCRIPT the remaining materials were washed to remove non-specifically adsorbed impurities. The cleaning effects of various washing solvents including methanol, water, acetonitrile, chloroform, and acetonitrile/water (1:1, v/v) were examined to obtain an optimum washing solvent. Finally, the syringe was disassembled to collect the

RI PT

AgNPs@MIP nanocomposites with the adsorbed caffeine from the membrane prior to obtaining the SERS spectra. The SERS signal of the AgNPs@MIP nanocomposites with adsorbed caffeine from the membrane could then be measured directly without the elution of the water matrix and the addition of Ag NPs as SERS substrates. This

SC

greatly simplifies the analysis process and shortens the required analysis time.

2.7 SERS spectra collection

M AN U

The SERS spectra were obtained using a portable BWS415 Raman spectrometer (B&W Tek, USA) with a 785 nm laser source, a resolution of 5 cm-1, and a 20× long working distance microscope objective. The measurements were conducted three times with a total accumulation time of 8 s per spectrum, and the average SERS intensity was recorded. All spectra were smoothed and the baselines were corrected.

D

3. Results and Discussion

TE

3.1 Synthesis of AgNPs@MIP nanocomposites

EP

Schematic illustration of the synthesis of AgNPs@MIP nanocomposites is shown in Fig. 1. Recently, the single-step MIP-SERS process was used for the

AC C

separation and detection of target analytes through bulk polymerization and surface molecular imprinting. Between them, bulk polymerization is the most common method for synthesizing MIP. This leads to poor recognition, tedious processing, irregular shapes and sizes, and low capacity for templates [39, 40]. Surface molecular imprinting can overcome the drawbacks of bulk

polymerization, but surface molecular imprinting technique has poor repeatability and difficult synthesis conditions. This is incompatible with industrial applications. In prior synthesis methods, precipitation techniques are the most convenient and offer only a single preparative step [39]. Versus other single-step polymerization methods, precipitation polymerization is simple and produces uniform products (Table. 1).

ACCEPTED MANUSCRIPT Therefore, this work used AgNPs@MIP nanocomposites selective for caffeine prepared via precipitation polymerization. Theophylline was used for control templates for several reasons: 1) The structural skeleton of theophylline is very similar to that of caffeine; 2) Theophylline is not a

RI PT

controlled substance like caffeine; and 3) Precipitation polymerization requires a large number of template molecules and theophylline is available at low cost.

3.2 Characterization of AgNPs@MIP nanocomposites

In the process of precipitation polymerization, Ag+ in AgNO3 formed an

SC

electrostatic interaction with the dummy template molecules. The Ag+ was reduced to AgNPs after treatment with NaBH4, and UV peaks were observed for AgNPs@MIP

M AN U

nanocomposites near 408 nm; there were no UV peaks at 408 nm for the MIP particles alone. As a result, the Ag NPs were successfully incubated into the MIP particles and formed a AgNPs@MIP nanocomposite as expected. The AgNPs@MIP nanocomposites ensure Raman enhancement because the caffeine molecules are preferentially adsorbed by the MIP particles. They are then combined onto the surface

D

of the AgNPs.

TE

The AgNPs@MIP nanocomposites were characterized by FTIR spectroscopy before and after the adsorption of caffeine to further confirm the specific affinity of

EP

AgNPs@MIP nanocomposites for caffeine (Fig. 3). There are three main absorption bands at 1047, 1558, and 1662 cm-1, and some fingerprints appear in the spectra after

AC C

the adsorption of caffeine. These main absorption bands are attributed to the stretching of N-CH3, C=C, and C=O, respectively. The N-CH3 bonds were mainly derived from the methyl functional groups of the pyrimidine ring, and the stretching vibration peaks of C=C at 1558 cm-1 resulted from the frame vibration of the pyrimidine ring. The new absorption band at 1662 cm-1 is attributed to the C=O bonds of caffeine. Furthermore, the strong and broad peak at 1722 cm−1 confirms the C=O vibrations for AgNPs@MIP nanocomposites (Fig. 3a); this vibration red shifts 4 cm-1 after the adsorption of caffeine (Fig. 3b). This shift is most likely due to the formation of hydrogen bonds between the AgNPs@MIP nanocomposites and the adsorbed

