Author’s Accepted Manuscript One-pot synthesis of mesoporous structured ratiometric fluorescence Molecularly imprinted sensor for highly sensitive detection of melamine from milk samples Shoufang Xu, Hongzhi Lu www.elsevier.com/locate/bios
PII: DOI: Reference:
S0956-5663(15)30162-7 http://dx.doi.org/10.1016/j.bios.2015.05.064 BIOS7728
To appear in: Biosensors and Bioelectronic Received date: 31 March 2015 Revised date: 19 May 2015 Accepted date: 28 May 2015 Cite this article as: Shoufang Xu and Hongzhi Lu, One-pot synthesis of mesoporous structured ratiometric fluorescence Molecularly imprinted sensor for highly sensitive detection of melamine from milk samples, Biosensors and Bioelectronic, http://dx.doi.org/10.1016/j.bios.2015.05.064 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 galley proof before it is published in its final citable 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.
One-pot synthesis of mesoporous structured ratiometric fluorescence molecularly imprinted sensor for highly sensitive detection of melamine from milk samples Shoufang Xu*, Hongzhi Lu School of Chemistry and Chemical Engineering, Linyi University, Linyi 276005, China. *Corresponding author Tel.: +86 539 8766000; fax: +86 539 8766000. E-mail address: [email protected] (S.F. Xu)
Abstract: A facile strategy was developed to prepare mesoporous structured ratiometric fluorescence molecularly imprinted sensor for highly sensitive and selective determination of melamine using CdTe QDs as target sensitive dye and hematoporphyrin as reference dyes. One-pot synthesis method was employed because it could simplify the imprinting process and shorten the experimental period. The as-prepared fluorescence MIPs sensor, which combined ratiometric fluorescence technique with mesoporous silica materials into one system, exhibited excellent selectivity and sensitivity. Under optimum conditions, these mesoporous structured ratiometric fluorescence MIP@QDs sensors showed detection limit as low as 38 nM, which was much lower than those non-mesoporous one. The recycling process was sustainable at least 10 times without obvious efficiency decrease. The feasibility of 1
the developed method in real samples was successfully evaluated through the analysis of melamine in raw milk and milk powder samples with satisfactory recoveries of 92– 101%. The developed method proposed in this work proved to be a convenient, rapid, reliable and practical way to prepared high sensitive and selective fluorescence sensors with potentially applicable for trace pollutants analysis in complicated samples.
Keywords: Molecularly imprinted polymers; Ratiometric fluorescence probe; mesoporous silica; Melamine; one-pot synthesis method 1. Introduction Molecularly imprinted polymers (MIPs) are synthetic polymeric materials with specific recognition sites complementary in shape, size, and functional groups to template molecules (Chen et al., 2011; Shen et al., 2014). Fluorescence sensors show the potential application in detection trace amounts of analytes because of their sensitivity, simplicity, and cost-effective instrumentation (Yuan et al., 2013; Kim et al., 2012). Combined the high sensitivity of fluorescence sensor with the high selectivity of MIPs, fluorescence MIPs sensors displayed apparent advantages when detection of trace substances in complicated matrix (Ivanova-Mitseva et al. 2012; Beyazit et al., 2014). Up to date, two methods were widely used to prepare fluorescence MIPs sensors. One was using fluorescence functional monomer for molecular imprinting. For example, Sellergren group (Wagner et al., 2013; Wan et al. 2013) prepared fluorescence MIPs sensors using naphthalimide-based fluorescent monomers, which 2
were equipped with a urea binding site to carboxylate-containing guests. Wu (Wu et al., 2015) prepared fluorescent MIPs using dansyl methacrylate as functional monomer,
which
displayed
selective,
concentration-dependent
fluorescence
quenching in response to bisphenol A. However, this method often involved in complicated synthesis procedure for fluorescence functional monomer. The other, also the most commonly used method for proposing fluorescence MIPs sensor was encapsulating semiconductor quantum dots (QDs) into MIPs. A number of MIPs@QDs sensors have been developed to detect various compounds. For example, MIPs coated CdTe QDs sensor have been used for detection of cytochrome c (Zhang et al., 2011), bovine hemoglobin (Li et al., 2013), TNT (Xu et al., 2013), clenbuterol and melamine (Huy et al., 2014); MIPs coated Mn-ZnS QDs sensors for 4nonylphenol (Liu et al., 2010), cyphenothrin (Ren et al., 2015), tetrabromobisphenol A (Chen et al., 2012), pesticides (Zhao et al., 2012), et.al also have been synthesized. This method of directly doping fluorescent substance into MIPs enjoys the advantages of simple synthesis procedure. However, the sensitivity of the MIPs@QDs sensors often be weakened. The reason would be the following. First, QDs were not uniformly embedded into highly cross-linked MIPs, so it is difficult to guarantee that each imprinted cavity was surrounded by fluorescent QDs. In the meanwhile, the morphology of MIPs@QDs sensor was often solid sphere or irregular bulk particles, which was not beneficial for mass transfer. The second reason may be that the mostly reported detection modes of MIPs@QDs sensors were based on signal “turn-off” attributed to fluorescent quenching. It was a common accepted fact that fluorescence 3
sensors based on signal “turn-on” were often able to achieve lower detection limits due to the lower optical background while ratiometric fluorescence sensor always have wide linear range and weaken interference. Considering that relative lower sensitivity hinder the application of fluorescence MIPs sensors, developing MIPs@QDs sensors with high sensitivity and selectivity were highly demanded. The morphology of sensor is significant for highly sensitive identification. So constructing hierarchical architectures for sensor, such as mesoporous structure, highly ordered 3D inverted opal structures or nanowire is an effective strategy to enhance the sensitivity of the sensor. For example, Zhang’s group constructed fluorescence sensor with high sensitivity for detection of explosives based on inverse opal structure (Fang et al., 2009); Yan’s group build mesoporous structured fluorescent MIPs sensor for highly sensitive detection of metal ions (Tan et al., 2009) and carbohydrates molecules (Tan et al., 2009). The results indicated that interconnected mesoporous structures can greatly decrease mass-transport resistance and provide easier accessibility to the target species. These features make mesoporous silica particularly suitable for the applications in chemo/biosensors. Ratiometric fluorescence technique has attracted increasing attention owing to its improved sensitivity. Ratiometric fluorescence QDs probes for copper ions (Yao et al., 2013) and TNT (Zhang et al., 2011) confirmed that ratiometric fluorescence QDs probes can achieve high sensitivity and visual signal output observed by the naked eye. Considering the above mentioned, in this paper, fluorescence MIPs sensor for high sensitive and selective detection of melamine were proposed. The selectivity was 4
guaranteed by molecular imprinting technique, and the sensitivity was improved by dual signal amplification method. One was ratiometric fluorescence technique, and the other was using periodic mesoporous silica particles as imprinting matrix. Melamine has been illegally added to food and feed products in order to increase apparent protein content (Jiang, el al., 2011). However, melamine could lead to crystallization and subsequent tissue injury, such as urolithiasis and bladder cancer. Thus, it is very important to establish a rapid and low-cost, sensitive and selective assay for it. So in this work, melamine was selected as the target molecule. The ratiometric fluorescence can be achieved by two ways. One was based on core–shell structured MIPs@QDs (Xu et al., 2015). However, the preparation process for the core-shell structured ratiometric fluorescence sensor involved in complicated two steps, which was time consuming and hard to control. The other way to build ratiometric fluorescence probe was incorporating two kinds of fluorophores, one as target sensitive fluorescence dye and the other as reference dye, into fluorescence probe by one pot synthesis procedure, which can simplify the synthetic process while maintain the sensitivity of ratiometric fluorescence probe. So in the present work, mesoporous structured ratiometric fluorescence MIPs sensor for melamine were prepared by one pot synthesis method using CdTe QDs as target sensitive fluorescence dye, hematoporphyrin (HP) as the reference dye. Compared the traditional MIPs@QDs sensor, the sensitivity of as-prepared mesoporous structured ratiometric fluorescence MIPs sensor were enhanced by the dual signal amplification method and the detection limit was down to nM level. 5
2. Experimental 2.1. Reagents and chemicals Tellurium
powder, Cd(NO3)2•4H2O, thioglycollic acid
(TGA),
sodium
borohydride, melamine, ammeline, cyanuric, atrazine and tetraethyl orthosilicate (TEOS) were purchased from Tianjin Reagent Plant (Tianjin, China). 3-aminopropyl triethoxy silane (APTES), hematoporphyrin (HP), cetyltrimethylammonium bromide (CTAB) and N,N-Dicyclohexylcarbodiimide (DCC, 99%）were purchased from J&K Technology Ltd. (Beijng, China). Double deionized water (DDW) was obtained with a Pall Cascada laboratory water system. All chemicals were of at least analytical grade and used without any purification. 2.2. Characterization Fluorescence
measurements
were
performed
with
a
Fluoromax-4
