Biosensors and Bioelectronics 64 (2015) 416–422
Contents lists available at ScienceDirect
Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios
Electrochemical sensor for chloramphenicol based on novel multiwalled carbon nanotubes@molecularly imprinted polymer Guangming Yang a,b, Faqiong Zhao a,n a Key Laboratory of Analytical Chemistry for Biology and Medicine (Ministry of Education), College of Chemistry and Molecular Sciences, Wuhan University, Wuhan 430072, PR China b Department of Resources and Environment, Baoshan University, Baoshan 678000, PR China
ar t ic l e i nf o
a b s t r a c t
Article history: Received 24 June 2014 Received in revised form 16 September 2014 Accepted 21 September 2014 Available online 2 October 2014
Herein, we present a novel electrochemical sensor for the determination of chloramphenicol (CAP), which is based on multiwalled carbon nanotubes@molecularly imprinted polymer (MWCNTs@MIP), mesoporous carbon (CKM-3) and three-dimensional porous graphene (P-r-GO). Firstly, 3-hexadecyl-1vinylimidazolium chloride (C16VimCl) was synthetized by using 1-vinylimidazole and 1-chlorohexadecane as precursors. Then, C16VImCl was used to improve the dispersion of MWCNT and as monomer to prepare MIP on MWCNT surface to obtain MWCNTs@MIP. After that, the obtained MWCNTs@MIP was coated on the CKM-3 and P-r-GO modified glassy carbon electrode to construct an electrochemical sensor for the determination of CAP. The parameters concerning this assay strategy were carefully considered. Under the optimal conditions, the electrochemical sensor offered an excellent response for CAP. The linear response ranges were 5.0 10 9–5 10 7 mol L 1 and 5.0 10 7–4.0 10 6, respectively, and the detection limit was 1.0 10 10 mol L 1. The electrochemical sensor was applied to determine CAP in real samples with satisfactory results. & 2014 Elsevier B.V. All rights reserved.
Keywords: Multiwalled carbon nanotubes Molecularly imprinted polymer Chloramphenicol Mesoporous carbon Porous graphene
1. Introduction Chloramphenicol (CAP) is a broad-spectrum antibiotic. It is effective against a wide variety of gram-positive and gramnegative bacteria and is widely used for the treatment of infectious diseases in humans and animals (Falagas et al., 2008; Turton et al., 2002). However, it also has serious side-effects on human beings such as gray baby syndrome, leukemia, and aplastic anemia (Falagas et al., 2008; Turton et al., 2002). For this reason, many countries such as USA, Canada and China have banned the use of CAP in the food producing animals and set the maximum residue limit for food-safety control programs (Falagas et al., 2008). Thus, it is important to develop sensitive and selective methods for the determination of this compound in food production. At present, several analytical techniques are used for the determination of CAP, including high-performance liquid chromatography (Chen et al., 2009), gas chromatography–mass spectrometry (Shen et al., 2009) and electrochemical technique (Borowiec et al., 2013; Pilehvar et al., 2012; Yadav et al., 2014; Zaidi 2013; Zhuang et al., 2014). Among them, electrochemical method offers the advantages of simplicity and low cost in comparison with n
Corresponding author: Fax: þ 86 27 68752701. E-mail address: [email protected] (F. Zhao).
http://dx.doi.org/10.1016/j.bios.2014.09.041 0956-5663/& 2014 Elsevier B.V. All rights reserved.
other methods (Zaidi, 2013). Many electrodes were applied to the directly electrochemical determination of CAP, such as activated carbon fiber electrode (Agüı ́ et al., 2002), singlewall carbon nanotube (SWCNT)–gold nanoparticle (GNP)–ionic liquid (Xiao et al., 2007) and nitrogen-doped graphene (r-GO) decorated with GNP (Borowiec et al., 2013) modified glassy carbon electrodes (GCEs), self-assembled monolayer modified gold electrode (Codognoto et al., 2010) and platinum electrode (Zhuang et al., 2014), but their selectivity and/or sensitivity were still not satisfactory. To solve this issue, some recognition receptors such as antibody (Chullasat et al., 2011; Liu et al., 2013; Zhang et al., 2011), aptamer (Pilehvar et al., 2012; Yadav et al., 2014; Yan et al., 2012) and molecular imprinted polymer (MIP) (Alizadeh et al., 2012; Zhao et al., 2012) were used to improve the sensing performance. MIP is cheap and stable in comparison with antibody and aptamer. However, the MIPs prepared with conventional methods for the determination of CAP had less effective binding sites as many binding sites were buried in polymeric matrix (Alizadeh et al., 2012; Zhao et al., 2012). To improve this situation, surface molecularly imprinting technique was developed (Chen et al., 2011; Cheong et al., 2013; Sharma et al., 2012). Surface imprinted polymers have higher binding capacity, faster mass transfer and binding kinetics (Sharma et al., 2012, 2013; Yang et al., 2014). However, these advantages most depend on the supporting materials, which should possess large surface,
G. Yang, F. Zhao / Biosensors and Bioelectronics 64 (2015) 416–422
good conductive property and easy functionalization. Multiwalled carbon nanotube (MWCNT) is an ideal supporting material based on these criterions (Chen et al., 2011; Sharma et al., 2013). Thus, it was widely applied for this purpose, but the functional monomer was firstly immobilized on MWCNT surface by using chemical bonding before the preparation of MWCNTs@MIP in most reports (Anirudhan and Alexander, 2014; Bali Prasad et al., 2013; Chi et al., 2012; Lee and Kim, 2009; Moreira et al., 2011; Prasad et al., 2010; Tong et al., 2013; Zhang et al., 2012; Zhao et al., 2014). Recently, Qian et al. reported the in situ chemical oxidative polymerization of pyrrole on MWCNT surface for the preparation of novel MWCNTs@MIP (Qian et al., 2014). Comparing with other methods, this one was simpler, but the obtained quantity of MIP was not enough as expected. Therefore, the facile preparation of MIPs on MWCNT surface is still in need of exploration. Here, we synthetized 3-hexadecyl-1-vinylimidazolium chloride (C16VimCl), which possessed double functional groups for the preparation of MWCMTs@MIP. C16VimCl could improve the dispersion of MWCNT and act as monomer to form MWCNTs@MIP. Then, the obtained MWCNTs@MIP was coated on the mesoporous carbon (CKM-3) and three-dimensional porous graphene (P-r-GO) modified GCE to construct an electrochemical sensor for the determination of CAP. In this strategy, C16VimCl facilitated the preparation of MWCNTs@MIP, and the CKM-3 and P-r-GO made the sensor more sensitive. To the best of our knowledge, this is the first time to use such a functional monomer to prepare MWCNTs@MIP for the determination of CAP.
417
Scheme 1. Schematic diagram of the construction procedure of sensor.
60 °C. The resulting crude product was purified by recrystallization with ethyl acetate for at least three times, giving the desired product as a white powder in 70.0% yield. 1H NMR (CDCl3, δ): 0.94 (t, 3H, –CH3), 1.10–1.41 (m, 26H, 13CH2), 1.94 (t, 2H, CH2), 4.40 (t, 2H, N–CH2), 5.45 (d, 1H, vinyl CH2), 5.96 (d, 1H, vinyl CH2), 7.40 (t, 1H, vinyl CH), 7.60 (s, 1H, imidazole N–CH), 8.02 (s, 1H, imidazole CH–N þ ), 11.20 (s, reduced intensity, imidazole N–CH–N þ ). 2.3. Preparation of MWCNTs@MIP
2. Experimental 2.1. Apparatus and reagents All electrochemical experiments were performed with a CHI 660D electrochemistry workstation (Shanghai CH Instruments Co., China). A three-electrode system was used, consisting of a modified glassy carbon electrode as working electrode, a saturated calomel electrode (SCE) as reference electrode and a platinum foil as counter electrode. Scanning electron microscope (SEM) was performed using a Zeiss (German) with an accelerating voltage of 10 kV. Transmission electron microscopy (TEM) was carried out by using a JEM-2100 transmission electron microscope with an accelerating voltage of 200 kV. 1H NMR spectra were recorded with a Varian Mercury 400 spectrometer and the solvent used was deuterated chloroform (CDCl3). Fourier transform infrared (FTIR) absorption spectra were recorded with a model Nexus-670 spectrometer (Nicolet, USA). Carboxylic acid-functionalized multiwalled carbon nanotube (MWCNT) and mesoporous carbon (CKM-3) were supplied by Nanjing XFNANO Materials Tech Co., Ltd. (Nanjing, China). Chloramphenicol (CAP), thiamphenicol, florfenicol, 1-vinylimidazole, 1-chlorohexadecane, 2.2-azobisisobutyronitrile (AIBN), ethylene glycol dimethacrylate (EGDMA) were purchased from Sigma (St. Louis, MO, USA). Unless stated otherwise, other reagents used were analytical grade. The milk samples and honey samples were purchased from local supermarket. The support electrolyte was 0.1 mol L 1 phosphate buffer solution (PBS, pH ¼7.0), which was prepared with NaH2PO4 and Na2HPO4. 2.2. Synthesis of C16VimCl C16VimCl was synthesized by referring the previous report about the synthesis of 3-hexyl-1-vinylimidazolium bromide (Jana et al., 2013). In a typical reaction, 6.03 mL (i.e. 20 mmol) 1-chlorohexadecane and 1.81 mL (i.e. 20 mmol) 1-vinylimidazole were added to 10 mL dry tetrahydrofuran and let them react for 48 h at
The preparation of MWCNTs@MIP was shown in Scheme 1 (Part A). Firstly, 0.1136 g (i.e. 4.00 mmol) C16VimCl and 40 mg MWCNTs were dispersed into 40 mL methanol/water (4/1, V/V)) by ultrasonication for 30 min to obtain a homogeneous suspension. Then, 0.0323 g (i.e. 1.00 mmol) CAP was completely dissolved into above suspension and let them prepolymerize for 60 min. After that, 10 mg AIBN and 3.1 mL (i.e. 20.00 mmol) EGDMA were dispersed into above solution. The resultant mixture was stirred at 60 °C for 24 h under nitrogen protection. Finally, the product was eluted by methanol/acetic acid (9/1, V/V) until no template molecule was detected in the washing solutions. For comparison, MWCNTs@NIP was prepared by the same manner except the absence of CAP. At the same time, other MWCNTs@MIPs were prepared by changing the mole ratio of CAP to C16VimCl (i.e. 1:2, 1:3, 1:5 and 1:6), the mole ratio of CAP to EGDMA (i.e. 1:15, 1:18, 1:22 and 1:25) and the mass ratio of MWCNT to C16VimCl (1:2, 1:2.5, 1:3 and 1:3.2), respectively. 2.4. Preparation of porous r-GO Graphene oxide (GO) was prepared by a modified Hummers method (Liang et al., 2009) and characterized as our previous report (Yang et al., 2014). According to the report (Chen and Qiao 2013), porous GO (P-GO) was obtained by oxidizing and etching GO with KMnO4 and HCl, but we lengthened oxidation time to 12 h in order to acquire bigger pore and better three-dimensional structure. Specifically, 100 mL of 0.5 mg mL 1 GO suspension was mixed with 0.5 g KMnO4 under magnetic stirring for 12 h. The above solution was merged with 30 mL HCl (36%, wt%) and 30 mL H2O2 (30%, wt%) for 3 h. After that, the products were separated by centrifugation, washed with water and dried in a vacuum oven at 60 °C. Next, 200 mL of 0.3 mg mL 1 P-GO suspension was mixed with 2.4 mL ammonia (28%, wt%) and 0.24 mL hydrazine hydrate (85%, wt%), followed by heating at 95 °C for 12 h. Then, the products were separated by centrifugation, washed with water and dried under a vacuum oven at 60 °C.