ACCEPTED MANUSCRIPT caffeine. These results demonstrate that the AgNPs@MIP nanocomposites include three-dimensional cavities whose hole sizes and hydrogen bonds were a match for caffeine molecules. The crystal structures of the AgNPs@MIP nanocomposites were further

RI PT

characterized by X-ray diffraction (XRD). The five predominant diffraction peaks of the AgNPs@MIP nanocomposites located at 2θ = 38.15°, 44.3°, 64.41°, 77.42°, and 81.60° nicely match the face-centered cubic (fcc) Ag crystal phase (111), (200), (220), (311), and (222) crystal planes (Fig. 4) indicating the formation of crystalline Ag NPs

SC

in the MIP particles. According to the working principle of SERS, Ag NPs of a few tens of nanometers can usually significantly enhance the Raman scattering of

M AN U

molecules approaching the surface of a SERS-active substrate because of the long-range enhancement effects generated by electromagnetic mechanisms [41]. Fig. 5 presents the UV-vis absorption spectra obtained during the removal of template molecules in the supernatant. The results can be clearly understood as a function of cleaning time where the intensity of the initially strongest peak gradually

D

declines with progressing elution—the elution of template creates imprinted cavities

TE

specific to caffeine molecules. This peak is almost completely eliminated after six elution steps, which demonstrates that the theophylline molecules have been

EP

thoroughly eliminated from the AgNPs@MIP nanocomposites for complete synthesis.

3.3 Specific molecular recognition capacity

AC C

The static and kinetic adsorption capacities of the synthesized AgNPs@MIP and AgNPs@NIP nanocomposites were investigated to evaluate their capacity to bind caffeine. Static adsorption testing assessed the equilibrium time for binding between the analyte and the AgNPs@MIP nanocomposites where C represents the initial concentration of caffeine, and Q is the adsorption capacity. Fig. 6A indicates that the binding equilibrium of AgNPs@MIP nanocomposites was attained in less than 20 min confirming that the nanocomposites offer rapid recognition and extraction of caffeine from complex water matrices. The kinetic adsorption capacity results are presented in Fig. 6B. The binding

ACCEPTED MANUSCRIPT capacity performance of AgNPs@MIP nanocomposites is always better than AgNPs@NIP nanocomposites, which further verifies the specific recognition ability of AgNPs@MIP nanocomposites for caffeine. The good adsorption performance is expected to be associated with interactions between AgNPs@MIP nanocomposites

RI PT

and the analyte. This adsorption results from specific hydrogen bonds and electrostatic interactions. However, the AgNPs@NIP nanocomposites reach equilibrium mainly due to the weak adsorption and non-specific recognition for caffeine. Thus, the AgNPs@MIP nanocomposites can be utilized for rapid and

SC

selective detection of caffeine.

3.4 Rebinding adsorption experiments by SERS

M AN U

We investigated the SERS activity of AgNPs@MIP and AgNPs@NIP nanocomposites using caffeine as the target analyte. The generation of localized excitation must usually meet two conditions. First, the distance between Ag NPs and target molecules must be less than 10 nm [42]. Second, the Ag NPs should be a few tens of nanometers in size [16]. The AgNPs@MIP nanocomposites had typical surface

D

plasmon resonance (SPR) adsorption bands of Ag NPs at 408 nm (Fig. 2), which

TE

indicates that the size of the Ag NPs provides excellent Raman enhancement. The AgNPs@MIP nanocomposites offer an obviously stronger SERS signal for caffeine

EP

than AgNPs@NIP nanocomposites (Fig. 7) further indicating that the adsorption capacity of AgNPs@MIP nanocomposites is higher than the AgNPs@NIP

AC C

nanocomposites.