Spectrofluorometer (Horiba Scientific). The morphological evaluation was examined with a scanning electron microscope (SEM, Hitachi S-4800 FE−SEM, operating at 5 kV) and a transmission electron microscope (TEM, JEM-2100F). Brunauer–Emmett– Teller (BET) surface area was determined by nitrogen adsorption/desorption at 77 K using a Micromeritics ASAP 2020 Sorptometer (Micromeritics, ASAP 2020, USA). HPLC condition employed for melamine detection was: C18 column (250 mm×4.6 mm i.d. ,Waters), mobile phase, ACN/water=10/90 (v/v); flow rate, 1.0 mL min-1; room temperature; UV detection, 236 nm; injection volume, 20 μL. 2.3 One-pot synthesis of mesoporous structured ratiometric fluorescence MIPs sensor CdTe QDs were used as melamine sensitive fluorescence substrate and HP as the 6
reference dye. Firstly, green emissive CdTe QDs were synthesized in aqueous phase by a previous reported method (Xu et al., 2015). HP was modified with APTES by the following step: 13 mg of HP dissolved in 1mL of dimethyl sulfoxide was allowed to react with 5 μL APTES with the help of 0.0045g DCC for overnight in the dark. A typical procedure for preparation of mesoporous structured ratiometric fluorescence MIPs sensor was as the following. 0.7 mL NaOH (2 M) was added to 100 mL of CTAB solution (5.5×10-3 M). The mixture was heated to 70 °C, followed by the addition of 2 mL green QDs solution, 100 μL of HP-APTES, 1.0 mL of TEOS, 200 μL of APTES and 7 mg melamine. After stirring 1 min, 1.0 mL of ethyl acetate was added, and the resulting mixture was stirred at 70 °C for 2 h. After the solution had cooled to room temperature, the precipitate was collected by centrifugation and washed with water and ethanol several times. The control melamine imprinted ratiometric fluorescence MIPs sensor without mesoporous structure was prepared in the same manner but without CTAB. The control melamine imprinted single fluorescence MIPs sensor was prepared using the above procedure, but without adding HP-APTES in the whole procedure. The control non-imprinted ratiometric fluorescence sensor was prepared as described above but without template melamine. Typically, washing with acid methanol and calcination were the commonly used method to remove template for molecular recognition and template CTAB for mesoporous structure. In the present work, considering that the fluorescence intensity of QDs could be quenched obviously after washing with acid ethanol or calcination, so the templates melamine for molecular recognition and template CTAB for 7
mesoporous structure were washed with water/ethanol under ultrasound assistance washing repeatedly. UV, IR and fluorescence spectrums were used to confirm the completely removal of melamine and CTAB. Melamine has strong UV absorbance at 236 nm, the washing solvent was detect by UV until no absorbance was detected at 236 nm. Considering that the low sensitivity of the UV absorption spectra cannot guarantee the complete elimination of the template. So fluorescence spectrum with higher sensitivity was also employed to confirm the completely removal of template. The fluorescence intensity of as-prepared mesoporous structured MIPs@QDs restored to the level of those NIPs can be seen as a symbol of completely removal of template. CTAB has typical IR absorption bands observed in the region 2800–3000 cm-1, which is attributed to the vibrations of -CH2 of CTAB templates. The MIPs or NIPs powder was detected by FT-IR until no adsorption peaks were observed in the range of 2800– 3000 cm-1. 2.4 Detection of melamine The disperse systems were optimized by monitoring the fluorescence stability and sensitivity in different media, including water, water/ethanol mixture and ethanol. The quenching amount, defined as (F0-F)/F0, increased with the increase of ethanol. Therefore, ethanol was selected as the dispersion medium. FL intensity of MIP@QDs were considerably stable in the pH interval between 6 and 9, so the influence of pH values between 6 and 9 on the recognition of melamine by the imprinted silica was examined. The results indicated that quenching amount reach the highest when pH was fixed at 8.0. Further experiments were carried out at pH 8.0. The amounts of 8
MIPs particles had an obvious effect on the quenching efficiency. After optimization using quenching amount as the index of quenching capacity, the concentration of MIPs was fixed at 50 mg/L. Melamine were added into 5 mL of 50 mg/L probe solution one by one with the final concentrations are 100 nM to 900 nM. The fluorescence spectra were collected 20 min after each addition because the fluorescence spectra became stable 15 min after the addition of melamine into the probe solution. 2.5 Reusability of the MIPs The recovery and reusability are an important characteristic of the MIPs based sensor. The removal−rebinding cycles were repeated ten times using the same MIPs or NIPs. During the process, the MIPs/NIPs could be recovered by centrifuging and washing to remove the rebinding melamine with an ethanol/water solvent. The reusable experiments were carried out in successive batches. In the first cycle, the florescence detection was conducted within 20 min. After separated and rinsed with ethanol/water solvent to remove the residual melamine, the MIPs@QDs particles were poured into the refreshed melamine solution and then the detection was continued again for 20 min as the second cycle. This process was repeated for 10 times and the quenching efficiency was recorded. 2.6 Selectivity experiments. In order to test of selectivity of the as-prepared fluorescence MIPs, the fluorescent responses to the other template analogues (ammeline, cyanuric, atrazine) and metal ions (Pb2+, Cu2+, Hg2+) were examined by a similar procedure mentioned 9
above. The concentration of metal ions fixed as 10 μM while the value was 500 nM for template analogues. The solution of Hg2+ was prepared in 0.1 M HNO3 while the solutions of other metal ions were prepared in DDW for the experiments. 2.7 Real sample detection Milk samples were diluted ten-fold using water prior to the analysis. 50 g of sample was mixed with10 mL acetonitrile/water (1:1, v/v) and then placed in a 50 mL beaker. Extraction was performed under stirring at 300 rpm for 12h; acetonitrile precipitated the proteins in this process. Subsequently, the mixture was centrifuged to remove the precipitated proteins. The supernatant were evaporated to dryness under nitrogen and the residues were redissolved in 5 mL ethanol for fluorescence analysis. With the addition of melamine in the real samples, the fluorescence intensity of the ratiometric probe decreased. The relative standard deviation (RSD) was obtained by repeating the experiment 5 times under the same condition. 3. Results and discussion 3.1 One pot preparation of mesoporous structured ratiometric fluorescence MIPs sensor For fabrication of ratiometric fluorescence sensor excited with a single wavelength, the selection of fluorescence dyes is an important step, which should be based on their spectroscopic properties. One fluorescence dye should can response to the target and the other as reference. It has been reported that melamine can quench the fluorescence intensity of CdTe QDs based on charge transfer mechanism (Huy et al. 2014), so CdTe QDs were selected as the target sensitive fluorescence dye. Three 10
requirements should be met by the reference dye: (1) fluorescence reference dye should not response to the target; (2) the emission spectrums of reference dye and target sensitive fluorescence dye can be clearly distinguished; (3) reference dye can be introduced into the polymer simply. According to the requirements mentioned above, HP with the maximum emission wavelength at 620 nm was selected as the reference dye in this work. The synthesis procedure and visual principle of mesoporous structured ratiometric fluorescence MIPs sensor were shown in Figure 1. First, HP was modified with APTES to form HP-APTES by covalent bond, meanwhile template melamine and functional monomer APTES were pre-polymerization to form template-functional monomer complex by hydrogen bond. Subsequently, melamine imprinted ratiometric fluorescence MIPs sensor was prepared by co-condensation reaction between TEOS, HP-APTS and template-functional complex in the presence of CdTe QDs (emission peak centered at 540 nm) by one step. To form the mesoporous structured imprinted silica, CTAB were added during the preparation process. After removal of CTAB for mesoporous silica and melamine for molecular recognition, mesoporous structured melamine imprinted ratiometric fluorescence MIPs were obtained. Yellow-green fluorescence can be observed under a 365 nm UV lamp for the obtained ratiometric fluorescence MIPs with dual emissions at 540 and 620 nm. When detection of melamine, the fluorescence of green CdTe QDs embedded in mesoporous silica can be selectively quenched, whereas HP remains constant. A noticeable fluorescence color change can be observed, which facilitating the visual detection of melamine. 11
Figure 1
The structure and optical properties of the obtained mesoporous structured melamine imprinted ratiometric fluorescence MIPs probe were studied in detail. First the morphology of fluorescence probe were characterized by TEM and SEM. The melamine imprinted (Figure 2A) and non-imprinted (Figure 2B) mesoporous structured silica share the similar diameter (about 220 nm calculated from the TEM photo) and similar morphology, which indicated that template has weal influence on the morphology of mesoporous structured silica. Figure 2D shown that non-mesoporous silica enjoy highly uniform spherical morphology with smooth surfaces. The rough surfaces and those hexagonal tunnels channel (Figure 2E) suggested the existence of mesoporous structure in the silica. For those single fluorescence MIPs (Figure 2C), mesoporous structured also can be seen clearly. The mesoporous structure of MIPs was further confirmed by nitrogen adsorption–desorption experiment (Figure 2F). The Brunauer–Emmett–Teller (BET) surface area obtained from the nitrogen isotherms was 853.27 m2/g and average pore diameter calculated from the adsorption branch of the isotherms was 2.6 nm. The typical type IV isotherm curves with an obvious adsorption hysteresis loop were observed, which indicated well defined porous structure. Figure 2
The fluorescence spectra of CdTe QDs, the ratiometric probes, and HP are 12