418
G. Yang, F. Zhao / Biosensors and Bioelectronics 64 (2015) 416–422
2.5. Preparation of MWCNTs@MIP/CKM-3/P-r-GO/GCE The preparation of MWCNTs@MIP/CKM-3/P-r-GO/GCE was shown in Scheme 1 (Part B). Firstly, 2 mg P-r-GO was dispersed in 10 mL N,N-dimethylformamide (DMF) to give 0.2 mg mL 1 P-rGO suspension. At the same time, 2.5 mg CKM-3 and 5 mg MWCNTs@MIP were dispersed in 5 mL DMF respectively, to obtain 0.5 mg mL 1 CKM-3 and 1 mg mL 1 MWCNTs@MIP. Then, 5 mL of 0.2 mg mL 1 P-r-GO suspension was coated on a clean GCE surface and dried at 80 °C. After that, 5 mL of 1 mg mL 1 CKM-3 suspension was coated on the P-r-GO modified GCE and dried at 80 °C. In succession, 5 mL of 1 mg mL 1 MWCNTs@MIP suspension was dropped on above modified electrode and dried at 80 °C. Thus, an MWCNTs@MIP/CKM-3/P-r-GO/GCE was obtained. For comparison, MWCNTs@MIP/r-GO/GCEs (the concentration of r-GO was 1.0 mg mL 1), MWCNTs@MIP/P-r-GO /GCE, MWCNTs@MIP/CKM3/GCE and MWCNTs@MIP/P-r-GO/CKM-3/GCE were prepared under same conditions. 2.6. Electrochemical measurements The sensor was immersed in a PBS (pH ¼7.0) containing a certain concentration of CAP for 200 s, under mildly magnetic stirring. Then, it was rinsed with H2O. After that, the threeelectrode system was assembled on a cell with 10 mL PBS (pH¼7.0), and cyclic voltammogram (CV) or differential pulse voltammogram (DPV) was recorded from 800 mV to 400 mV (vs SCE).
3. Results and discussion 3.1. Morphological and structural characterization Fig. 1 shows the TEM of MWCNTs and MWCNTs@MIP. It was clear that the diameter of MWCNTs was about 10–15 nm (Fig. 1A) and MIP was successfully prepared on MWCNTs surface (B) and its thickness was about 20 nm. At the same time, the FTIR absorption spectra of MWCNTs@MIP also confirmed this point (Fig. S1A). The characteristic peaks of the bonded EGDMA (i.e., 1710 cm 1 (C=O) and 1230/1153 cm 1) (C‒O‒C) were observed. At the same time, the peaks corresponding to the C=N stretching and imidazole ring stretching (1554 and 1460 cm 1) and C=C stretching (1628 cm 1) of C16VimCl (Fig. S1A). Next, the FTIR absorption spectra of P-GO and P-r-GO were recorded (Fig. S1B). Results indicated that the absorption bands of the oxo-groups on P-r-GO were obviously decreased or
disappeared in comparison with P-GO (i.e., 1725 cm 1 (C‒O), 3430, 1393 and 1053 cm 1 (–OH)), meaning that the P-GO was reduced successfully. Then, the TEM of P-r-GO was recorded. It presented three-dimensional and porous structure and the pore diameters were about 20–30 nm (Fig. 2A). The evolution of modified electrode was recorded by using SEM. The P-r-GO modified GCE showed three-dimensional surface (Fig. 2B). Fig. 2C indicates that CKM-3 tightly adhered on P-r-GO surface. When MWCNTs@MIP suspension was coated on the modified electrode, the MWCNTs@MIP showed three-dimensional nanonetwork structure (Fig. 2D). 3.2. Adsorption study Adsorption experiments were used to evaluate the molecularly imprinted effect. Results indicated that the adsorption amount of MWCNTs@MIPs (or MWCNTs@NIP) increased with CAP concentration increasing (Fig. S1C). When the CAP concentration exceeded 1.0 10 4 mol L 1, the adsorption amount of MWCNTs @NIP kept unchanged, meaning that it reached adsorption equilibrium and the value was about 25 μmol g 1. However, for MWCNTs@MIPs prepared with different mole ratio of CAP to C16VimCl (1:1–1:6), when the CAP concentration exceeded 1.4 10 4 mol L 1 the adsorption amount increased from 62 to 91 μmol g 1, indicating that the imprinted effect was obvious. It achieved a maximum when the ratio of CAP to C16VimCl was 1: 4. This was related to the change of the number of available binding sites. When the amount of functional monomer was too little to combine enough template molecules, the available binding sites were less. On the contrary, the MIP film became thick and many binding sites were buried in polymeric matrix so that the adsorption amount decreased. For MWCNTs@MIPs prepared with different mole ratio of CAP to EGDMA (i.e. 1:15–1:25) (Fig. S1D), the adsorption amount of MWCNTs@MIPs increased from 65 to 91 μ mol g 1 and then decreased. It achieved the maximum when the ratio was 1:20. This should be ascribed to the effect of the ratio of CAP to EGDMA on the formation of available binding sites, like the effect of the ratio of CAP to C16VimCl. For the MWCNTs@MIPs prepared with different mole ratios of MWCNT to C16VimCl (i.e. 1:2–1:3.2) (Fig. S1E), the adsorption amount of MWCNTs@MIPs increased from 60 to 91 μmol g 1 and then decreased. It was thought that the number of available binding sites increased with enhancing the amount of C16VimCl. However, when the ratio of MWCNT to C16VimCl was lower than 1:2.8, some binding sites were buried in polymeric matrix due to the increased MIP film thickness. Therefore, the adsorption amount decreased.
Fig. 1. TEM images of MWCNT (A) and MWCNTs@MIP (B).
G. Yang, F. Zhao / Biosensors and Bioelectronics 64 (2015) 416–422
419
Fig. 2. TEM image of P-r-GO (A), SEM images of P-r-GO/GCE (B), CKM-3/P-r-GO/GCE and MWCNTs@MIP/CKM-3/P-r-GO/GCE (C).
3.3. Voltammetric characterization Molecularly imprinted effect of different sensors was also reflected by the voltammetric behavior of CAP (Fig. 3(A and B)). The MWCNTs@MIP/CKM-3/P-r-GO/GCE showed a small background current signal in PBS (Fig. 3A(a)), but when the electrode was soaked in a CAP solution (1.0 10 6 mol L 1) for adsorption,
it exhibited a obvious oxidation peak at about 92 mV and two reduction peaks at 104 mV and 640 mV (Fig. 3A(b)). This was related to the oxidation/reduction reactions of CAP. The peak at 640 mV could be attributed to the reduction of the nitro group of CAP, and the other two peaks were produced by the reversible redox of the reduction product (i.e. hydroxylamine derivative) (Borowiec et al., 2013; Xiao et al., 2007). However, the
Fig. 3. (A) Voltammograms of different electrodes in PBS (pH ¼7.0). Electrodes: (a) MWCNTs@MIP/CKM-3/P-r-GO/GCE in blank PBS; (b) MWCNTs@MIP/CKM-3/P-r-GO/GCE after binding CAP in 1.0 10 6 mol L 1 CAP; (c) MWCNTs@NIP/CKM-3/P-r-GO/GCE after binding CAP in 1.0 10 6 mol L 1 CAP; (B) DPVs of different MIP electrodes in PBS (pH ¼7.0) after binding CAP in 1.0 10 6 mol L 1 CAP. Electrodes: (a) r-GO/GCE; (b) P-r-GO/GCE; (c) CKM-3/GCE; (d) CKM-3/P-r-GO/GCE; (e) P-r-GO/CKM-3/GCE.