3.5 The repeatability of AgNPs@MISPE method The repeatability of the AgNPs@MISPE method was investigated in a 1 mg L-1

caffeine solution for 20 min. The repeatability was studied via the Raman signal between different points on the AgNPs@MIP substrate and the SERS signal of multiple repeated experiments at the same concentration. We studied variation in signal on the SERS substrate. The Raman signal of caffeine in the AgNPs@MIP nanocomposites was obtained after the SPE was washed. Fig. 8A shows the Raman spectra obtained from 9 random points.

ACCEPTED MANUSCRIPT The Raman bands at 554 cm-1 represent the strongest characteristic caffeine peak, and this peak was selected as the characteristic peak for calculation of RSD. The values of RSD for a peak intensity at 554 cm-1 of caffeine was 4.8%. Similarly, the repeatability of the AgNPs@MISPE method was evaluated by comparing the SERS

RI PT

spectra for nine parallel experiments (RSD of 8.7 %; Fig. 8B). The repeatability of the as-prepared single-step AgNPs@MISPE-SERS was higher than that reported in the

caffeine and AgNPs.

3.6 Measurement of caffeine in river water

SC

two-step method [24, 43]. The RSD is mainly a function of the distance between

The simple SPE cartridge discussed in Subsection 2.6 was used to detect caffeine

M AN U

in river water. The cleaning process involved various washing solvents, and acetonitrile/water (1:1, v/v) best removed impurities from the sample leading to a minimal impact on the SERS signal. In addition, the acetonitrile/water solution provided the best cleaning effect. Several background peaks from the wastewater spectrum were removed with cleaning (Fig. 10), several main peaks of the wastewater

D

signal were nearly unobservableafter cleaning, while the SERS signal for caffeine

TE

remained largely unchanged. the SERS signal for caffeine remained largely unchanged. The results indicate that the AgNPs@MIPs nanocomposites offer a

EP

significantly selective uptake of caffeine over impurities in wastewater. Moreover, the entire SPE-based analysis process was only 23 min, which is less than the 25 minutes

AC C

reported in the single-step method using bulk polymerization [28]. Next, the LOD of the AgNPs@MISPE-SERS was determined via serial dilutions.

In Fig. 9, the intensity of the characteristic peak decreased with decreasing caffeine concentration. The Raman band at 554 cm-1 was selected as the characteristic peak for the determination of the LOD. When the caffeine concentration was diluted to 10 ng L-1, no Raman bands at 554 cm-1 could be identified (less than three times signal-to-noise ratio). However, the characteristic caffeine peak in the SERS spectra can be clearly observed at a concentration as low as 100 ng L-1. Therefore, the LOD of this method was 100 ng L-1. Although the LOD of this single-step

ACCEPTED MANUSCRIPT AgNPs@MISPE-SERS is worse than other reports [44, 45], it is still lower than standard caffeine concentrations found in water [5, 7, 46-49]. The results implied that the proposed SERS nanosensors based on SPE using AgNPs@MIP nanocomposites are very promising for analytical applications involving complex water samples.

RI PT

4. Conclusion

This article presented an effective single-step approach for the selective and

SC

sensitive detection of caffeine in complex water matrices. The method employed nanocomposites of MIP particles doped with Ag NPs via precipitation polymerization

M AN U

as an active SERS substrate. Theophylline was applied as a dummy template molecule in the synthesis process due to its high structural similarity with caffeine and greater availability. Moreover, the resulting AgNPs@MIP nanocomposites were employed as sorbents for SPE, which facilitated integrated separation, cleaning, and SERS detection of caffeine. The LOD of the SPE detection process for caffeine was

D

100 ng L-1, which is less than the working concentration of caffeine in many studies. The method has good repeatability, and it can rapidly detect caffeine in river water.