characterized in Figure S1. Green fluorescence QDs show fluorescence maximum at 540 nm while red fluorescence HP show fluorescence maximum at 620 nm. When the green emissive QDs and HP incorporated into silica together, the ratiometric fluorescence probe displays well-resolved dual emission bands under a single wavelength excitation at 365 nm and displayed yellow-green fluorescence color. These results all indicate that mesoporous structured ratiometric fluorescence probe were successfully prepared. 3.2 Analytical performance of mesoporous structured ratiometric fluorescence MIPs The dose response of the mesoporous structured ratiometric fluorescence MIPs to melamine were examined. Firstly, variables such as the dispersion medium, response time, pH and amount of MIPs which may affect the fluorescence-quenching efficiency, were optimized using quenching amount, defined as (F0-F)/F0, as the index. Experiment results indicated that ethanol was the ideal dispersion medium, 20 min was enough for response time, pH 8.0 was suitable and the ideal concentration of MIPs was 50 mg/L. It should be noted that mesoporous structured fluorescence probes with equilibrium adsorption time 15 min had faster response than non-mesoporous structured MIPs with equilibrium adsorption time 30 min, which can be attributed to the advantage of mesoporous structure. The doping ratio of two dyes had importance influence to the ratiometric detection, so MIPs with different doping ratios were prepared. The experimental results indicated that both the linear range and detection limit varies with variation of the target sensitive dye proportion. When QDs were a large proportion of the total dye, 13
the linear range was limited because the strong intensity of QDs masked the emission of reference HPs. In the meanwhile, overmuch increase of HP also narrowed the linear range because the decrease of QDs led to low intensity of QDs. In the present work, the probes are the most sensitive when 2 mL green QDs solution were added while 100 μL of HP-APTES was used as reference. In the absence of melamine, the ratiometric probe emited two well-resolved emission peaks centered at 540 and 620 nm. Upon the addition of melamine, the fluorescence intensity at 540 nm was continuously quenched while the intensity at 620 nm still remained unchanged, as shown in Figure 3A. The ratio of the fluorescence intensity was closely related to the amount of melamine ranging from 100 nM to 800 nM, which can be used for the quantification of melamine with the detection limit as low as 38 nM (3S/N). Owing to the changes in the intensity ratio of the two dyes, the fluorescence colors of the ratiometric probe solution changed continuously as demonstrated in the inset of Figure 3A. A slight decrease of the emission intensity at 540 nm could result in distinguishable color changes from the original background. Therefore, the visual detection of melamine by the naked eye is feasible. The dose response of the mesoporous structured ratiometric NIPs to melamine also has been examined. From Figure 3B we can see that fluorescence intensity of NIPs also can be quenched by melamine. However, the sensitivity was much lower and the linear range was narrower, meanwhile the fluorescence color change was not obviously. This phenomenon can be explained as no recognition sites existed in the 14
NIPs, so melamine cannot enter into the inner of the NIPs. Only the fluorescence intensity of those QDs located on the surface of materials can be quenched while the most of QDs keep constant. Figure 3.
Meanwhile, the advantages of the ratiometric fluorescence probe was verified by comparison with single fluorescence probes, in which only QDs were embedded. As can be seen from Figure 3C, the color changes of the single fluorescence probe upon the addition of melamine were hard to observe, meanwhile the linear range was narrow. The comparison clearly showed that ratiometric fluorescence probe possessed higher sensitivity and reliability than single fluorescence quenching probe for visual detection. In order to verify the advantage of mesoporous structured ratiometric fluorescence MIPs, non-mesoporous structured ratiometric fluorescence MIPs were prepared as control. The linear range 5−45 μM and the detection limit of 1.5 μM were attained for those non-mesoporous structured ratiometric fluorescence MIPs, which can much higher than those mesoporous structured one. For the mesoporous imprinted silica with higher surface area, most of the recognition sites were located on the surface of silica matrix, so target molecules have more chance to enter the recognition site, and to quench the fluorescence intensity of QDs. So higher surface area and better accessibility for the binding sites are the reason for the better sensitivity. The reusability of the MIPs are important because synthesis of the material is 15
laborious and time consuming. The life time of mesoporous structured ratiometric fluorescence MIPs was evaluated by using it for 500 nM melamine detection while monitoring their fluorescence intensity and quenching efficiency for 10 times. Figure S2 showed that the sensor could retain its fluorescence intensity and detection sensitivity without obvious drop (RSD=5.9) during those ten recycles. 3.3 Molecular selectivity and sensing specificity of the sensor. The selectivity of mesoporous structured ratiometric fluorescence MIPs and NIPs were demonstrated by single component measurement. The fluorescence intensity ratios of the probe were recorded (I540/I620) in the presence of melamine and other template analogues (ammeline, cyanuric, atrazine) and metal ions (Pb2+, Cu2+, Hg2+) . It can be seen from Figure 4 that 60% of the fluorescence intensity of the MIPs solution was quenched by 500 nM melamine, accompanied by remarkable fluorescence color changes under UV light. In contrast, no obvious changes about fluorescence intensity ratio and color can be observed when detection of template analogues and metal ions. The results implied the high selectivity of the imprinted ratiometric fluorescence probe. It is well known that some metal ions, such as Cu2+, Hg2+ also can quenching the fluorescence intensity of CdTe QDs. Admittedly the diameters of metal ions are lower than melamine, it is easy for them entering into recognition sites. However, there is no specific interaction between functional monomer and metal ions in the recognition sites, and the hole is not complementary to metal ions in shape and size. So metal ions can desorption from the recognition sites easily. So the quenching resulted from metal ions can be neglected. Competitive 16
molecules including ammeline, cyanuric and atrazine also caused negligence quenching efficiency when compared to melamine. The phenomenon can be explained as the recognition sites of melamine imprinted cavities were not complementary to them, resulting in less chance to quench the fluorescence of the QDs. Figure 4
3.4 Practical application to real sample analysis To investigate the practical application of this method in real samples, the sensing of melamine in milk samples was carried out. Melamine does not be detected in the milk powder and row milk samples. Then, spiked experiments were carried out. The estimated recoveries of the measurements and the RSD are satisfactory (shown in Table 1) and the corresponding fluorescence spectra of the ratiometric probe upon the exposure to different real samples were displayed in Figure S3. The recoveries of melamine for real samples are statistically close to those values added. And the value is much higher agree with obtained using the HPLC method. The comparison shows that this sensing method has promising potential for use in melamine detection. Table 1
Analytical performances of the developed methods for detection of melamine were mainly compared with the previously methods including MIP@QDs sensors (Huy et al., 2014), MIP-SPE-HPLC (Zhang et al., 2011; Cheng et al., 2013), MIP/ 17
electrochemical sensor (Xu et al., 2014), MIP/SPR optic sensors (Shrivastav et al., 2015) and HPLC-MS-MS (Wu et al., 2009), which were summarized in Table 2. Without expensive or sophisticated instrumentation, extensive sample preparation, or the need for highly skilled personnel, the LOD obtained in the present work is on same order as those of MIP electrochemical sensor, but much lower than those of MIP-SPE-HPLC method and MIP@QD sensor without mesoporous structure. Although the lower detection for melamine was achieved by MIP@SPR sensors, it needs elaborated and sophisticated instrumentation. When compared to HPLC method, the presented work do not need time consuming sample pretreatment process and expensive instrument. Compared to MIP@QDs without mesoporous structure, the mesoporous structured MIP@QDs sensor displayed higher sensitivity. Our proposed method is sensitive enough and highly selective for practical uses to determine melamine in various matrices. The developed method requires no sample pretreatments including extraction and derivatization and would be a promising alternative for analyzing melamine quantitatively in food sample. Table 2
4 Conclusion In summary, melamine imprinted mesoporous structured ratiometric fluorescence probes have been constructed by one pot synthesis method, and its utility has been established for detection of melamine in real milk samples. The novel fluorescence MIPs sensor integrated the high selectivity of MIPs and the high sensitivity of 18
ratiometric fluorescence probe and mesoporous silica into one system. The mesoporous structured radiometric fluorescence MIP@QDs probes demonstrated prominence advantages over determination of melamine from real milk samples. The one pot synthesis method confirmed to be simple, rapid, and reliable to build mesoporous structured imprinted ratiometric fluorescence probe. In the meanwhile, the method presented here opens up attractive perspectives for melamine monitoring, and it would be an universe method for other substance analysis.