420
G. Yang, F. Zhao / Biosensors and Bioelectronics 64 (2015) 416–422
MWCNTs@NIP/CKM-3/P-r-GO/GCE (Fig. 3A(c)) showed small peaks under the same conditions, meaning that the nonspecifically adsorption of CAP was weak and the effect of MIP was very obvious. DPVs of the MIP sensors prepared with different modified interfaces were recorded (Fig. 3B). Results indicated that the peak current of MWCNTs@MIP/P-r-GO/GCE (Fig. 3B(b)) was larger than that of MWCNTs@MIP/r-GO/GCE (B(a)) (r-GO concentration was 1 mg mL 1, which was optimized), meaning the three-dimensional porous film had large surface area and enhanced masstransfer rate. When P-r-GO was replaced by CKM-3, the obtained MWCNTs@MIP/CKM/GCE (B(c)) showed better response than MWCNTs@MIP/P-r-GO/GCE (B(b)). This should be ascribed to the excellence electro-catalysis of CKM-3 for nitrobenzene compounds due to its unique properties (Walcarius, 2012; Zang et al., 2011; Zhang et al., 2013). Compared with other MIP sensors, the MWCNTs@MIP/CKM-3/P-r-GO/GCE showed higher sensitivity (Fig. 3B(d)). However, when CKM-3 was firstly modified and then P-r-GO (i.e. MWCNTs@MIP/P-r-GO/CKM/GCE (Fig. 3B(e))), the CKM-3 was buried by P-r-GO, which made response current decrease in comparison with MWCNTs@MIP/CKM/GCE (B(c)) and MWCNTs@MIP/CKM/P-r-GO/GCE (B(d)). For this strategy, P-r-GO enhanced electrode surface area and benefited the dispersion of CKM-3. When the MWCNTs@MIP was dropped on this interface, it well contacted CKM-3 and CKM-3 could catalyze the reaction of CAP on MWCNTs@MIPs. At the same time, three-dimensional modified electrode benefited mass transfer, enhancing the sensitivity of sensor. 3.4. Optimization of experimental variables 3.4.1. Effect of P-r-GO concentration The P-r-GO concentration was optimized in the range of 0.1–0.3 mg mL 1 for the preparation of sensor (Fig. S2A). Result indicated that the peak current increased with the concentration increasing up to 0.2 mg mL 1, and then gradually decreased. This should be ascribed to the change of electrode surface area and film resistance. Hence, 5 mL of 0.2 mg L 1 P-r-GO was selected for preparing modified electrode. 3.4.2. Effect of CKM-3 concentration The effect of CKM-3 concentration was discussed in the range from 0.1 to 1.0 mg mL 1 (Fig. S2B). It was found that the peak current increased with the concentration increasing, and when the concentration increased to 0.5 mg mL 1, the optimum response value was obtained. Then, the response signal gradually decreased. This was also related to the change of electrode surface area and resistance. In the experiments, 5 mL of 0.5 mg mL 1 CKM-3 was selected. 3.4.3. Effect of mole ratio of CAP to C16VimCl and EGDMA The mole ratio of CAP to C16VimCl was changed from 1:2 to 1:6 to test its effect (Fig. S2C). As a result, the response current increased with the ratio decreasing, and it achieved a maximum at 1:4. Then, the current response gradually decreased. This was related to the change of the number of available binding sites, which was mentioned above (Fig. S1C). Additionally, the thickness of MIP film increased with the ratio decreasing, leading to the decrease of the conductive property. Here, the mole ratio of 1:4 was chosen. At the same time, the effect of the mole ratio of CAP to EGDMA was also considered in the range of 1:10–1:25 (Fig. S2D). Results indicated that the peak current increased with the ratio decreasing, and it achieved a maximum at 1:20. In succession, the current response gradually decreased. This could be explained by the change of the number of available binding sites as mentioned
previously (Fig. S1D). In addition, the conductive property of MWCNTs@MIP became weak with the EGDMA increasing. Therefore, the mole ratio of 1:20 was chosen. 3.4.4. Effect of mass ratio of MWCNT to C16VimCl The mass ratio of MWCNT to C16VimCl was changed from 1:2 to 1:3.2 to test its effect (Fig. S2E). Result indicated that the peak current of CAP increased when the ratio changed from 1:2 to 1:2.8. Further increase caused a gradual decrease in peak current. This was related to the change of density of imprinted sites on MWCNTs (Fig. S1E) and the conductive property of resulting MWCNTs@MIP. When the ratio was smaller the site density was higher, but the resistance of MWCNTs@MIP increased. In this work, the mass ratio of 1:2.8 was chosen. 3.4.5. Effect of MWCNTs@MIP concentration The effect of MWCNTs@MIP concentration was discussed in the range of 0.2–2.0 mg mL 1 (Fig. S2F). The peak current of CAP increased when its concentration changed from 0.2 to 1.0 mg mL 1. Further increase caused a gradual decrease in peak current. The reason was that the adsorption amount of CAP increased with the amount of MWCNTs@MIP increasing, but overmuch MWCNTs@MIP made the resistance increase, which resulted in response current reducing. Therefore, 1 mg mL 1 was a balanced point and was adopted. At the same time, the effect of MWCNTs@MIP concentration on selectivity was also considered. It was found that 50-fold of p-nitrophenol (i.e. 5 10 5 mol L 1) showed weak few interference for the detection of CAP (1.0 10 6 mol L 1) when the MWCNTs@MIP concentration was lower (i.e. 0.2 mg mL 1), but when the concentration was up to 0.5 mg mL 1, the p-nitrophenol almost did not influence the detection (signal change was 2.3%), meaning that the MIP film became enough thick to block the electrode reaction of most p-nitrophenol molecules so that the selectivity was enhanced. 3.4.6. Effect of adsorption time The effect of adsorption time on the response current was also considered in the range of 60–200 s (Fig. S2G). It was found that the peak current increased gradually with extending adsorption time from 60 to 120 s for a 1 10 6 mol L 1 CAP solution (Fig. S2G (a)). Then, it kept almost unchanged, which indicated adsorption of CAP was saturated. However, the saturated adsorption time was up to 140 s for a lower concentration solution (i.e. 2.5 10 8 mol L 1) (Fig. S2G(b)). Thus, 200 s was chosen as adsorption time so that the CAP solutions with lower concentrations had enough time to reach saturated adsorption. 3.4.7. Effect of pH The effect of pH value of supporting electrolyte was investigated over the range from pH 5.5 to pH 8.5 (Fig. S3). It was found that the optimum pH was 7.0. At the same time, the peak potential changed with pH value. The slope of the peak potential vs pH plot was 55.9 mV pH 1, indicating that the number of electron transfer was equal to that of hydrogen ion taking part in the electrode reaction. 3.5. The optimal response characteristics of MIP sensor Under optimal experimental conditions, the DPVs of CAP were recorded (Fig. 4). As expected, the peak current increased upon the increase of CAP concentration. The calibration curve is shown in Fig. 4, and the linear response ranges were 5.0 10 9–5.0 10 7 mol L 1 and 5.0 10 7–4.0 10 6 mol L 1, respectively, and their regression equations were ip (mA)¼ 10.92þ128.18C (mmol L 1), (r¼0.9973) and ip (mA)¼ 62.29þ9.8C (mmol L 1), (r¼0.9921), respectively. The detection
G. Yang, F. Zhao / Biosensors and Bioelectronics 64 (2015) 416–422
421
Table 1 Recovery tests of CAP in milk and honey samples. Samples
Added (μmol L 1) Found (μmol L 1)
RSD (%, n ¼3)
Recovery (%)
Milk 1
0.00 0.025 0.10 1.00 2.50
0.00 0.023 0.094 0.95 2.54
4.3 3.5 2.8 2.5
92 94 95 102
0.00 0.025 0.10 2.50
0.00 0.024 0.91 2.33
5.0 4.2 5.1
96 91 93
Honey 1 0.00 0.025 0.100 1.00 2.50
0.00 0.026 0.095 0.97 2.38
2.4 5.2 4.1 4.3
104 95 97 95
Honey 2 0.00 0.025 0.10 2.50
0.00 0.023 1.04 2.40
3.3 2.2 3.8
92 104 96
Milk 2
Fig. 4. DPVs of the MWCNTs@MIP/CKM-3/P-r-GO/GCE in PBS (pH¼ 7.0) after binding different CAP concentrations. CAP concentrations: 0–4.8 10 6 mol L 1. Inset is the calibration curve of CAP. Error bars represent standard deviation, n¼ 3.