TE

The entire analysis process took only 23 min. Thus, this method is fast, simple, and low-cost. It has substantial potential for determining trace levels of other PPCPs in

AC C

EP

complex water matrices.

Acknowledgments This research was supported by research funds from the National Natural

Science Foundation of China (grant no 81573598) and the Ministry of Science and Technology of the People’s Republic of China (grant no. 2018ZX09J18112, 2012YQ180132).

Reference

ACCEPTED MANUSCRIPT

AC C

EP

TE

D

M AN U

SC

RI PT

[1] J.L. Liu, M.H. Wong, Pharmaceuticals and personal care products (PPCPs): a review on environmental contamination in China, Environ. Int. 59 (2013) 208-224. [2] R. Ma, B. Wang, L. Yin, Y. Zhang, S. Deng, J. Huang, Y. Wang, G. Yu, Characterization of pharmaceutically active compounds in Beijing, China: Occurrence pattern, spatiotemporal distribution and its environmental implication, J. Hazard. Mater. 323 (2016) 147-155. [3] A. Carmalin Sophia, E.C. Lima, N. Allaudeen, S. Rajan, Application of graphene based materials for adsorption of pharmaceutical traces from water and wastewater- a review, Desalin. Water Trea. 57 (2016) 1-14. [4] G. Archana, R. Dhodapkar, A. Kumar, Offline solid-phase extraction for preconcentration of pharmaceuticals and personal care products in environmental water and their simultaneous determination using the reversed phase high-performance liquid chromatography method, Environ. Monit. Assess. 188 (2016) 512. [5] A.M. Ali, H.T. Ronning, W. Alarif, R. Kallenborn, S.S. Al-Lihaibi, Occurrence of pharmaceuticals and personal care products in effluent-dominated Saudi Arabian coastal waters of the Red Sea, Chemosphere 175 (2017) 505-513. [6] Y.Y. Yang, W.R. Liu, Y.S. Liu, J.L. Zhao, Q.Q. Zhang, M. Zhang, J.N. Zhang, Y.X. Jiang, L.J. Zhang, G.G. Ying, Suitability of pharmaceuticals and personal care products (PPCPs) and artificial sweeteners (ASs) as wastewater indicators in the Pearl River Delta, South China, Sci. Total Environ. 590-591 (2017) 611-619. [7] N.H. Tran, J. Li, J. Hu, S.L. Ong, Occurrence and suitability of pharmaceuticals and personal care products as molecular markers for raw wastewater contamination in surface water and groundwater, Environ. Sci. Pollut. Res. Int. 21 (2014) 4727-4740. [8] M. Lv, Q. Sun, A.Y. Hu, L.Y. Hou, J.W. Li, X. Cai, C.P. Yu, Pharmaceuticals and personal care products in a mesoscale subtropical watershed and their application as sewage markers, J. Hazard. Mater. 280 (2014) 696-705. [9] H.S. Ali, A.A. Abdullah, P.T. Pinar, Y. Yardim, Z. Senturk, Simultaneous voltammetric determination of vanillin and caffeine in food products using an anodically pretreated boron-doped diamond electrode: Its comparison with HPLC-DAD, Talanta 170 (2017) 384-391. [10] S.C. Escobar, L.R. Cubides, C.P. Perez, Optimization and Validation of a Simple and Fast RP-HPLC Method for Simultaneous Determination of Acetaminophen and Caffeine in Tablets, Indian J. Pharm. Sci. 79 (2017) 731-739. [11] E. Lopez-Garcia, N. Mastroianni, C. Postigo, Y. Valcarcel, S. Gonzalez-Alonso, D. Barcelo, M.L. de Alda, Simultaneous LC-MS/MS determination of 40 legal and illegal psychoactive drugs in breast and bovine milk, Food Chem. 245 (2018) 159-167. [12] K. Dashtian, R. Zare-Dorabei, Preparation and characterization of a novel optical chemical sensor for determination of trace amounts of Praseodymium ion by UV/Vis spectrophotometry, Sensor. Actuat. B: Chem. 242 (2017) 586-594. [13] F. Nemati, R. Zare-Dorabei, M. Hosseini, M.R. Ganjali, Fluorescence turn-on sensing of thiamine based on Arginine – functionalized graphene quantum dots