Acknowledgements We thanks NSFC (NO. 21307052), the Natural Science Foundation of Shandong Province of China (ZR2013BL006).
References Chen, L.X., Xu, S.F., Li, J.X., 2011.Chem. Soc. Rev. 40, 2922-2942. Chen, Y., Wang, D., Yin, Y., Wang, L., Wang, X., Xie, M., 2012. J. Agric. Food Chem. 60, 10472-10479. Cheng, W., Liu, Z., Wang, Y., 2013. Talanta 116, 396–402. Fang, Q., Geng, J., Liu, B., 2009. Chem.-Eur. J. 15, 11507-11514. Huy, B., Seo, M., Zhang, X., Lee, Y., 2014. Biosens. Bioelectron. 57, 310-316. Jiang, Z., Zhou, L., Liang, A., 2011. Chem. Commun. 47, 3162-3164. Kim, H., Ren, W., Kim, J., Yoon, J., 2012. Chem. Soc. Rev. 41, 3210-3244. Li, D., He, X., Chen, Y., Li, W., Zhang, Y., 2013. ACS Appl. Mater. Interfaces. 5, 12609-12616. 19
Liu, J., Chen, H., Lin, Z., Lin, J. 2010. Anal. Chem. 82, 7380-7386. Petya, K., Ivanova, M., Antonio, G., Elena, V., Piletska, Michael, Whitcombe, J., Zhou, Z.X., Petar, A., Mitsev, Frank, D. and Sergey, A., Piletsky, 2012. Angew. Chem. Int. Ed. 51, 5196-5199. Ren, X., Chen, L., 2015. Biosens. Bioelectron. 64, 182-188. Selim, B., Serena, A., Nataliya, M., Emilia, P.,Vishal, K.,Tero, S., Bernadette, T.S.B., Karsten, H., 2014. Angew. Chem. Int. Ed. 53, 8919-8923. Shen, X.T., Johan S. B., Tripta, K., Leif B., Jack C. L., Dirk L. and Ye, L. , 2014. Angew. Chem. Int. Ed. 53, 10687-10690. Shrivastav, A., Mishra, S., Gupta, B., 2015, Sensors and Actuators B 212, 404–410. Tan, J., Wang, H., Yan, X. P., 2009. Anal. Chem. 81, 5273-5280. Tan, J., Wang, H., Yan, X.P., 2009. Biosens. Bioelectron. 24, 3316-3321. Wagner, R., Wan, W., Biyikal, M., Benito-Peña, E., Moreno-Bondi, M., Lazraq, I., Rurack, K., Sellergren, B., 2013. J. Org. Chem. 78, 1377-1389. Wan, W., Mustafa, B., Ricarda, W., Bçrje, S., Knut, R., 2013. Angew. Chem. Int. Ed. 52, 7023-7027. Wu, Y.; Huang, C.; Lin, C.; Ho, W.; Lin, L.; Chiu, T.; Tarng, D.; Lin, C.; Tsai, T., 2009. J. Chromatogr. A 1216, 7595–7601. Wu, Y., Liu, Y., Gao, X., Gao, K., Xia, H., Luo, M., Wang, X., Ye, L., Shi, Y., Lu, B., 2015. Chemosphere 119, 515-523. Xu, G., Zhang, H., Zhong, M., Zhang, T., Lu, X., Kan, X., 2014. J. Electroanal. Chem. 713, 112–118. Xu, S.F., Lu, H.Z., 2015. Chem. Commun. 51, 3200-3203. 20
Xu, S.F., Lu, H.Z., Li, J.H., Song, X.L., Wang, A.X., Chen, L.X., Han, S.B., 2013. ACS Appl. Mater. Interfaces. 5, 8146-8154. Yao, J., Zhang, K., Zhu, H., Ma, F., Sun, M., Yu, H., Sun, J., Wang, S., 2013. Anal. Chem. 85, 6461-6468. Yuan, L., Lin, W., Zheng, K., He, L. and Huang, W., 2013. Chem. Soc. Rev. 42, 622-661. Zhang, H., Zhang, Z., Hu, Y., Yang, X., Yao, S., 2011. J. Agric. Food Chem. 59, 1063– 1071. Zhang, K., Zhou, H., Mei, Q., Wang, S., Guan, G., Liu, R., Zhang, J., 2011. J. Am. Chem. Soc. 133, 8424-8427. Zhang, W., He, X., Chen, Y., Li, W., Zhang, Y., 2011. Biosens. Bioelectron. 26, 2553-2558. Zhao, Y., Ma, Y., Li, H., Wang, L., 2012. Anal. Chem. 84, 386-395. Figure captions Figure 1 Schematic illustration for the one pot preparation of mesoporous sturctured ratiometric fluorescence MIPs probe. Figure 2 TEM images of (A) mesoporous structured ratiometric fluorescence MIPs; (B) mesoporous structured ratiometric fluorescence NIPs; (C) mesoporous structured single fluorescence MIPs; (D) non-mesoporous structured ratiometric fluorescence MIPs; (E) the high resolution TEM for mesoporous structured ratiometric fluorescence MIPs; (F) N2 sorption isotherms and the pore size distribution for mesoporous structured ratiometric fluorescence MIPs. The insets were SEM images for the corresponding TEM images. Figure 3. The fluorescence colors and the corresponding fluorescence spectra (λex = 490 nm) of the ratiometric probe upon the exposure to different concentrations of melamine. A was the mesoporous stuctured ratiometric fluorescence MIPs; B was mesoprous stuctured ratiometric fluorescence NIPs; C was mesoprous stuctured single fluorescence MIPs and D was 21
non-mesoprous stuctured ratiometric fluorescence MIPs. The experimental conditions were MIPs or NIPs 50 mg/L; excited light, 490 nm; silt widths of excitation and emission, 5 and 3 nm, respectively. Figure 4 The selectivity of the mesoporous structured ratiometric fluorescence MIPs and NIPs to other template analogues and metal ions. The experimental conditions were MIPs or NIPs, 50 mg/L; excited light, 490 nm; silt widths of excitation and emission, 5 and 3 nm, respectively. The concentration of metal ions fixed as 10 μM while the value was 500 nM for template analogues. The inset images show the corresponding fluorescence colors under a UV lamp (from left to right were: cyanuric, ammeline, atrazine, melamine, Cu2+, Hg2+, Pb2+).
Figure 1
Figure 2
22
Figure 3
23
Figure 4
NH2
OH N HO
N
N N
OH
cyanuric
HO
Cl N
N N
NH2
HN
N
atrazine
ammeline
NH2
N
N NH H 2N
N N
NH2
melamine
Table 1 Spiked recoveries and relative standard deviations for detection of melamine in spiked raw milk and milk powder samples by ratiometric fluorescence MIPs 24
probe and HPLC.
Sample
Recovery, mean±SD (%)
Spiked
Raw milk
Milk powder
(nM)
Ratiometric probe
HPLC
50
92.7±3.4
94.2±1.9
100
98.2±2.6
97.5±2.9
200
101.3±3.7
99.8±3.2
50
95.4±5.2
96.8±4.6
100
96.8±3.1
97.1±2.8
200
97.4±3.4
97.5±2.6
Table 2 Comparison of determination of melamine to other methods
a
Methods
LOD
Matrix
Ref.
SPEa HPLC-MS-MS
4.2 μg/L
Milk
Wu el al.,2009
MIP@SPRb sensor
9.87nM
Milk
Shrivastav el al., 2015
MIP electrochemical sensor
68 nM
Raw milk
MIP@CNT c -SPE-HPLC
0.3 μg/L
Milk powder
Zhang el al., 2011
MIP@SiO2-SPE-HPLC
NGd
Milk powder
Cheng el al., 2013
MIP@QDs
75 μg/L
Milk drink
Huy et al., 2014
Mesoporous MIP@QDs
38 nM
Milk powder
This work
SPE: solid phase extraction; nanotube; d NG: not given.
b
SPR: surface plasmon resonance;
25
Xu el al.,2014
c
CNT: multiwalled carbon
Highlights
Mesoporous structured ratiometric fluorescence MIPs sensor was prepared. One pot synthesis method was employed. Mesoporous structured ratiometric fluorescence MIPs sensor displayed high sensitivity and high selectivity.