limit was 1.0 10 10 mol L 1 (S/N¼3). Compared with other electrochemical sensors, this sensor offered a reasonable linear range and a lower detection limit for the directly electrochemical sensors (Agüı ́ et al., 2002; Borowiec et al., 2013; Codognoto et al., 2010; Xiao et al., 2007; Zhuang et al., 2014), MIP based sensors (Alizadeh et al., 2012; Zhao et al., 2012) and aptamer based sensor (Pilehvar et al., 2012; Yan et al., 2012). 3.6. Selectivity To assess the selectivity of this method, the interference of some potential organic compounds in real samples was tested. Results indicated that 100-fold of glucose, ascorbic acid, uric acid and glutamic acid did not affect the detection of CAP (1.0 10 6 mol L 1, signal change below 5%). At the same time, the influence of similar compounds (i.e. florfenicol, thiamphenicol, 4-nitrobenzamide and p-nitrophenol) was also investigated. As a result, 50-fold of these compounds had no interference (Fig. S4). This should be ascribed to the effect of MIP, which enhanced the selectivity of sensor. 3.7. Repeatability and stability The repeatability of the sensor was investigated by detecting 1.0 10 6 mol L 1 CAP and a variation coefficient of 3.4% was observed for five successive assays. The current response of the sensor still remained up to 96.4% of its initial value after 15 successive assays (RSD ¼ 2.1%). The stability of the sensor was investigated by measuring its response current to 1.0 10 6 mol L 1 CAP over 20 days. When not in use, the sensor was stored at room temperature. After 20 days, the current response of the sensor remained up to 95.1% (RSD ¼2.8%, n ¼3) of its initial value. The fabrication reproducibility was also estimated with five different electrodes, which were fabricated independently by the same procedure. The RSD was 6.1% for the peak current of 1.0 10 6 mol L 1 CAP, which reflected the reliability of the fabrication procedure.
precipitation (Alizadeh et al., 2012) and extraction of CAP (Xiao et al., 2007). Specifically, 10 mL milk sample and 1 mL of CAP standard solution with different concentrations was added in 7 mL water. Then, 2 mL of trichloroacetic acid (20%, wt%) solution was added to the mixed solution for protein precipitation. The mixture was vortexed for 1 min, and then centrifuged at 4000 rpm for 10 min. The supernatant was collected and filtered through a 0.22 μm membrane filter. After that, the CAP was extracted by using ethyl acetate (the amount is 10 mL for one time) as an extraction liquid for three times. Next, the organic extraction was collected and evaporated to dryness using a gentle stream of nitrogen gas. Afterwards, 10.0 mL PBS (0.10 M PBS) was added and vigorously vortexed for 10 s. Then, the resultant PBS containing CAP was transferred to a cell for the recovery determination. For honey samples, 1.00 g sample and 1 mL of CAP standard solution with different concentrations were added into 9.0 mL water. Then, the treatment was similar to that for milk sample except the protein precipitation. The results are shown in Table 1. The recoveries for CAP standards added was 91–104%, which demonstrates that the sensor may provide a useful tool for determining CAP in real samples.
4. Conclusions In this paper, a facile method was used to prepare MWCNTs@MIP for the electrochemical determination of CAP. C16VimCl not only acted as novel functional monomer, but also promoted the dispersion of MWCNT as it tightly adhered on MWCNT surface. At the same time, three-dimensional P-r-GO and CKM-3 were applied to modify GCE to enhance the response signal. The resulting sensor exhibited sensitive and selective response to CAP and was applied to the detection of CAP in real samples. This work provided a useful platform for the preparation of MWCNTs@MIP based sensors.
3.8. Detection of real samples The proposed MIP electrochemical sensor was evaluated by performing recovery tests for CAP in milk and honey samples. Firstly, the milk samples were treated according to the two reports (Alizadeh et al., 2012; Xiao et al., 2007), which involved protein
Acknowledgements This work was jointly supported by the National Natural Science Foundation of China (Grant nos. 21075092 and 21277105) and the
422
G. Yang, F. Zhao / Biosensors and Bioelectronics 64 (2015) 416–422
Youth Foundation of Science Commission of Yunnan Province (Grant no. 2012FD061).
Appendix A. Supplementary information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.bios.2014.09.041.
References Agüı,́ L., Guzmán, A., Yáñez-Sedeño, P., Pingarrón, J.M., 2002. Anal. Chim. Acta 461, 65–73. Alizadeh, T., Ganjali, M.R., Zare, M., Norouzi, P., 2012. Food Chem. 130, 1108–1114. Anirudhan, T.S., Alexander, S., 2014. Appl. Surf. Sci. 303, 180–186. Bali Prasad, B., Jauhari, D., Tiwari, M.P., 2013. Biosens. Bioelectron. 50, 19–27. Borowiec, J., Wang, R., Zhu, L., Zhang, J., 2013. Electrochim. Acta 99, 138–144. Chen, H., Chen, H., Ying, J., Huang, J., Liao, L., 2009. Anal. Chim. Acta 632, 80–85. Chen, L., Xu, S., Li, J., 2011. Chem. Soc. Rev. 40, 2922–2942. Chen, S., Qiao, S., 2013. ACS Nano 7, 10190–10196. Cheong, W.J., Yang, S.H., Ali, F., 2013. J. Sep. Sci. 36, 609–628. Chi, W., Shi, H., Shi, W., Guo, Y., Guo, T., 2012. J. Hazard. Mater. 227–228, 243–249. Chullasat, K., Kanatharana, P., Limbut, W., Numnuam, A., Thavarungkul, P., 2011. Biosens. Bioelectron. 26, 4571–4578. Codognoto, L., Winter, E., Doretto, K.M., Monteiro, G.B., Rath, S., 2010. Microchim. Acta 169, 345–351. Falagas, M.E., Grammatikos, A.P., Michalopoulos, A., 2008. Expert Rev. Anti-Infect. Ther. 6, 593–600. Jana, S., Vasantha, V.A., Stubbs, L.P., Parthiban, A., Vancso, J.G., 2013. J. Polym. Sci. Part A: Polym. Chem. 51, 3260–3273. Lee, H.Y., Kim, B.S., 2009. Biosens. Bioelectron. 25, 587–591. Liang, Y., Wu, D., Feng, X., Müllen, K., 2009. Adv. Mater. 21, 1679–1683.
Liu, B., Zhang, B., Chen, G., Tang, D., 2013. Microchim. Acta 181, 257–262. Moreira, F.T., Dutra, R.A., Noronha, J.P., Cunha, A.L., Sales, M.G., 2011. Biosens. Bioelectron., 243–250. Pilehvar, S., Mehta, J., Dardenne, F., Robbens, J., Blust, R., De Wael, K., 2012. Anal. Chem. 84, 6753–6758. Prasad, B.B., Madhuri, R., Tiwari, M.P., Sharma, P.S., 2010. Electrochim. Acta 55, 9146–9156. Qian, T., Yu, C., Zhou, X., Ma, P., Wu, S., Xu, L., Shen, J., 2014. Biosens. Bioelectron. 58, 237–241. Sharma, P.S., D'Souza, F., Kutner, W., 2012. TrAC—Trends Anal. Chem. 34, 59–77. Sharma, P.S., Dabrowski, M., D'Souza, F., Kutner, W., 2013. TrAC—Trends Anal. Chem. 51, 146–157. Shen, J., Xia, X., Jiang, H., Li, C., Li, J., Li, X., Ding, S., 2009. J. Chromatogr. B 877, 1523–1529. Tong, Y., Li, H., Guan, H., Zhao, J., Majeed, S., Anjum, S., Liang, F., Xu, G., 2013. Biosens. Bioelectron. 47, 553–558. Turton, J.A., Andrews, C.M., Havard, A.C., Williams, T.C., 2002. Int. J. Exp. Pathol. 83, 225–238. Walcarius, A., 2012. TrAC—Trends Anal. Chem. 38, 79–97. Xiao, F., Zhao, F., Li, J., Yan, R., Yu, J., Zeng, B., 2007. Anal. Chim. Acta 596, 79–85. Yadav, S.K., Agrawal, B., Chandra, P., Goyal, R.N., 2014. Biosens. Bioelectron. 55, 337–342. Yan, L., Luo, C., Cheng, W., Mao, W., Zhang, D., Ding, S., 2012. J. Electroanal. Chem. 687, 89–94. Yang, G., Zhao, F., Zeng, B., 2014. Electrochim. Acta 135, 154–160. Zaidi, S.A., 2013. Int. J. Electrochem. Sci. 8, 9936–9955. Zang, J., Guo, C.X., Hu, F., Yu, L., Li, C.M., 2011. Anal. Chim. Acta 683, 187–191. Zhang, D., Yu, D., Zhao, W., Yang, Q., Kajiura, H., Li, Y., Zhou, T., Shi, G., 2012. Analyst 137, 2629–2636. Zhang, N., Xiao, F., Bai, J., Lai, Y., Hou, J., Xian, Y., Jin, L., 2011. Talanta 87, 100–105. Zhang, T., Lang, Q., Yang, D., Li, L., Zeng, L., Zheng, C., Li, T., Wei, M., Liu, A., 2013. Electrochim. Acta 106, 127–134. Zhao, H., Chen, Y., Tian, J., Yu, H., Quan, X., 2012. J. Electrochem. Soc. 159, J231–J235. Zhao, L., Zhao, F., Zeng, B., 2014. Biosens. Bioelectron. 60, 71–76. Zhuang, Y., Cai, L., Cao, G., 2014. J. Electrochem. Soc. 161, H129–H132.