ACCEPTED MANUSCRIPT

AC C

EP

TE

D

M AN U

SC

RI PT

(Arg-GQDs): Central composite design for process optimization, Sensor. Actuat. B: Chem. 255 (2018) 2078-2085. [14] M.R. Ganjali, R. Zare-Dorabei, P. Norouzi, Design and construction of a novel optical sensor for determination of trace amounts of dysprosium ion, Sensors Actuat. B-Chem. 143 (2009) 233-238. [15] M. Noroozifar, M.K. Motlagh, A. Taheri, R.Z. Dorabei, Diphenylthiocarbazone immobilized on the triacetyl cellulose membrane as an optical silver sensor, Turk. J. Chem. 32 (2008) 249-257. [16] W.B. Li, X.C. Zhao, Z.F. Yi, A.M. Glushenkov, L.X. Kong, Plasmonic substrates for surface enhanced Raman scattering, Anal. Chim. Acta 984 (2017) 19-41. [17] C.D.L. de Albuoluerque, R.G. Sobral, R.J. Poppi, A.G. Brolo, Digital Protocol for Chemical Analysis at Ultralow Concentrations by Surface-Enhanced Raman Scattering, Anal. Chem. 90 (2018) 1248-1254. [18] Z. Wu, E. Xu, J. Li, J. Long, A. Jiao, Z. Jin, Highly sensitive determination of ethyl carbamate in alcoholic beverages by surface-enhanced Raman spectroscopy combined with a molecular imprinting polymer, RSC Adv. 6 (2016) 109442-109452. [19] M.K. Bojdi, M. Behbahani, G. Hesam, M.H. Mashhadizadeh, Application of magnetic lamotrigine-imprinted polymer nanoparticles as an electrochemical sensor for trace determination of lamotrigine in biological samples, Rsc Adv. 6 (2016) 32374-32380. [20] F. Gao, Y. Hu, D. Chen, E.C.Y. Li-Chan, E. Grant, X. Lu, Determination of Sudan I in paprika powder by molecularly imprinted polymers-thin layer chromatography-surface enhanced Raman spectroscopic biosensor, Talanta 143 (2015) 344-352. [21] Y. Hu, S. Feng, F. Gao, E.C. Li-Chan, E. Grant, X. Lu, Detection of melamine in milk using molecularly imprinted polymers-surface enhanced Raman spectroscopy, Food Chem. 176 (2015) 123-129. [22] Y. Xie, M. Zhao, Q. Hu, Y. Cheng, Y. Guo, H. Qian, W. Yao, Selective detection of chloramphenicol in milk based on a molecularly imprinted polymer-surface-enhanced Raman spectroscopic nanosensor, J. Raman Spectrosc. 48 (2017) 204-210. [23] F. Gao, E. Grant, X. Lu, Determination of histamine in canned tuna by molecularly imprinted polymers-surface enhanced Raman spectroscopy, Anal. Chim. Acta 901 (2015) 68-75. [24] X. Xiao, K. Yan, X. Xu, G. Li, Rapid analysis of ractopamine in pig tissues by dummy-template imprinted solid-phase extraction coupling with surface-enhanced Raman spectroscopy, Talanta 138 (2015) 40-45. [25] F. Gao, S. Feng, Z. Chen, E.C. Li-Chan, E. Grant, X. Lu, Detection and quantification of chloramphenicol in milk and honey using molecularly imprinted polymers: Canadian penny-based SERS nano-biosensor, J. Food. Sci. 79 (2014) 2542-2549. [26] T. Kamra, T. Zhou, L. Montelius, J. Schnadt, L. Ye, Implementation of molecularly imprinted polymer beads for surface enhanced Raman detection, Anal. Chem. 87 (2015) 5056-5061.