Mesoporous structured ratiometric fluorescence MIPs sensor has potentially applicable for trace pollutants analysis in complicated samples
26
PII: DOI: Reference:
S0956-5663(15)30162-7 http://dx.doi.org/10.1016/j.bios.2015.05.064 BIOS7728
To appear in: Biosensors and Bioelectronic Received date: 31 March 2015 Revised date: 19 May 2015 Accepted date: 28 May 2015 Cite this article as: Shoufang Xu and Hongzhi Lu, One-pot synthesis of mesoporous structured ratiometric fluorescence Molecularly imprinted sensor for highly sensitive detection of melamine from milk samples, Biosensors and Bioelectronic, http://dx.doi.org/10.1016/j.bios.2015.05.064 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 galley proof before it is published in its final citable 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.
One-pot synthesis of mesoporous structured ratiometric fluorescence molecularly imprinted sensor for highly sensitive detection of melamine from milk samples Shoufang Xu*, Hongzhi Lu School of Chemistry and Chemical Engineering, Linyi University, Linyi 276005, China. *Corresponding author Tel.: +86 539 8766000; fax: +86 539 8766000. E-mail address: [email protected] (S.F. Xu)
Abstract: A facile strategy was developed to prepare mesoporous structured ratiometric fluorescence molecularly imprinted sensor for highly sensitive and selective determination of melamine using CdTe QDs as target sensitive dye and hematoporphyrin as reference dyes. One-pot synthesis method was employed because it could simplify the imprinting process and shorten the experimental period. The as-prepared fluorescence MIPs sensor, which combined ratiometric fluorescence technique with mesoporous silica materials into one system, exhibited excellent selectivity and sensitivity. Under optimum conditions, these mesoporous structured ratiometric fluorescence MIP@QDs sensors showed detection limit as low as 38 nM, which was much lower than those non-mesoporous one. The recycling process was sustainable at least 10 times without obvious efficiency decrease. The feasibility of 1
the developed method in real samples was successfully evaluated through the analysis of melamine in raw milk and milk powder samples with satisfactory recoveries of 92– 101%. The developed method proposed in this work proved to be a convenient, rapid, reliable and practical way to prepared high sensitive and selective fluorescence sensors with potentially applicable for trace pollutants analysis in complicated samples.
Keywords: Molecularly imprinted polymers; Ratiometric fluorescence probe; mesoporous silica; Melamine; one-pot synthesis method 1. Introduction Molecularly imprinted polymers (MIPs) are synthetic polymeric materials with specific recognition sites complementary in shape, size, and functional groups to template molecules (Chen et al., 2011; Shen et al., 2014). Fluorescence sensors show the potential application in detection trace amounts of analytes because of their sensitivity, simplicity, and cost-effective instrumentation (Yuan et al., 2013; Kim et al., 2012). Combined the high sensitivity of fluorescence sensor with the high selectivity of MIPs, fluorescence MIPs sensors displayed apparent advantages when detection of trace substances in complicated matrix (Ivanova-Mitseva et al. 2012; Beyazit et al., 2014). Up to date, two methods were widely used to prepare fluorescence MIPs sensors. One was using fluorescence functional monomer for molecular imprinting. For example, Sellergren group (Wagner et al., 2013; Wan et al. 2013) prepared fluorescence MIPs sensors using naphthalimide-based fluorescent monomers, which 2
were equipped with a urea binding site to carboxylate-containing guests. Wu (Wu et al., 2015) prepared fluorescent MIPs using dansyl methacrylate as functional monomer,
which
displayed
selective,
concentration-dependent
fluorescence
quenching in response to bisphenol A. However, this method often involved in complicated synthesis procedure for fluorescence functional monomer. The other, also the most commonly used method for proposing fluorescence MIPs sensor was encapsulating semiconductor quantum dots (QDs) into MIPs. A number of MIPs@QDs sensors have been developed to detect various compounds. For example, MIPs coated CdTe QDs sensor have been used for detection of cytochrome c (Zhang et al., 2011), bovine hemoglobin (Li et al., 2013), TNT (Xu et al., 2013), clenbuterol and melamine (Huy et al., 2014); MIPs coated Mn-ZnS QDs sensors for 4nonylphenol (Liu et al., 2010), cyphenothrin (Ren et al., 2015), tetrabromobisphenol A (Chen et al., 2012), pesticides (Zhao et al., 2012), et.al also have been synthesized. This method of directly doping fluorescent substance into MIPs enjoys the advantages of simple synthesis procedure. However, the sensitivity of the MIPs@QDs sensors often be weakened. The reason would be the following. First, QDs were not uniformly embedded into highly cross-linked MIPs, so it is difficult to guarantee that each imprinted cavity was surrounded by fluorescent QDs. In the meanwhile, the morphology of MIPs@QDs sensor was often solid sphere or irregular bulk particles, which was not beneficial for mass transfer. The second reason may be that the mostly reported detection modes of MIPs@QDs sensors were based on signal “turn-off” attributed to fluorescent quenching. It was a common accepted fact that fluorescence 3
sensors based on signal “turn-on” were often able to achieve lower detection limits due to the lower optical background while ratiometric fluorescence sensor always have wide linear range and weaken interference. Considering that relative lower sensitivity hinder the application of fluorescence MIPs sensors, developing MIPs@QDs sensors with high sensitivity and selectivity were highly demanded. The morphology of sensor is significant for highly sensitive identification. So constructing hierarchical architectures for sensor, such as mesoporous structure, highly ordered 3D inverted opal structures or nanowire is an effective strategy to enhance the sensitivity of the sensor. For example, Zhang’s group constructed fluorescence sensor with high sensitivity for detection of explosives based on inverse opal structure (Fang et al., 2009); Yan’s group build mesoporous structured fluorescent MIPs sensor for highly sensitive detection of metal ions (Tan et al., 2009) and carbohydrates molecules (Tan et al., 2009). The results indicated that interconnected mesoporous structures can greatly decrease mass-transport resistance and provide easier accessibility to the target species. These features make mesoporous silica particularly suitable for the applications in chemo/biosensors. Ratiometric fluorescence technique has attracted increasing attention owing to its improved sensitivity. Ratiometric fluorescence QDs probes for copper ions (Yao et al., 2013) and TNT (Zhang et al., 2011) confirmed that ratiometric fluorescence QDs probes can achieve high sensitivity and visual signal output observed by the naked eye. Considering the above mentioned, in this paper, fluorescence MIPs sensor for high sensitive and selective detection of melamine were proposed. The selectivity was 4
guaranteed by molecular imprinting technique, and the sensitivity was improved by dual signal amplification method. One was ratiometric fluorescence technique, and the other was using periodic mesoporous silica particles as imprinting matrix. Melamine has been illegally added to food and feed products in order to increase apparent protein content (Jiang, el al., 2011). However, melamine could lead to crystallization and subsequent tissue injury, such as urolithiasis and bladder cancer. Thus, it is very important to establish a rapid and low-cost, sensitive and selective assay for it. So in this work, melamine was selected as the target molecule. The ratiometric fluorescence can be achieved by two ways. One was based on core–shell structured MIPs@QDs (Xu et al., 2015). However, the preparation process for the core-shell structured ratiometric fluorescence sensor involved in complicated two steps, which was time consuming and hard to control. The other way to build ratiometric fluorescence probe was incorporating two kinds of fluorophores, one as target sensitive fluorescence dye and the other as reference dye, into fluorescence probe by one pot synthesis procedure, which can simplify the synthetic process while maintain the sensitivity of ratiometric fluorescence probe. So in the present work, mesoporous structured ratiometric fluorescence MIPs sensor for melamine were prepared by one pot synthesis method using CdTe QDs as target sensitive fluorescence dye, hematoporphyrin (HP) as the reference dye. Compared the traditional MIPs@QDs sensor, the sensitivity of as-prepared mesoporous structured ratiometric fluorescence MIPs sensor were enhanced by the dual signal amplification method and the detection limit was down to nM level. 5
2. Experimental 2.1. Reagents and chemicals Tellurium
powder, Cd(NO3)2•4H2O, thioglycollic acid
(TGA),
sodium
borohydride, melamine, ammeline, cyanuric, atrazine and tetraethyl orthosilicate (TEOS) were purchased from Tianjin Reagent Plant (Tianjin, China). 3-aminopropyl triethoxy silane (APTES), hematoporphyrin (HP), cetyltrimethylammonium bromide (CTAB) and N,N-Dicyclohexylcarbodiimide (DCC, 99%）were purchased from J&K Technology Ltd. (Beijng, China). Double deionized water (DDW) was obtained with a Pall Cascada laboratory water system. All chemicals were of at least analytical grade and used without any purification. 2.2. Characterization Fluorescence
measurements
were
performed
with
a
Fluoromax-4
Spectrofluorometer (Horiba Scientific). The morphological evaluation was examined with a scanning electron microscope (SEM, Hitachi S-4800 FE−SEM, operating at 5 kV) and a transmission electron microscope (TEM, JEM-2100F). Brunauer–Emmett– Teller (BET) surface area was determined by nitrogen adsorption/desorption at 77 K using a Micromeritics ASAP 2020 Sorptometer (Micromeritics, ASAP 2020, USA). HPLC condition employed for melamine detection was: C18 column (250 mm×4.6 mm i.d. ,Waters), mobile phase, ACN/water=10/90 (v/v); flow rate, 1.0 mL min-1; room temperature; UV detection, 236 nm; injection volume, 20 μL. 2.3 One-pot synthesis of mesoporous structured ratiometric fluorescence MIPs sensor CdTe QDs were used as melamine sensitive fluorescence substrate and HP as the 6