Contents lists available at ScienceDirect
Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios
Electrochemical sensor for chloramphenicol based on novel multiwalled carbon nanotubes@molecularly imprinted polymer Guangming Yang a,b, Faqiong Zhao a,n a Key Laboratory of Analytical Chemistry for Biology and Medicine (Ministry of Education), College of Chemistry and Molecular Sciences, Wuhan University, Wuhan 430072, PR China b Department of Resources and Environment, Baoshan University, Baoshan 678000, PR China
ar t ic l e i nf o
a b s t r a c t
Article history: Received 24 June 2014 Received in revised form 16 September 2014 Accepted 21 September 2014 Available online 2 October 2014
Herein, we present a novel electrochemical sensor for the determination of chloramphenicol (CAP), which is based on multiwalled carbon nanotubes@molecularly imprinted polymer (MWCNTs@MIP), mesoporous carbon (CKM-3) and three-dimensional porous graphene (P-r-GO). Firstly, 3-hexadecyl-1vinylimidazolium chloride (C16VimCl) was synthetized by using 1-vinylimidazole and 1-chlorohexadecane as precursors. Then, C16VImCl was used to improve the dispersion of MWCNT and as monomer to prepare MIP on MWCNT surface to obtain MWCNTs@MIP. After that, the obtained MWCNTs@MIP was coated on the CKM-3 and P-r-GO modified glassy carbon electrode to construct an electrochemical sensor for the determination of CAP. The parameters concerning this assay strategy were carefully considered. Under the optimal conditions, the electrochemical sensor offered an excellent response for CAP. The linear response ranges were 5.0 10 9–5 10 7 mol L 1 and 5.0 10 7–4.0 10 6, respectively, and the detection limit was 1.0 10 10 mol L 1. The electrochemical sensor was applied to determine CAP in real samples with satisfactory results. & 2014 Elsevier B.V. All rights reserved.
Keywords: Multiwalled carbon nanotubes Molecularly imprinted polymer Chloramphenicol Mesoporous carbon Porous graphene
1. Introduction Chloramphenicol (CAP) is a broad-spectrum antibiotic. It is effective against a wide variety of gram-positive and gramnegative bacteria and is widely used for the treatment of infectious diseases in humans and animals (Falagas et al., 2008; Turton et al., 2002). However, it also has serious side-effects on human beings such as gray baby syndrome, leukemia, and aplastic anemia (Falagas et al., 2008; Turton et al., 2002). For this reason, many countries such as USA, Canada and China have banned the use of CAP in the food producing animals and set the maximum residue limit for food-safety control programs (Falagas et al., 2008). Thus, it is important to develop sensitive and selective methods for the determination of this compound in food production. At present, several analytical techniques are used for the determination of CAP, including high-performance liquid chromatography (Chen et al., 2009), gas chromatography–mass spectrometry (Shen et al., 2009) and electrochemical technique (Borowiec et al., 2013; Pilehvar et al., 2012; Yadav et al., 2014; Zaidi 2013; Zhuang et al., 2014). Among them, electrochemical method offers the advantages of simplicity and low cost in comparison with n
Corresponding author: Fax: þ 86 27 68752701. E-mail address: [email protected] (F. Zhao).
http://dx.doi.org/10.1016/j.bios.2014.09.041 0956-5663/& 2014 Elsevier B.V. All rights reserved.
other methods (Zaidi, 2013). Many electrodes were applied to the directly electrochemical determination of CAP, such as activated carbon fiber electrode (Agüı ́ et al., 2002), singlewall carbon nanotube (SWCNT)–gold nanoparticle (GNP)–ionic liquid (Xiao et al., 2007) and nitrogen-doped graphene (r-GO) decorated with GNP (Borowiec et al., 2013) modified glassy carbon electrodes (GCEs), self-assembled monolayer modified gold electrode (Codognoto et al., 2010) and platinum electrode (Zhuang et al., 2014), but their selectivity and/or sensitivity were still not satisfactory. To solve this issue, some recognition receptors such as antibody (Chullasat et al., 2011; Liu et al., 2013; Zhang et al., 2011), aptamer (Pilehvar et al., 2012; Yadav et al., 2014; Yan et al., 2012) and molecular imprinted polymer (MIP) (Alizadeh et al., 2012; Zhao et al., 2012) were used to improve the sensing performance. MIP is cheap and stable in comparison with antibody and aptamer. However, the MIPs prepared with conventional methods for the determination of CAP had less effective binding sites as many binding sites were buried in polymeric matrix (Alizadeh et al., 2012; Zhao et al., 2012). To improve this situation, surface molecularly imprinting technique was developed (Chen et al., 2011; Cheong et al., 2013; Sharma et al., 2012). Surface imprinted polymers have higher binding capacity, faster mass transfer and binding kinetics (Sharma et al., 2012, 2013; Yang et al., 2014). However, these advantages most depend on the supporting materials, which should possess large surface,
G. Yang, F. Zhao / Biosensors and Bioelectronics 64 (2015) 416–422
good conductive property and easy functionalization. Multiwalled carbon nanotube (MWCNT) is an ideal supporting material based on these criterions (Chen et al., 2011; Sharma et al., 2013). Thus, it was widely applied for this purpose, but the functional monomer was firstly immobilized on MWCNT surface by using chemical bonding before the preparation of MWCNTs@MIP in most reports (Anirudhan and Alexander, 2014; Bali Prasad et al., 2013; Chi et al., 2012; Lee and Kim, 2009; Moreira et al., 2011; Prasad et al., 2010; Tong et al., 2013; Zhang et al., 2012; Zhao et al., 2014). Recently, Qian et al. reported the in situ chemical oxidative polymerization of pyrrole on MWCNT surface for the preparation of novel MWCNTs@MIP (Qian et al., 2014). Comparing with other methods, this one was simpler, but the obtained quantity of MIP was not enough as expected. Therefore, the facile preparation of MIPs on MWCNT surface is still in need of exploration. Here, we synthetized 3-hexadecyl-1-vinylimidazolium chloride (C16VimCl), which possessed double functional groups for the preparation of MWCMTs@MIP. C16VimCl could improve the dispersion of MWCNT and act as monomer to form MWCNTs@MIP. Then, the obtained MWCNTs@MIP was coated on the mesoporous carbon (CKM-3) and three-dimensional porous graphene (P-r-GO) modified GCE to construct an electrochemical sensor for the determination of CAP. In this strategy, C16VimCl facilitated the preparation of MWCNTs@MIP, and the CKM-3 and P-r-GO made the sensor more sensitive. To the best of our knowledge, this is the first time to use such a functional monomer to prepare MWCNTs@MIP for the determination of CAP.
417
Scheme 1. Schematic diagram of the construction procedure of sensor.
60 °C. The resulting crude product was purified by recrystallization with ethyl acetate for at least three times, giving the desired product as a white powder in 70.0% yield. 1H NMR (CDCl3, δ): 0.94 (t, 3H, –CH3), 1.10–1.41 (m, 26H, 13CH2), 1.94 (t, 2H, CH2), 4.40 (t, 2H, N–CH2), 5.45 (d, 1H, vinyl CH2), 5.96 (d, 1H, vinyl CH2), 7.40 (t, 1H, vinyl CH), 7.60 (s, 1H, imidazole N–CH), 8.02 (s, 1H, imidazole CH–N þ ), 11.20 (s, reduced intensity, imidazole N–CH–N þ ). 2.3. Preparation of MWCNTs@MIP
2. Experimental 2.1. Apparatus and reagents All electrochemical experiments were performed with a CHI 660D electrochemistry workstation (Shanghai CH Instruments Co., China). A three-electrode system was used, consisting of a modified glassy carbon electrode as working electrode, a saturated calomel electrode (SCE) as reference electrode and a platinum foil as counter electrode. Scanning electron microscope (SEM) was performed using a Zeiss (German) with an accelerating voltage of 10 kV. Transmission electron microscopy (TEM) was carried out by using a JEM-2100 transmission electron microscope with an accelerating voltage of 200 kV. 1H NMR spectra were recorded with a Varian Mercury 400 spectrometer and the solvent used was deuterated chloroform (CDCl3). Fourier transform infrared (FTIR) absorption spectra were recorded with a model Nexus-670 spectrometer (Nicolet, USA). Carboxylic acid-functionalized multiwalled carbon nanotube (MWCNT) and mesoporous carbon (CKM-3) were supplied by Nanjing XFNANO Materials Tech Co., Ltd. (Nanjing, China). Chloramphenicol (CAP), thiamphenicol, florfenicol, 1-vinylimidazole, 1-chlorohexadecane, 2.2-azobisisobutyronitrile (AIBN), ethylene glycol dimethacrylate (EGDMA) were purchased from Sigma (St. Louis, MO, USA). Unless stated otherwise, other reagents used were analytical grade. The milk samples and honey samples were purchased from local supermarket. The support electrolyte was 0.1 mol L 1 phosphate buffer solution (PBS, pH ¼7.0), which was prepared with NaH2PO4 and Na2HPO4. 2.2. Synthesis of C16VimCl C16VimCl was synthesized by referring the previous report about the synthesis of 3-hexyl-1-vinylimidazolium bromide (Jana et al., 2013). In a typical reaction, 6.03 mL (i.e. 20 mmol) 1-chlorohexadecane and 1.81 mL (i.e. 20 mmol) 1-vinylimidazole were added to 10 mL dry tetrahydrofuran and let them react for 48 h at
The preparation of MWCNTs@MIP was shown in Scheme 1 (Part A). Firstly, 0.1136 g (i.e. 4.00 mmol) C16VimCl and 40 mg MWCNTs were dispersed into 40 mL methanol/water (4/1, V/V)) by ultrasonication for 30 min to obtain a homogeneous suspension. Then, 0.0323 g (i.e. 1.00 mmol) CAP was completely dissolved into above suspension and let them prepolymerize for 60 min. After that, 10 mg AIBN and 3.1 mL (i.e. 20.00 mmol) EGDMA were dispersed into above solution. The resultant mixture was stirred at 60 °C for 24 h under nitrogen protection. Finally, the product was eluted by methanol/acetic acid (9/1, V/V) until no template molecule was detected in the washing solutions. For comparison, MWCNTs@NIP was prepared by the same manner except the absence of CAP. At the same time, other MWCNTs@MIPs were prepared by changing the mole ratio of CAP to C16VimCl (i.e. 1:2, 1:3, 1:5 and 1:6), the mole ratio of CAP to EGDMA (i.e. 1:15, 1:18, 1:22 and 1:25) and the mass ratio of MWCNT to C16VimCl (1:2, 1:2.5, 1:3 and 1:3.2), respectively. 2.4. Preparation of porous r-GO Graphene oxide (GO) was prepared by a modified Hummers method (Liang et al., 2009) and characterized as our previous report (Yang et al., 2014). According to the report (Chen and Qiao 2013), porous GO (P-GO) was obtained by oxidizing and etching GO with KMnO4 and HCl, but we lengthened oxidation time to 12 h in order to acquire bigger pore and better three-dimensional structure. Specifically, 100 mL of 0.5 mg mL 1 GO suspension was mixed with 0.5 g KMnO4 under magnetic stirring for 12 h. The above solution was merged with 30 mL HCl (36%, wt%) and 30 mL H2O2 (30%, wt%) for 3 h. After that, the products were separated by centrifugation, washed with water and dried in a vacuum oven at 60 °C. Next, 200 mL of 0.3 mg mL 1 P-GO suspension was mixed with 2.4 mL ammonia (28%, wt%) and 0.24 mL hydrazine hydrate (85%, wt%), followed by heating at 95 °C for 12 h. Then, the products were separated by centrifugation, washed with water and dried under a vacuum oven at 60 °C.