ACCEPTED MANUSCRIPT

AC C

EP

TE

D

M AN U

SC

RI PT

[27] T. Shahar, T. Sicron, D. Mandler, Nanosphere molecularly imprinted polymers doped with gold nanoparticles for high selectivity molecular sensors, Nano Res. 10 (2017) 1056-1063. [28] Y. Hu, X. Lu, Rapid Detection of Melamine in Tap Water and Milk Using Conjugated "One-Step" Molecularly Imprinted Polymers-Surface Enhanced Raman Spectroscopic Sensor, J. Food Sci. 81 (2016) 1272-1280. [29] S. Chen, X. Li, Y. Guo, J. Qi, A Ag-molecularly imprinted polymer composite for efficient surface-enhanced Raman scattering activities under a low-energy laser, Analyst 140 (2015) 3239-3243. [30] S. Chen, X. Li, Y. Zhao, L. Chang, J. Qi, High performance surface-enhanced Raman scattering via dummy molecular imprinting onto silver microspheres, Chem. Commun. 50 (2014) 14331-14333. [31] S.N. Chen, X. Li, S. Han, J.H. Liu, Y.Y. Zhao, Synthesis of surface-imprinted Ag nanoplates for detecting organic pollutants in water environments based on surface enhanced Raman scattering, RSC Adv. 5 (2015) 99914-99919. [32] L. Chang, Y. Ding, X. Li, Surface molecular imprinting onto silver microspheres for surface enhanced Raman scattering applications, Biosens. Bioelectron. 50 (2013) 106-110. [33] P. Liu, R. Liu, G. Guan, C. Jiang, S. Wang, Z. Zhang, Surface-enhanced Raman scattering sensor for theophylline determination by molecular imprinting on silver nanoparticles, Analyst 136 (2011) 4152-4158. [34] T. Kamra, C. Xu, L. Montelius, J. Schnadt, S.A. Wijesundera, M. Yan, L. Ye, Photoconjugation of Molecularly Imprinted Polymer Nanoparticles for Surface-Enhanced Raman Detection of Propranolol, ACS Appl. Mater. Interfaces 7 (2015) 27479-27485. [35] L. Zhang, Y. Jin, X. Huang, Y. Zhou, S. Du, Z. Zhang, Ligand Replacement Approach to Raman-Responded Molecularly Imprinted Monolayer for Rapid Determination of Penicilloic Acid in Penicillin, Anal. Chem. 87 (2015) 11763-11770. [36] J.Q. Xue, D.W. Li, L.L. Qu, Y.T. Long, Surface-imprinted core-shell Au nanoparticles for selective detection of bisphenol A based on surface-enhanced Raman scattering, Anal. Chim. Acta 777 (2013) 57-62. [37] Y. Guo, L. Kang, S. Chen, X. Li, High performance surface-enhanced Raman scattering from molecular imprinting polymer capsulated silver spheres, Phys. Chem. Chem. Phys. 17 (2015) 21343-21347. [38] S.j. Li, S. Gong, A Substrate-Selective Nanoreactor Made of Molecularly Imprinted Polymer Containing Catalytic Silver Nanoparticles, Adv. Funct. Mater. 19 (2009) 2601-2606. [39] M.K. Bojdi, M. Behbahani, M. Najafi, A. Bagheri, F. Omidi, S. Salimi, Selective and Sensitive Determination of Uranyl Ions in Complex Matrices by Ion Imprinted Polymers-Based Electrochemical Sensor, Electroanal. 27 (2015) 2458-2467. [40] M.K. Bojdi, M. Behbahani, A. Sahragard, B.G. Amin, A. Fakhari, A. Bagheri, A palladium imprinted polymer for highly selective and sensitive electrochemical determination of ultra-trace of palladium ions, Electrochim. Acta 149 (2014) 108-116.