reference dye. Firstly, green emissive CdTe QDs were synthesized in aqueous phase by a previous reported method (Xu et al., 2015). HP was modified with APTES by the following step: 13 mg of HP dissolved in 1mL of dimethyl sulfoxide was allowed to react with 5 μL APTES with the help of 0.0045g DCC for overnight in the dark. A typical procedure for preparation of mesoporous structured ratiometric fluorescence MIPs sensor was as the following. 0.7 mL NaOH (2 M) was added to 100 mL of CTAB solution (5.5×10-3 M). The mixture was heated to 70 °C, followed by the addition of 2 mL green QDs solution, 100 μL of HP-APTES, 1.0 mL of TEOS, 200 μL of APTES and 7 mg melamine. After stirring 1 min, 1.0 mL of ethyl acetate was added, and the resulting mixture was stirred at 70 °C for 2 h. After the solution had cooled to room temperature, the precipitate was collected by centrifugation and washed with water and ethanol several times. The control melamine imprinted ratiometric fluorescence MIPs sensor without mesoporous structure was prepared in the same manner but without CTAB. The control melamine imprinted single fluorescence MIPs sensor was prepared using the above procedure, but without adding HP-APTES in the whole procedure. The control non-imprinted ratiometric fluorescence sensor was prepared as described above but without template melamine. Typically, washing with acid methanol and calcination were the commonly used method to remove template for molecular recognition and template CTAB for mesoporous structure. In the present work, considering that the fluorescence intensity of QDs could be quenched obviously after washing with acid ethanol or calcination, so the templates melamine for molecular recognition and template CTAB for 7
mesoporous structure were washed with water/ethanol under ultrasound assistance washing repeatedly. UV, IR and fluorescence spectrums were used to confirm the completely removal of melamine and CTAB. Melamine has strong UV absorbance at 236 nm, the washing solvent was detect by UV until no absorbance was detected at 236 nm. Considering that the low sensitivity of the UV absorption spectra cannot guarantee the complete elimination of the template. So fluorescence spectrum with higher sensitivity was also employed to confirm the completely removal of template. The fluorescence intensity of as-prepared mesoporous structured MIPs@QDs restored to the level of those NIPs can be seen as a symbol of completely removal of template. CTAB has typical IR absorption bands observed in the region 2800–3000 cm-1, which is attributed to the vibrations of -CH2 of CTAB templates. The MIPs or NIPs powder was detected by FT-IR until no adsorption peaks were observed in the range of 2800– 3000 cm-1. 2.4 Detection of melamine The disperse systems were optimized by monitoring the fluorescence stability and sensitivity in different media, including water, water/ethanol mixture and ethanol. The quenching amount, defined as (F0-F)/F0, increased with the increase of ethanol. Therefore, ethanol was selected as the dispersion medium. FL intensity of MIP@QDs were considerably stable in the pH interval between 6 and 9, so the influence of pH values between 6 and 9 on the recognition of melamine by the imprinted silica was examined. The results indicated that quenching amount reach the highest when pH was fixed at 8.0. Further experiments were carried out at pH 8.0. The amounts of 8
MIPs particles had an obvious effect on the quenching efficiency. After optimization using quenching amount as the index of quenching capacity, the concentration of MIPs was fixed at 50 mg/L. Melamine were added into 5 mL of 50 mg/L probe solution one by one with the final concentrations are 100 nM to 900 nM. The fluorescence spectra were collected 20 min after each addition because the fluorescence spectra became stable 15 min after the addition of melamine into the probe solution. 2.5 Reusability of the MIPs The recovery and reusability are an important characteristic of the MIPs based sensor. The removal−rebinding cycles were repeated ten times using the same MIPs or NIPs. During the process, the MIPs/NIPs could be recovered by centrifuging and washing to remove the rebinding melamine with an ethanol/water solvent. The reusable experiments were carried out in successive batches. In the first cycle, the florescence detection was conducted within 20 min. After separated and rinsed with ethanol/water solvent to remove the residual melamine, the MIPs@QDs particles were poured into the refreshed melamine solution and then the detection was continued again for 20 min as the second cycle. This process was repeated for 10 times and the quenching efficiency was recorded. 2.6 Selectivity experiments. In order to test of selectivity of the as-prepared fluorescence MIPs, the fluorescent responses to the other template analogues (ammeline, cyanuric, atrazine) and metal ions (Pb2+, Cu2+, Hg2+) were examined by a similar procedure mentioned 9
above. The concentration of metal ions fixed as 10 μM while the value was 500 nM for template analogues. The solution of Hg2+ was prepared in 0.1 M HNO3 while the solutions of other metal ions were prepared in DDW for the experiments. 2.7 Real sample detection Milk samples were diluted ten-fold using water prior to the analysis. 50 g of sample was mixed with10 mL acetonitrile/water (1:1, v/v) and then placed in a 50 mL beaker. Extraction was performed under stirring at 300 rpm for 12h; acetonitrile precipitated the proteins in this process. Subsequently, the mixture was centrifuged to remove the precipitated proteins. The supernatant were evaporated to dryness under nitrogen and the residues were redissolved in 5 mL ethanol for fluorescence analysis. With the addition of melamine in the real samples, the fluorescence intensity of the ratiometric probe decreased. The relative standard deviation (RSD) was obtained by repeating the experiment 5 times under the same condition. 3. Results and discussion 3.1 One pot preparation of mesoporous structured ratiometric fluorescence MIPs sensor For fabrication of ratiometric fluorescence sensor excited with a single wavelength, the selection of fluorescence dyes is an important step, which should be based on their spectroscopic properties. One fluorescence dye should can response to the target and the other as reference. It has been reported that melamine can quench the fluorescence intensity of CdTe QDs based on charge transfer mechanism (Huy et al. 2014), so CdTe QDs were selected as the target sensitive fluorescence dye. Three 10
requirements should be met by the reference dye: (1) fluorescence reference dye should not response to the target; (2) the emission spectrums of reference dye and target sensitive fluorescence dye can be clearly distinguished; (3) reference dye can be introduced into the polymer simply. According to the requirements mentioned above, HP with the maximum emission wavelength at 620 nm was selected as the reference dye in this work. The synthesis procedure and visual principle of mesoporous structured ratiometric fluorescence MIPs sensor were shown in Figure 1. First, HP was modified with APTES to form HP-APTES by covalent bond, meanwhile template melamine and functional monomer APTES were pre-polymerization to form template-functional monomer complex by hydrogen bond. Subsequently, melamine imprinted ratiometric fluorescence MIPs sensor was prepared by co-condensation reaction between TEOS, HP-APTS and template-functional complex in the presence of CdTe QDs (emission peak centered at 540 nm) by one step. To form the mesoporous structured imprinted silica, CTAB were added during the preparation process. After removal of CTAB for mesoporous silica and melamine for molecular recognition, mesoporous structured melamine imprinted ratiometric fluorescence MIPs were obtained. Yellow-green fluorescence can be observed under a 365 nm UV lamp for the obtained ratiometric fluorescence MIPs with dual emissions at 540 and 620 nm. When detection of melamine, the fluorescence of green CdTe QDs embedded in mesoporous silica can be selectively quenched, whereas HP remains constant. A noticeable fluorescence color change can be observed, which facilitating the visual detection of melamine. 11
Figure 1
The structure and optical properties of the obtained mesoporous structured melamine imprinted ratiometric fluorescence MIPs probe were studied in detail. First the morphology of fluorescence probe were characterized by TEM and SEM. The melamine imprinted (Figure 2A) and non-imprinted (Figure 2B) mesoporous structured silica share the similar diameter (about 220 nm calculated from the TEM photo) and similar morphology, which indicated that template has weal influence on the morphology of mesoporous structured silica. Figure 2D shown that non-mesoporous silica enjoy highly uniform spherical morphology with smooth surfaces. The rough surfaces and those hexagonal tunnels channel (Figure 2E) suggested the existence of mesoporous structure in the silica. For those single fluorescence MIPs (Figure 2C), mesoporous structured also can be seen clearly. The mesoporous structure of MIPs was further confirmed by nitrogen adsorption–desorption experiment (Figure 2F). The Brunauer–Emmett–Teller (BET) surface area obtained from the nitrogen isotherms was 853.27 m2/g and average pore diameter calculated from the adsorption branch of the isotherms was 2.6 nm. The typical type IV isotherm curves with an obvious adsorption hysteresis loop were observed, which indicated well defined porous structure. Figure 2
The fluorescence spectra of CdTe QDs, the ratiometric probes, and HP are 12