418
G. Yang, F. Zhao / Biosensors and Bioelectronics 64 (2015) 416–422
2.5. Preparation of MWCNTs@MIP/CKM-3/P-r-GO/GCE The preparation of MWCNTs@MIP/CKM-3/P-r-GO/GCE was shown in Scheme 1 (Part B). Firstly, 2 mg P-r-GO was dispersed in 10 mL N,N-dimethylformamide (DMF) to give 0.2 mg mL 1 P-rGO suspension. At the same time, 2.5 mg CKM-3 and 5 mg MWCNTs@MIP were dispersed in 5 mL DMF respectively, to obtain 0.5 mg mL 1 CKM-3 and 1 mg mL 1 MWCNTs@MIP. Then, 5 mL of 0.2 mg mL 1 P-r-GO suspension was coated on a clean GCE surface and dried at 80 °C. After that, 5 mL of 1 mg mL 1 CKM-3 suspension was coated on the P-r-GO modified GCE and dried at 80 °C. In succession, 5 mL of 1 mg mL 1 MWCNTs@MIP suspension was dropped on above modified electrode and dried at 80 °C. Thus, an MWCNTs@MIP/CKM-3/P-r-GO/GCE was obtained. For comparison, MWCNTs@MIP/r-GO/GCEs (the concentration of r-GO was 1.0 mg mL 1), MWCNTs@MIP/P-r-GO /GCE, MWCNTs@MIP/CKM3/GCE and MWCNTs@MIP/P-r-GO/CKM-3/GCE were prepared under same conditions. 2.6. Electrochemical measurements The sensor was immersed in a PBS (pH ¼7.0) containing a certain concentration of CAP for 200 s, under mildly magnetic stirring. Then, it was rinsed with H2O. After that, the threeelectrode system was assembled on a cell with 10 mL PBS (pH¼7.0), and cyclic voltammogram (CV) or differential pulse voltammogram (DPV) was recorded from 800 mV to 400 mV (vs SCE).
3. Results and discussion 3.1. Morphological and structural characterization Fig. 1 shows the TEM of MWCNTs and MWCNTs@MIP. It was clear that the diameter of MWCNTs was about 10–15 nm (Fig. 1A) and MIP was successfully prepared on MWCNTs surface (B) and its thickness was about 20 nm. At the same time, the FTIR absorption spectra of MWCNTs@MIP also confirmed this point (Fig. S1A). The characteristic peaks of the bonded EGDMA (i.e., 1710 cm 1 (C=O) and 1230/1153 cm 1) (C‒O‒C) were observed. At the same time, the peaks corresponding to the C=N stretching and imidazole ring stretching (1554 and 1460 cm 1) and C=C stretching (1628 cm 1) of C16VimCl (Fig. S1A). Next, the FTIR absorption spectra of P-GO and P-r-GO were recorded (Fig. S1B). Results indicated that the absorption bands of the oxo-groups on P-r-GO were obviously decreased or
disappeared in comparison with P-GO (i.e., 1725 cm 1 (C‒O), 3430, 1393 and 1053 cm 1 (–OH)), meaning that the P-GO was reduced successfully. Then, the TEM of P-r-GO was recorded. It presented three-dimensional and porous structure and the pore diameters were about 20–30 nm (Fig. 2A). The evolution of modified electrode was recorded by using SEM. The P-r-GO modified GCE showed three-dimensional surface (Fig. 2B). Fig. 2C indicates that CKM-3 tightly adhered on P-r-GO surface. When MWCNTs@MIP suspension was coated on the modified electrode, the MWCNTs@MIP showed three-dimensional nanonetwork structure (Fig. 2D). 3.2. Adsorption study Adsorption experiments were used to evaluate the molecularly imprinted effect. Results indicated that the adsorption amount of MWCNTs@MIPs (or MWCNTs@NIP) increased with CAP concentration increasing (Fig. S1C). When the CAP concentration exceeded 1.0 10 4 mol L 1, the adsorption amount of MWCNTs @NIP kept unchanged, meaning that it reached adsorption equilibrium and the value was about 25 μmol g 1. However, for MWCNTs@MIPs prepared with different mole ratio of CAP to C16VimCl (1:1–1:6), when the CAP concentration exceeded 1.4 10 4 mol L 1 the adsorption amount increased from 62 to 91 μmol g 1, indicating that the imprinted effect was obvious. It achieved a maximum when the ratio of CAP to C16VimCl was 1: 4. This was related to the change of the number of available binding sites. When the amount of functional monomer was too little to combine enough template molecules, the available binding sites were less. On the contrary, the MIP film became thick and many binding sites were buried in polymeric matrix so that the adsorption amount decreased. For MWCNTs@MIPs prepared with different mole ratio of CAP to EGDMA (i.e. 1:15–1:25) (Fig. S1D), the adsorption amount of MWCNTs@MIPs increased from 65 to 91 μ mol g 1 and then decreased. It achieved the maximum when the ratio was 1:20. This should be ascribed to the effect of the ratio of CAP to EGDMA on the formation of available binding sites, like the effect of the ratio of CAP to C16VimCl. For the MWCNTs@MIPs prepared with different mole ratios of MWCNT to C16VimCl (i.e. 1:2–1:3.2) (Fig. S1E), the adsorption amount of MWCNTs@MIPs increased from 60 to 91 μmol g 1 and then decreased. It was thought that the number of available binding sites increased with enhancing the amount of C16VimCl. However, when the ratio of MWCNT to C16VimCl was lower than 1:2.8, some binding sites were buried in polymeric matrix due to the increased MIP film thickness. Therefore, the adsorption amount decreased.
Fig. 1. TEM images of MWCNT (A) and MWCNTs@MIP (B).
G. Yang, F. Zhao / Biosensors and Bioelectronics 64 (2015) 416–422
419
Fig. 2. TEM image of P-r-GO (A), SEM images of P-r-GO/GCE (B), CKM-3/P-r-GO/GCE and MWCNTs@MIP/CKM-3/P-r-GO/GCE (C).
3.3. Voltammetric characterization Molecularly imprinted effect of different sensors was also reflected by the voltammetric behavior of CAP (Fig. 3(A and B)). The MWCNTs@MIP/CKM-3/P-r-GO/GCE showed a small background current signal in PBS (Fig. 3A(a)), but when the electrode was soaked in a CAP solution (1.0 10 6 mol L 1) for adsorption,
it exhibited a obvious oxidation peak at about 92 mV and two reduction peaks at 104 mV and 640 mV (Fig. 3A(b)). This was related to the oxidation/reduction reactions of CAP. The peak at 640 mV could be attributed to the reduction of the nitro group of CAP, and the other two peaks were produced by the reversible redox of the reduction product (i.e. hydroxylamine derivative) (Borowiec et al., 2013; Xiao et al., 2007). However, the
Fig. 3. (A) Voltammograms of different electrodes in PBS (pH ¼7.0). Electrodes: (a) MWCNTs@MIP/CKM-3/P-r-GO/GCE in blank PBS; (b) MWCNTs@MIP/CKM-3/P-r-GO/GCE after binding CAP in 1.0 10 6 mol L 1 CAP; (c) MWCNTs@NIP/CKM-3/P-r-GO/GCE after binding CAP in 1.0 10 6 mol L 1 CAP; (B) DPVs of different MIP electrodes in PBS (pH ¼7.0) after binding CAP in 1.0 10 6 mol L 1 CAP. Electrodes: (a) r-GO/GCE; (b) P-r-GO/GCE; (c) CKM-3/GCE; (d) CKM-3/P-r-GO/GCE; (e) P-r-GO/CKM-3/GCE.