ACCEPTED MANUSCRIPT

AC C

EP

TE

D

M AN U

SC

RI PT

[41] Y. Wang, W. Ji, H.M. Sui, Y. Kitahama, W.D. Ruan, Y. Ozaki, B. Zhao, Exploring the Effect of Intermolecular H-Bonding: A Study on Charge-Transfer Contribution to Surface-Enhanced Raman Scattering of p-Mercaptobenzoic Acid, J. Phys. Chem. C 118 (2014) 10191-10197. [42] H.J. Li, X.N. Wang, Z.R. Wang, J.Q. Jiang, Y. Qiao, M.B. Wei, Y.S. Yan, C.X. Li, A high-performance SERS-imprinted sensor doped with silver particles of different surface morphologies for selective detection of pyrethroids in rivers, New J. Chem. 41 (2017) 14342-14350. [43] D. Bekana, R. Liu, S.S. Li, Y.J. Lai, J.F. Liu, Facile fabrication of silver nanoparticle decorated alpha-Fe2O3 nanoflakes as ultrasensitive surface-enhanced Raman spectroscopy substrates, Anal. Chim. Acta 1006 (2018) 74-82. [44] M.C.C.V. Marcia S. Bispo, Heloísa Lúcia C. Pinheiro1, Rodolfo F.S. DeOliveira, JoséOscar N. Reis, and Jailson B. De Andrade, Simultaneous Determination of Caffeine, Theobromine, and Theophylline by High-Performance Liquid Chromatography, J. Chromatogr. Sci. 40 (2002) 45-48. [45] X.Z. Piero R. Gardinalia, Trace determination of caffeine in surface water samples by liquid chromatography–atmospheric pressure chemical ionization–mass spectrometry (LC–APCI–MS), Environ. Int. 28 (2002) 521-528. [46] E. Berger, P. Haase, M. Kuemmerlen, M. Leps, R.B. Schafer, A. Sundermann, Water quality variables and pollution sources shaping stream macroinvertebrate communities, Sci. Total Environ. 587 (2017) 1-10. [47] E.S.G.S.V.R.E.V.d. Silva-Filho, The use of caffeine as a chemical marker of domestic wastewater contamination in surface waters: seasonal and spatial variations in Teresópolis, Brazil, Ambiente & Água (2016). [48] L.C. Bodhipaksha, C.M. Sharpless, Y.P. Chin, A.A. MacKay, Role of effluent organic matter in the photochemical degradation of compounds of wastewater origin, Water Res. 110 (2017) 170-179. [49] A.Y.C. Lin, C.F. Lin, Y.T. Tsai, H.H.H. Lin, J. Chen, X.H. Wang, T.H. Yu, Fate of selected pharmaceuticals and personal care products after secondary wastewater treatment processes in Taiwan, Water Sci. Technol. 62 (2010) 2450-2458.

D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

TE

Table. 1

Comparison of the disadvantage of Two-step and Single-step MISPE-SERS Ref.

Bulk polymerization

Laborious elution step and subsequent addition

Precipitation polymerization

of SERS substrates; complicated operation and

Surface molecular imprinting

long analysis time

AC C

Two-step

Method

EP

MISPE-SERS

[20-26]

technique

Bulk polymerization

Single-step

Surface molecular imprinting technique (Not coupled with SPE)

Time-consuming,

difficult

to

elute,

low

imprinting efficiency；Laborious elution step and the final introduction of SERS substrates Difficult

synthesis

conditions;

poor

reproducibility; not suitable for industrial applications Simple synthesis; mono-dispersity; no need for

Our work

Precipitation polymerization

[28, 33, 38]

elution step and final introduction of SERS substrates

[18, 29, 31, 35, 37]

Fig. 1

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

Synthesis of theophylline-templated AgNP@MIP nanocomposites and

AC C

EP

TE

D

subsequent SPE of caffeine prior to SERS detection.