characterized in Figure S1. Green fluorescence QDs show fluorescence maximum at 540 nm while red fluorescence HP show fluorescence maximum at 620 nm. When the green emissive QDs and HP incorporated into silica together, the ratiometric fluorescence probe displays well-resolved dual emission bands under a single wavelength excitation at 365 nm and displayed yellow-green fluorescence color. These results all indicate that mesoporous structured ratiometric fluorescence probe were successfully prepared. 3.2 Analytical performance of mesoporous structured ratiometric fluorescence MIPs The dose response of the mesoporous structured ratiometric fluorescence MIPs to melamine were examined. Firstly, variables such as the dispersion medium, response time, pH and amount of MIPs which may affect the fluorescence-quenching efficiency, were optimized using quenching amount, defined as (F0-F)/F0, as the index. Experiment results indicated that ethanol was the ideal dispersion medium, 20 min was enough for response time, pH 8.0 was suitable and the ideal concentration of MIPs was 50 mg/L. It should be noted that mesoporous structured fluorescence probes with equilibrium adsorption time 15 min had faster response than non-mesoporous structured MIPs with equilibrium adsorption time 30 min, which can be attributed to the advantage of mesoporous structure. The doping ratio of two dyes had importance influence to the ratiometric detection, so MIPs with different doping ratios were prepared. The experimental results indicated that both the linear range and detection limit varies with variation of the target sensitive dye proportion. When QDs were a large proportion of the total dye, 13
the linear range was limited because the strong intensity of QDs masked the emission of reference HPs. In the meanwhile, overmuch increase of HP also narrowed the linear range because the decrease of QDs led to low intensity of QDs. In the present work, the probes are the most sensitive when 2 mL green QDs solution were added while 100 μL of HP-APTES was used as reference. In the absence of melamine, the ratiometric probe emited two well-resolved emission peaks centered at 540 and 620 nm. Upon the addition of melamine, the fluorescence intensity at 540 nm was continuously quenched while the intensity at 620 nm still remained unchanged, as shown in Figure 3A. The ratio of the fluorescence intensity was closely related to the amount of melamine ranging from 100 nM to 800 nM, which can be used for the quantification of melamine with the detection limit as low as 38 nM (3S/N). Owing to the changes in the intensity ratio of the two dyes, the fluorescence colors of the ratiometric probe solution changed continuously as demonstrated in the inset of Figure 3A. A slight decrease of the emission intensity at 540 nm could result in distinguishable color changes from the original background. Therefore, the visual detection of melamine by the naked eye is feasible. The dose response of the mesoporous structured ratiometric NIPs to melamine also has been examined. From Figure 3B we can see that fluorescence intensity of NIPs also can be quenched by melamine. However, the sensitivity was much lower and the linear range was narrower, meanwhile the fluorescence color change was not obviously. This phenomenon can be explained as no recognition sites existed in the 14
NIPs, so melamine cannot enter into the inner of the NIPs. Only the fluorescence intensity of those QDs located on the surface of materials can be quenched while the most of QDs keep constant. Figure 3.
Meanwhile, the advantages of the ratiometric fluorescence probe was verified by comparison with single fluorescence probes, in which only QDs were embedded. As can be seen from Figure 3C, the color changes of the single fluorescence probe upon the addition of melamine were hard to observe, meanwhile the linear range was narrow. The comparison clearly showed that ratiometric fluorescence probe possessed higher sensitivity and reliability than single fluorescence quenching probe for visual detection. In order to verify the advantage of mesoporous structured ratiometric fluorescence MIPs, non-mesoporous structured ratiometric fluorescence MIPs were prepared as control. The linear range 5−45 μM and the detection limit of 1.5 μM were attained for those non-mesoporous structured ratiometric fluorescence MIPs, which can much higher than those mesoporous structured one. For the mesoporous imprinted silica with higher surface area, most of the recognition sites were located on the surface of silica matrix, so target molecules have more chance to enter the recognition site, and to quench the fluorescence intensity of QDs. So higher surface area and better accessibility for the binding sites are the reason for the better sensitivity. The reusability of the MIPs are important because synthesis of the material is 15
laborious and time consuming. The life time of mesoporous structured ratiometric fluorescence MIPs was evaluated by using it for 500 nM melamine detection while monitoring their fluorescence intensity and quenching efficiency for 10 times. Figure S2 showed that the sensor could retain its fluorescence intensity and detection sensitivity without obvious drop (RSD=5.9) during those ten recycles. 3.3 Molecular selectivity and sensing specificity of the sensor. The selectivity of mesoporous structured ratiometric fluorescence MIPs and NIPs were demonstrated by single component measurement. The fluorescence intensity ratios of the probe were recorded (I540/I620) in the presence of melamine and other template analogues (ammeline, cyanuric, atrazine) and metal ions (Pb2+, Cu2+, Hg2+) . It can be seen from Figure 4 that 60% of the fluorescence intensity of the MIPs solution was quenched by 500 nM melamine, accompanied by remarkable fluorescence color changes under UV light. In contrast, no obvious changes about fluorescence intensity ratio and color can be observed when detection of template analogues and metal ions. The results implied the high selectivity of the imprinted ratiometric fluorescence probe. It is well known that some metal ions, such as Cu2+, Hg2+ also can quenching the fluorescence intensity of CdTe QDs. Admittedly the diameters of metal ions are lower than melamine, it is easy for them entering into recognition sites. However, there is no specific interaction between functional monomer and metal ions in the recognition sites, and the hole is not complementary to metal ions in shape and size. So metal ions can desorption from the recognition sites easily. So the quenching resulted from metal ions can be neglected. Competitive 16
molecules including ammeline, cyanuric and atrazine also caused negligence quenching efficiency when compared to melamine. The phenomenon can be explained as the recognition sites of melamine imprinted cavities were not complementary to them, resulting in less chance to quench the fluorescence of the QDs. Figure 4
3.4 Practical application to real sample analysis To investigate the practical application of this method in real samples, the sensing of melamine in milk samples was carried out. Melamine does not be detected in the milk powder and row milk samples. Then, spiked experiments were carried out. The estimated recoveries of the measurements and the RSD are satisfactory (shown in Table 1) and the corresponding fluorescence spectra of the ratiometric probe upon the exposure to different real samples were displayed in Figure S3. The recoveries of melamine for real samples are statistically close to those values added. And the value is much higher agree with obtained using the HPLC method. The comparison shows that this sensing method has promising potential for use in melamine detection. Table 1
Analytical performances of the developed methods for detection of melamine were mainly compared with the previously methods including MIP@QDs sensors (Huy et al., 2014), MIP-SPE-HPLC (Zhang et al., 2011; Cheng et al., 2013), MIP/ 17
electrochemical sensor (Xu et al., 2014), MIP/SPR optic sensors (Shrivastav et al., 2015) and HPLC-MS-MS (Wu et al., 2009), which were summarized in Table 2. Without expensive or sophisticated instrumentation, extensive sample preparation, or the need for highly skilled personnel, the LOD obtained in the present work is on same order as those of MIP electrochemical sensor, but much lower than those of MIP-SPE-HPLC method and MIP@QD sensor without mesoporous structure. Although the lower detection for melamine was achieved by MIP@SPR sensors, it needs elaborated and sophisticated instrumentation. When compared to HPLC method, the presented work do not need time consuming sample pretreatment process and expensive instrument. Compared to MIP@QDs without mesoporous structure, the mesoporous structured MIP@QDs sensor displayed higher sensitivity. Our proposed method is sensitive enough and highly selective for practical uses to determine melamine in various matrices. The developed method requires no sample pretreatments including extraction and derivatization and would be a promising alternative for analyzing melamine quantitatively in food sample. Table 2
4 Conclusion In summary, melamine imprinted mesoporous structured ratiometric fluorescence probes have been constructed by one pot synthesis method, and its utility has been established for detection of melamine in real milk samples. The novel fluorescence MIPs sensor integrated the high selectivity of MIPs and the high sensitivity of 18
ratiometric fluorescence probe and mesoporous silica into one system. The mesoporous structured radiometric fluorescence MIP@QDs probes demonstrated prominence advantages over determination of melamine from real milk samples. The one pot synthesis method confirmed to be simple, rapid, and reliable to build mesoporous structured imprinted ratiometric fluorescence probe. In the meanwhile, the method presented here opens up attractive perspectives for melamine monitoring, and it would be an universe method for other substance analysis.