420
G. Yang, F. Zhao / Biosensors and Bioelectronics 64 (2015) 416–422
MWCNTs@NIP/CKM-3/P-r-GO/GCE (Fig. 3A(c)) showed small peaks under the same conditions, meaning that the nonspecifically adsorption of CAP was weak and the effect of MIP was very obvious. DPVs of the MIP sensors prepared with different modified interfaces were recorded (Fig. 3B). Results indicated that the peak current of MWCNTs@MIP/P-r-GO/GCE (Fig. 3B(b)) was larger than that of MWCNTs@MIP/r-GO/GCE (B(a)) (r-GO concentration was 1 mg mL 1, which was optimized), meaning the three-dimensional porous film had large surface area and enhanced masstransfer rate. When P-r-GO was replaced by CKM-3, the obtained MWCNTs@MIP/CKM/GCE (B(c)) showed better response than MWCNTs@MIP/P-r-GO/GCE (B(b)). This should be ascribed to the excellence electro-catalysis of CKM-3 for nitrobenzene compounds due to its unique properties (Walcarius, 2012; Zang et al., 2011; Zhang et al., 2013). Compared with other MIP sensors, the MWCNTs@MIP/CKM-3/P-r-GO/GCE showed higher sensitivity (Fig. 3B(d)). However, when CKM-3 was firstly modified and then P-r-GO (i.e. MWCNTs@MIP/P-r-GO/CKM/GCE (Fig. 3B(e))), the CKM-3 was buried by P-r-GO, which made response current decrease in comparison with MWCNTs@MIP/CKM/GCE (B(c)) and MWCNTs@MIP/CKM/P-r-GO/GCE (B(d)). For this strategy, P-r-GO enhanced electrode surface area and benefited the dispersion of CKM-3. When the MWCNTs@MIP was dropped on this interface, it well contacted CKM-3 and CKM-3 could catalyze the reaction of CAP on MWCNTs@MIPs. At the same time, three-dimensional modified electrode benefited mass transfer, enhancing the sensitivity of sensor. 3.4. Optimization of experimental variables 3.4.1. Effect of P-r-GO concentration The P-r-GO concentration was optimized in the range of 0.1–0.3 mg mL 1 for the preparation of sensor (Fig. S2A). Result indicated that the peak current increased with the concentration increasing up to 0.2 mg mL 1, and then gradually decreased. This should be ascribed to the change of electrode surface area and film resistance. Hence, 5 mL of 0.2 mg L 1 P-r-GO was selected for preparing modified electrode. 3.4.2. Effect of CKM-3 concentration The effect of CKM-3 concentration was discussed in the range from 0.1 to 1.0 mg mL 1 (Fig. S2B). It was found that the peak current increased with the concentration increasing, and when the concentration increased to 0.5 mg mL 1, the optimum response value was obtained. Then, the response signal gradually decreased. This was also related to the change of electrode surface area and resistance. In the experiments, 5 mL of 0.5 mg mL 1 CKM-3 was selected. 3.4.3. Effect of mole ratio of CAP to C16VimCl and EGDMA The mole ratio of CAP to C16VimCl was changed from 1:2 to 1:6 to test its effect (Fig. S2C). As a result, the response current increased with the ratio decreasing, and it achieved a maximum at 1:4. Then, the current response gradually decreased. This was related to the change of the number of available binding sites, which was mentioned above (Fig. S1C). Additionally, the thickness of MIP film increased with the ratio decreasing, leading to the decrease of the conductive property. Here, the mole ratio of 1:4 was chosen. At the same time, the effect of the mole ratio of CAP to EGDMA was also considered in the range of 1:10–1:25 (Fig. S2D). Results indicated that the peak current increased with the ratio decreasing, and it achieved a maximum at 1:20. In succession, the current response gradually decreased. This could be explained by the change of the number of available binding sites as mentioned
previously (Fig. S1D). In addition, the conductive property of MWCNTs@MIP became weak with the EGDMA increasing. Therefore, the mole ratio of 1:20 was chosen. 3.4.4. Effect of mass ratio of MWCNT to C16VimCl The mass ratio of MWCNT to C16VimCl was changed from 1:2 to 1:3.2 to test its effect (Fig. S2E). Result indicated that the peak current of CAP increased when the ratio changed from 1:2 to 1:2.8. Further increase caused a gradual decrease in peak current. This was related to the change of density of imprinted sites on MWCNTs (Fig. S1E) and the conductive property of resulting MWCNTs@MIP. When the ratio was smaller the site density was higher, but the resistance of MWCNTs@MIP increased. In this work, the mass ratio of 1:2.8 was chosen. 3.4.5. Effect of MWCNTs@MIP concentration The effect of MWCNTs@MIP concentration was discussed in the range of 0.2–2.0 mg mL 1 (Fig. S2F). The peak current of CAP increased when its concentration changed from 0.2 to 1.0 mg mL 1. Further increase caused a gradual decrease in peak current. The reason was that the adsorption amount of CAP increased with the amount of MWCNTs@MIP increasing, but overmuch MWCNTs@MIP made the resistance increase, which resulted in response current reducing. Therefore, 1 mg mL 1 was a balanced point and was adopted. At the same time, the effect of MWCNTs@MIP concentration on selectivity was also considered. It was found that 50-fold of p-nitrophenol (i.e. 5 10 5 mol L 1) showed weak few interference for the detection of CAP (1.0 10 6 mol L 1) when the MWCNTs@MIP concentration was lower (i.e. 0.2 mg mL 1), but when the concentration was up to 0.5 mg mL 1, the p-nitrophenol almost did not influence the detection (signal change was 2.3%), meaning that the MIP film became enough thick to block the electrode reaction of most p-nitrophenol molecules so that the selectivity was enhanced. 3.4.6. Effect of adsorption time The effect of adsorption time on the response current was also considered in the range of 60–200 s (Fig. S2G). It was found that the peak current increased gradually with extending adsorption time from 60 to 120 s for a 1 10 6 mol L 1 CAP solution (Fig. S2G (a)). Then, it kept almost unchanged, which indicated adsorption of CAP was saturated. However, the saturated adsorption time was up to 140 s for a lower concentration solution (i.e. 2.5 10 8 mol L 1) (Fig. S2G(b)). Thus, 200 s was chosen as adsorption time so that the CAP solutions with lower concentrations had enough time to reach saturated adsorption. 3.4.7. Effect of pH The effect of pH value of supporting electrolyte was investigated over the range from pH 5.5 to pH 8.5 (Fig. S3). It was found that the optimum pH was 7.0. At the same time, the peak potential changed with pH value. The slope of the peak potential vs pH plot was 55.9 mV pH 1, indicating that the number of electron transfer was equal to that of hydrogen ion taking part in the electrode reaction. 3.5. The optimal response characteristics of MIP sensor Under optimal experimental conditions, the DPVs of CAP were recorded (Fig. 4). As expected, the peak current increased upon the increase of CAP concentration. The calibration curve is shown in Fig. 4, and the linear response ranges were 5.0 10 9–5.0 10 7 mol L 1 and 5.0 10 7–4.0 10 6 mol L 1, respectively, and their regression equations were ip (mA)¼ 10.92þ128.18C (mmol L 1), (r¼0.9973) and ip (mA)¼ 62.29þ9.8C (mmol L 1), (r¼0.9921), respectively. The detection
G. Yang, F. Zhao / Biosensors and Bioelectronics 64 (2015) 416–422
421
Table 1 Recovery tests of CAP in milk and honey samples. Samples
Added (μmol L 1) Found (μmol L 1)
RSD (%, n ¼3)
Recovery (%)
Milk 1
0.00 0.025 0.10 1.00 2.50
0.00 0.023 0.094 0.95 2.54
4.3 3.5 2.8 2.5
92 94 95 102
0.00 0.025 0.10 2.50
0.00 0.024 0.91 2.33
5.0 4.2 5.1
96 91 93
Honey 1 0.00 0.025 0.100 1.00 2.50
0.00 0.026 0.095 0.97 2.38
2.4 5.2 4.1 4.3
104 95 97 95
Honey 2 0.00 0.025 0.10 2.50
0.00 0.023 1.04 2.40
3.3 2.2 3.8
92 104 96
Milk 2
Fig. 4. DPVs of the MWCNTs@MIP/CKM-3/P-r-GO/GCE in PBS (pH¼ 7.0) after binding different CAP concentrations. CAP concentrations: 0–4.8 10 6 mol L 1. Inset is the calibration curve of CAP. Error bars represent standard deviation, n¼ 3.