Fig. 2

UV-vis spectra of AgNPs@MIP nanocomposites before (a) and after (b) the

reduction of Ag+. The strongest adsorption peak at 408 nm shows the surface plasmon resonance of silver nanoparticles

Fig. 3

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

FTIR spectra of AgNPs@MIP nanocomposites before (a) and after (b) the

adsorption of caffeine. The variation in four main absorption bands at 1047, 1558,

AC C

EP

TE

D

1662 and 1722 cm-1 is mainly due to the adsorption of caffeine.

Fig. 4

XRD patterns of Ag and AgNPs@MIP nanocomposites. (a) the

AgNPs@MIP particles doped with silver nanoparticles; (b) the MIP particles without silver nanoparticles doped.

Fig. 5

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

UV-vis spectra of theophylline in the supernatant liquid with increasing

elution times. The UV spectra curves from top to bottom are the first to the sixth

AC C

EP

TE

D

elution.

Fig. 6

(A) Static adsorption capacity of AgNPs@MIP nanocomposites for caffeine

at different time points (3 to 120 min). (B) Kinetic adsorption capacity of the synthesized AgNPs@MIP and AgNPs@NIP nanocomposites for caffeine, and the concentrations range from 0.1 to 0.35 mmol L-1. Here, C represents the initial concentration of caffeine and Q is the adsorption capacity.

Fig. 7

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

SERS spectra of AgNPs@MIP and AgNPs@NIP nanocomposites before and

after rebinding caffeine in 1 mg L−1 caffeine aqueous solution.: (a) AgNPs@MIP nanocomposites after rebinding caffeine in 1 mg L−1 caffeine aqueous solution; (b) SERS spectra of caffeine 1 mg L−1 caffeine aqueous solution; (c) AgNPs@NIP

D

nanocomposites after rebinding caffeine in 1 mg L−1 caffeine aqueous solution; (d)

AC C

EP

TE

AgNPs@MIP nanocomposites before rebinding caffeine.

Fig. 8

The Raman signal between different points on the AgNPs@MIP substrate

and the SERS signal of multiple repeated experiments with caffeine aqueous solution at the concentration of 1 mg L−1. (A) SERS spectra of caffeine collected from 9 random points on the AgNPs@MIP substrate using AgNPs@MISPE-SERS method; (B) SERS spectra of caffeine collected from 9 repeated experiments using

ACCEPTED MANUSCRIPT

Fig. 9

M AN U

SC

RI PT

AgNPs@MISPE-SERS method.

Concentration-dependent SERS spectra of AgNPs@MIP nanocomposites

incubated with caffeine aqueous solution at the concentration within the range from 0

D

to 100 g L-1, and the Raman shift at 554 cm-1 is considered as the characteristic

AC C

EP

TE

bands of caffeine. Each data point is the average result of three SERS spectra.

ACCEPTED MANUSCRIPT Fig. 10 Representative SERS spectral features of 0.2 mmol L-1 caffeine in wastewater after SPE by the AgNPs@MIPs nanocomposites as a positive control (a), SERS spectral features of 0.2 mmol L-1 caffeine in waste-water before SPE by the AgNPs@MIPs nanocomposites (b), and non-spiked wastewater as a negative control

AC C

EP

TE

D

M AN U

SC

RI PT

(c).

ACCEPTED MANUSCRIPT 

We describe novel molecularly-imprinted polymer (MIP) surface enhanced Raman scattering (SERS) nanosensors employing an integrated single-step detection



Silver nanoparticles are introduced into the MIP material using precipitation



RI PT

polymerization. The integrated approach greatly reduces the overall required analysis time versus conventional two-step processes by omitting the separate eluting step and subsequent addition of SERS substrates.

EP

TE

D

M AN U

personal care product pollutant.

SC

The proposed nanosensors rapidly detect caffeine as a pharmaceutical and

AC C