Acknowledgements We thanks NSFC (NO. 21307052), the Natural Science Foundation of Shandong Province of China (ZR2013BL006).
References Chen, L.X., Xu, S.F., Li, J.X., 2011.Chem. Soc. Rev. 40, 2922-2942. Chen, Y., Wang, D., Yin, Y., Wang, L., Wang, X., Xie, M., 2012. J. Agric. Food Chem. 60, 10472-10479. Cheng, W., Liu, Z., Wang, Y., 2013. Talanta 116, 396–402. Fang, Q., Geng, J., Liu, B., 2009. Chem.-Eur. J. 15, 11507-11514. Huy, B., Seo, M., Zhang, X., Lee, Y., 2014. Biosens. Bioelectron. 57, 310-316. Jiang, Z., Zhou, L., Liang, A., 2011. Chem. Commun. 47, 3162-3164. Kim, H., Ren, W., Kim, J., Yoon, J., 2012. Chem. Soc. Rev. 41, 3210-3244. Li, D., He, X., Chen, Y., Li, W., Zhang, Y., 2013. ACS Appl. Mater. Interfaces. 5, 12609-12616. 19
Liu, J., Chen, H., Lin, Z., Lin, J. 2010. Anal. Chem. 82, 7380-7386. Petya, K., Ivanova, M., Antonio, G., Elena, V., Piletska, Michael, Whitcombe, J., Zhou, Z.X., Petar, A., Mitsev, Frank, D. and Sergey, A., Piletsky, 2012. Angew. Chem. Int. Ed. 51, 5196-5199. Ren, X., Chen, L., 2015. Biosens. Bioelectron. 64, 182-188. Selim, B., Serena, A., Nataliya, M., Emilia, P.,Vishal, K.,Tero, S., Bernadette, T.S.B., Karsten, H., 2014. Angew. Chem. Int. Ed. 53, 8919-8923. Shen, X.T., Johan S. B., Tripta, K., Leif B., Jack C. L., Dirk L. and Ye, L. , 2014. Angew. Chem. Int. Ed. 53, 10687-10690. Shrivastav, A., Mishra, S., Gupta, B., 2015, Sensors and Actuators B 212, 404–410. Tan, J., Wang, H., Yan, X. P., 2009. Anal. Chem. 81, 5273-5280. Tan, J., Wang, H., Yan, X.P., 2009. Biosens. Bioelectron. 24, 3316-3321. Wagner, R., Wan, W., Biyikal, M., Benito-Peña, E., Moreno-Bondi, M., Lazraq, I., Rurack, K., Sellergren, B., 2013. J. Org. Chem. 78, 1377-1389. Wan, W., Mustafa, B., Ricarda, W., Bçrje, S., Knut, R., 2013. Angew. Chem. Int. Ed. 52, 7023-7027. Wu, Y.; Huang, C.; Lin, C.; Ho, W.; Lin, L.; Chiu, T.; Tarng, D.; Lin, C.; Tsai, T., 2009. J. Chromatogr. A 1216, 7595–7601. Wu, Y., Liu, Y., Gao, X., Gao, K., Xia, H., Luo, M., Wang, X., Ye, L., Shi, Y., Lu, B., 2015. Chemosphere 119, 515-523. Xu, G., Zhang, H., Zhong, M., Zhang, T., Lu, X., Kan, X., 2014. J. Electroanal. Chem. 713, 112–118. Xu, S.F., Lu, H.Z., 2015. Chem. Commun. 51, 3200-3203. 20
Xu, S.F., Lu, H.Z., Li, J.H., Song, X.L., Wang, A.X., Chen, L.X., Han, S.B., 2013. ACS Appl. Mater. Interfaces. 5, 8146-8154. Yao, J., Zhang, K., Zhu, H., Ma, F., Sun, M., Yu, H., Sun, J., Wang, S., 2013. Anal. Chem. 85, 6461-6468. Yuan, L., Lin, W., Zheng, K., He, L. and Huang, W., 2013. Chem. Soc. Rev. 42, 622-661. Zhang, H., Zhang, Z., Hu, Y., Yang, X., Yao, S., 2011. J. Agric. Food Chem. 59, 1063– 1071. Zhang, K., Zhou, H., Mei, Q., Wang, S., Guan, G., Liu, R., Zhang, J., 2011. J. Am. Chem. Soc. 133, 8424-8427. Zhang, W., He, X., Chen, Y., Li, W., Zhang, Y., 2011. Biosens. Bioelectron. 26, 2553-2558. Zhao, Y., Ma, Y., Li, H., Wang, L., 2012. Anal. Chem. 84, 386-395. Figure captions Figure 1 Schematic illustration for the one pot preparation of mesoporous sturctured ratiometric fluorescence MIPs probe. Figure 2 TEM images of (A) mesoporous structured ratiometric fluorescence MIPs; (B) mesoporous structured ratiometric fluorescence NIPs; (C) mesoporous structured single fluorescence MIPs; (D) non-mesoporous structured ratiometric fluorescence MIPs; (E) the high resolution TEM for mesoporous structured ratiometric fluorescence MIPs; (F) N2 sorption isotherms and the pore size distribution for mesoporous structured ratiometric fluorescence MIPs. The insets were SEM images for the corresponding TEM images. Figure 3. The fluorescence colors and the corresponding fluorescence spectra (λex = 490 nm) of the ratiometric probe upon the exposure to different concentrations of melamine. A was the mesoporous stuctured ratiometric fluorescence MIPs; B was mesoprous stuctured ratiometric fluorescence NIPs; C was mesoprous stuctured single fluorescence MIPs and D was 21
non-mesoprous stuctured ratiometric fluorescence MIPs. The experimental conditions were MIPs or NIPs 50 mg/L; excited light, 490 nm; silt widths of excitation and emission, 5 and 3 nm, respectively. Figure 4 The selectivity of the mesoporous structured ratiometric fluorescence MIPs and NIPs to other template analogues and metal ions. The experimental conditions were MIPs or NIPs, 50 mg/L; excited light, 490 nm; silt widths of excitation and emission, 5 and 3 nm, respectively. The concentration of metal ions fixed as 10 μM while the value was 500 nM for template analogues. The inset images show the corresponding fluorescence colors under a UV lamp (from left to right were: cyanuric, ammeline, atrazine, melamine, Cu2+, Hg2+, Pb2+).
Figure 1
Figure 2
22
Figure 3
23
Figure 4
NH2
OH N HO
N
N N
OH
cyanuric
HO
Cl N
N N
NH2
HN
N
atrazine
ammeline
NH2
N
N NH H 2N
N N
NH2
melamine
Table 1 Spiked recoveries and relative standard deviations for detection of melamine in spiked raw milk and milk powder samples by ratiometric fluorescence MIPs 24
probe and HPLC.
Sample
Recovery, mean±SD (%)
Spiked
Raw milk
Milk powder
(nM)
Ratiometric probe
HPLC
50
92.7±3.4
94.2±1.9
100
98.2±2.6
97.5±2.9
200
101.3±3.7
99.8±3.2
50
95.4±5.2
96.8±4.6
100
96.8±3.1
97.1±2.8
200
97.4±3.4
97.5±2.6
Table 2 Comparison of determination of melamine to other methods
a
Methods
LOD
Matrix
Ref.
SPEa HPLC-MS-MS
4.2 μg/L
Milk
Wu el al.,2009
MIP@SPRb sensor
9.87nM
Milk
Shrivastav el al., 2015
MIP electrochemical sensor
68 nM
Raw milk
MIP@CNT c -SPE-HPLC
0.3 μg/L
Milk powder
Zhang el al., 2011
MIP@SiO2-SPE-HPLC
NGd
Milk powder
Cheng el al., 2013
MIP@QDs
75 μg/L
Milk drink
Huy et al., 2014
Mesoporous MIP@QDs
38 nM
Milk powder
This work
SPE: solid phase extraction; nanotube; d NG: not given.
b
SPR: surface plasmon resonance;
25
Xu el al.,2014
c
CNT: multiwalled carbon
Highlights
Mesoporous structured ratiometric fluorescence MIPs sensor was prepared. One pot synthesis method was employed. Mesoporous structured ratiometric fluorescence MIPs sensor displayed high sensitivity and high selectivity.
Mesoporous structured ratiometric fluorescence MIPs sensor has potentially applicable for trace pollutants analysis in complicated samples
26