limit was 1.0 10 10 mol L 1 (S/N¼3). Compared with other electrochemical sensors, this sensor offered a reasonable linear range and a lower detection limit for the directly electrochemical sensors (Agüı ́ et al., 2002; Borowiec et al., 2013; Codognoto et al., 2010; Xiao et al., 2007; Zhuang et al., 2014), MIP based sensors (Alizadeh et al., 2012; Zhao et al., 2012) and aptamer based sensor (Pilehvar et al., 2012; Yan et al., 2012). 3.6. Selectivity To assess the selectivity of this method, the interference of some potential organic compounds in real samples was tested. Results indicated that 100-fold of glucose, ascorbic acid, uric acid and glutamic acid did not affect the detection of CAP (1.0 10 6 mol L 1, signal change below 5%). At the same time, the influence of similar compounds (i.e. florfenicol, thiamphenicol, 4-nitrobenzamide and p-nitrophenol) was also investigated. As a result, 50-fold of these compounds had no interference (Fig. S4). This should be ascribed to the effect of MIP, which enhanced the selectivity of sensor. 3.7. Repeatability and stability The repeatability of the sensor was investigated by detecting 1.0 10 6 mol L 1 CAP and a variation coefficient of 3.4% was observed for five successive assays. The current response of the sensor still remained up to 96.4% of its initial value after 15 successive assays (RSD ¼ 2.1%). The stability of the sensor was investigated by measuring its response current to 1.0 10 6 mol L 1 CAP over 20 days. When not in use, the sensor was stored at room temperature. After 20 days, the current response of the sensor remained up to 95.1% (RSD ¼2.8%, n ¼3) of its initial value. The fabrication reproducibility was also estimated with five different electrodes, which were fabricated independently by the same procedure. The RSD was 6.1% for the peak current of 1.0 10 6 mol L 1 CAP, which reflected the reliability of the fabrication procedure.
precipitation (Alizadeh et al., 2012) and extraction of CAP (Xiao et al., 2007). Specifically, 10 mL milk sample and 1 mL of CAP standard solution with different concentrations was added in 7 mL water. Then, 2 mL of trichloroacetic acid (20%, wt%) solution was added to the mixed solution for protein precipitation. The mixture was vortexed for 1 min, and then centrifuged at 4000 rpm for 10 min. The supernatant was collected and filtered through a 0.22 μm membrane filter. After that, the CAP was extracted by using ethyl acetate (the amount is 10 mL for one time) as an extraction liquid for three times. Next, the organic extraction was collected and evaporated to dryness using a gentle stream of nitrogen gas. Afterwards, 10.0 mL PBS (0.10 M PBS) was added and vigorously vortexed for 10 s. Then, the resultant PBS containing CAP was transferred to a cell for the recovery determination. For honey samples, 1.00 g sample and 1 mL of CAP standard solution with different concentrations were added into 9.0 mL water. Then, the treatment was similar to that for milk sample except the protein precipitation. The results are shown in Table 1. The recoveries for CAP standards added was 91–104%, which demonstrates that the sensor may provide a useful tool for determining CAP in real samples.
4. Conclusions In this paper, a facile method was used to prepare MWCNTs@MIP for the electrochemical determination of CAP. C16VimCl not only acted as novel functional monomer, but also promoted the dispersion of MWCNT as it tightly adhered on MWCNT surface. At the same time, three-dimensional P-r-GO and CKM-3 were applied to modify GCE to enhance the response signal. The resulting sensor exhibited sensitive and selective response to CAP and was applied to the detection of CAP in real samples. This work provided a useful platform for the preparation of MWCNTs@MIP based sensors.
3.8. Detection of real samples The proposed MIP electrochemical sensor was evaluated by performing recovery tests for CAP in milk and honey samples. Firstly, the milk samples were treated according to the two reports (Alizadeh et al., 2012; Xiao et al., 2007), which involved protein
Acknowledgements This work was jointly supported by the National Natural Science Foundation of China (Grant nos. 21075092 and 21277105) and the
422
G. Yang, F. Zhao / Biosensors and Bioelectronics 64 (2015) 416–422
Youth Foundation of Science Commission of Yunnan Province (Grant no. 2012FD061).
Appendix A. Supplementary information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.bios.2014.09.041.
References Agüı,́ L., Guzmán, A., Yáñez-Sedeño, P., Pingarrón, J.M., 2002. Anal. Chim. Acta 461, 65–73. Alizadeh, T., Ganjali, M.R., Zare, M., Norouzi, P., 2012. Food Chem. 130, 1108–1114. Anirudhan, T.S., Alexander, S., 2014. Appl. Surf. Sci. 303, 180–186. Bali Prasad, B., Jauhari, D., Tiwari, M.P., 2013. Biosens. Bioelectron. 50, 19–27. Borowiec, J., Wang, R., Zhu, L., Zhang, J., 2013. Electrochim. Acta 99, 138–144. Chen, H., Chen, H., Ying, J., Huang, J., Liao, L., 2009. Anal. Chim. Acta 632, 80–85. Chen, L., Xu, S., Li, J., 2011. Chem. Soc. Rev. 40, 2922–2942. Chen, S., Qiao, S., 2013. ACS Nano 7, 10190–10196. Cheong, W.J., Yang, S.H., Ali, F., 2013. J. Sep. Sci. 36, 609–628. Chi, W., Shi, H., Shi, W., Guo, Y., Guo, T., 2012. J. Hazard. Mater. 227–228, 243–249. Chullasat, K., Kanatharana, P., Limbut, W., Numnuam, A., Thavarungkul, P., 2011. Biosens. Bioelectron. 26, 4571–4578. Codognoto, L., Winter, E., Doretto, K.M., Monteiro, G.B., Rath, S., 2010. Microchim. Acta 169, 345–351. Falagas, M.E., Grammatikos, A.P., Michalopoulos, A., 2008. Expert Rev. Anti-Infect. Ther. 6, 593–600. Jana, S., Vasantha, V.A., Stubbs, L.P., Parthiban, A., Vancso, J.G., 2013. J. Polym. Sci. Part A: Polym. Chem. 51, 3260–3273. Lee, H.Y., Kim, B.S., 2009. Biosens. Bioelectron. 25, 587–591. Liang, Y., Wu, D., Feng, X., Müllen, K., 2009. Adv. Mater. 21, 1679–1683.
Liu, B., Zhang, B., Chen, G., Tang, D., 2013. Microchim. Acta 181, 257–262. Moreira, F.T., Dutra, R.A., Noronha, J.P., Cunha, A.L., Sales, M.G., 2011. Biosens. Bioelectron., 243–250. Pilehvar, S., Mehta, J., Dardenne, F., Robbens, J., Blust, R., De Wael, K., 2012. Anal. Chem. 84, 6753–6758. Prasad, B.B., Madhuri, R., Tiwari, M.P., Sharma, P.S., 2010. Electrochim. Acta 55, 9146–9156. Qian, T., Yu, C., Zhou, X., Ma, P., Wu, S., Xu, L., Shen, J., 2014. Biosens. Bioelectron. 58, 237–241. Sharma, P.S., D'Souza, F., Kutner, W., 2012. TrAC—Trends Anal. Chem. 34, 59–77. Sharma, P.S., Dabrowski, M., D'Souza, F., Kutner, W., 2013. TrAC—Trends Anal. Chem. 51, 146–157. Shen, J., Xia, X., Jiang, H., Li, C., Li, J., Li, X., Ding, S., 2009. J. Chromatogr. B 877, 1523–1529. Tong, Y., Li, H., Guan, H., Zhao, J., Majeed, S., Anjum, S., Liang, F., Xu, G., 2013. Biosens. Bioelectron. 47, 553–558. Turton, J.A., Andrews, C.M., Havard, A.C., Williams, T.C., 2002. Int. J. Exp. Pathol. 83, 225–238. Walcarius, A., 2012. TrAC—Trends Anal. Chem. 38, 79–97. Xiao, F., Zhao, F., Li, J., Yan, R., Yu, J., Zeng, B., 2007. Anal. Chim. Acta 596, 79–85. Yadav, S.K., Agrawal, B., Chandra, P., Goyal, R.N., 2014. Biosens. Bioelectron. 55, 337–342. Yan, L., Luo, C., Cheng, W., Mao, W., Zhang, D., Ding, S., 2012. J. Electroanal. Chem. 687, 89–94. Yang, G., Zhao, F., Zeng, B., 2014. Electrochim. Acta 135, 154–160. Zaidi, S.A., 2013. Int. J. Electrochem. Sci. 8, 9936–9955. Zang, J., Guo, C.X., Hu, F., Yu, L., Li, C.M., 2011. Anal. Chim. Acta 683, 187–191. Zhang, D., Yu, D., Zhao, W., Yang, Q., Kajiura, H., Li, Y., Zhou, T., Shi, G., 2012. Analyst 137, 2629–2636. Zhang, N., Xiao, F., Bai, J., Lai, Y., Hou, J., Xian, Y., Jin, L., 2011. Talanta 87, 100–105. Zhang, T., Lang, Q., Yang, D., Li, L., Zeng, L., Zheng, C., Li, T., Wei, M., Liu, A., 2013. Electrochim. Acta 106, 127–134. Zhao, H., Chen, Y., Tian, J., Yu, H., Quan, X., 2012. J. Electrochem. Soc. 159, J231–J235. Zhao, L., Zhao, F., Zeng, B., 2014. Biosens. Bioelectron. 60, 71–76. Zhuang, Y., Cai, L., Cao, G., 2014. J. Electrochem. Soc. 161, H129–H132.
