Biosensors and Bioelectronics 74 (2015) 85–90
Contents lists available at ScienceDirect
Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios
A high-performance nonenzymatic piezoelectric sensor based on molecularly imprinted transparent TiO2 film for detection of urea Zheng-peng Yang a, Xuan Liu a, Chun-jing Zhang a,b,n, Bao-zhong Liu a,nn a b
Institute of Materials Science and Engineering, Henan Polytechnic University, Jiaozuo 454000, China College of Chemistry and Chemical Engineering, Central South University, Changsha 410083, China
art ic l e i nf o
a b s t r a c t
Article history: Received 18 March 2015 Received in revised form 21 May 2015 Accepted 9 June 2015 Available online 19 June 2015
Transparent photocatalytic surfaces are of ever increasing importance for the enhancement of the photocatalytic efficiency. Here, the highly ordered transparent TiO2 nanotube arrays were prepared by the anodization and thermal annealing of titanium layer deposited onto the glass substrate, and a novel nonenzymatic piezoelectric sensor was developed for urea detection based on the modification of molecularly imprinted TiO2 thin film onto transparent TiO2 nanotube arrays. The performance of the fabricated sensor was evaluated and the results indicated that the sensor exhibited high sensitivity in urea detection, with a linear range from 0.04 to 120 μM and a limit of detection of 0.01 μM. Moreover, the sensor presented outstanding selectivity while used in coexisting systems containing various interferents with high concentration. The analytical application of the urea sensor confirmed the feasibility of urea detection in urine sample. & 2015 Elsevier B.V. All rights reserved.
Keywords: Urea sensor Molecular imprinting Transparent TiO2 nanotube Piezoelectric sensing Photocatalysis
1. Introduction Urea, a main end product of protein metabolism, can act as an important indicator of liver and kidney function (Gutierrez et al., 2007; Ciana and Caputo, 1996). A physiological urea level is in a range from 2.5 to 7.5 mM, depending on the build and relative health of the body (Dhawan et al., 2009). A decreased urea level can be associated with severe liver disease or insufficient protein intake, and a high urea concentration can cause renal dysfunction, urinary tract obstruction, dehydration, shock and gastrointestinal bleeding (Soldatkin et al., 2014; Ahmad et al., 2014). Therefore, it is very important to monitor the level of urea to determine the health of the livers and kidneys in the human body. In the past years, some analytical methods have been developed for urea detection, such as enzymatic assay (Hamilton and Breslin, 2014; Yang and Zhang, 2013; Siqueira et al., 2014; Lin et al., 2013), gas chromatography (Kessler and Siekmann, 1999), calorimetry (Francis et al., 2002), high-performance liquid chromatography (Zhang et al., 2014), fluorimetry (Roch-Ramel, 1967), surface plasmon resonance (Verma and Gupta, 2014) and electrochemistry (Mondal and Sangaranarayanan, 2013; Srivastava et al., 2012). Despite many advances in urea detection, many of these methods n Corresponding author at: Institute of Materials Science and Engineering, Henan Polytechnic University, Jiaozuo 454000, China. nn Corresponding author. E-mail addresses: [email protected] (C.-j. Zhang), [email protected] (B.-z. Liu).
http://dx.doi.org/10.1016/j.bios.2015.06.022 0956-5663/& 2015 Elsevier B.V. All rights reserved.
still do not meet the growing demand for more sensitive and selective detection of urea. Thus, the further development of simple, facile and stable methods with high sensitivity and selectivity for urea detection is highly desirable. TiO2 is an ever growing research field due to its remarkable photocatalytic activity, chemical stability, structural rigidity, excellent biocompatibility and relatively low price (Yang et al., 2011). It is wellknown that the properties and performance of TiO2 partly depend on its crystallinity and morphology (Li et al., 2009). Currently, the highly ordered transparent TiO2 nanotube arrays have attracted considerable scientific interest in such fields as solar energy conversion, electrochromic device, antireflection coating and selfcleaning applications because the arrays possess highly active surfaces with a large surface to volume ratio, aligned tubular nature, high transparency and photocatalytic activity (Ji et al., 2010; Berger et al., 2009; Mor et al., 2005). Still, the photocatalysis of transparent TiO2 nanotube lacks selectivity due to the nature of hydroxyl radicals or holes generated in photocatalytic oxidation, which will limit its application (TranT. et al., 2014). Molecular imprinting technique, the design and construction of mimetic receptor system with predetermined recognition for target molecule, has been proposed and developed rapidly (Yang et al., 2014). Recently, TiO2 sol–gel materials have been used to prepare molecularly imprinted thin film, which was used for the detection and separation of organisms (Zhang et al., 2015; Liu et al., 2010). The electrode-separated piezoelectric sensor (ESPS) is sensitive sensor and can monitor the frequency shift caused by conductive change of solution (Lamas-Ardisana and Costa-Garcia, 2006; de Jesus
86
Z.-p. Yang et al. / Biosensors and Bioelectronics 74 (2015) 85–90
et al., 2001). The frequency response of ESPS depends on the highfrequency excitation electricfield, which is applied to quartz crystal via solution conduction and closely related to the conductivity of solution. The ESPS does not produce thermal effects for the biological materials and exhibits high sensitivity and accuracy (Shen et al., 1993). In this work, molecularly imprinted TiO2 (MIT) was selected as sensing film and the ESPS/flow injection analysis (FIA) system was employed to monitor the photocatalytic reaction. A novel nonenzymatic piezoelectric sensor has been developed for the detection of urea based on the modification of molecularly imprinted TiO2 thin film onto transparent TiO2 nanotube arrays. The constructed ESPS sensor with high sensitivity, selectivity and stability showed excellent performance for urea detection.
2. Experimental 2.1. Materials and apparatus High purity titanium (99.99%) was purchased from Sigma-Aldrich. Indium tin oxide (ITO) glass was obtained from Shenzhen Jianshengda Technology Co. Ltd. Glycerol, NH4F and Ti(O-nBu)4 were purchased from Sinopharm Co. Urea, glucose, uric acid, creatinine, thiourea, lactic acid, glycine and alanine were purchased from Shanghai Aladdin Chemical Co., and used without further purification. All other chemicals were of analytical grade and deionized (DI) water was used throughout the experiments. Piezoelectric measurements were performed in an ESPS/FIA sensing system which was described in our earlier report (Yang et al., 2007). The setup of the detection system is illustrated in Fig. 1. A MIT/TiO2 nanotube/ITO/glass electrode with an effective working area of 2.0 2.0 cm2 was used as the separated electrode. UV–vis transmission spectra were recorded by an UV/Vis spectrophotometer (UV-1601, Shimadzu, Japan). Scanning electron micrograph (SEM) was carried out on a Hitachi S-520 scanning electron microscope (Hitachi Ltd., Tokyo, Japan) working at 20 kV. X-ray diffraction (XRD) was performed on a Rigaku D/MAX-RC X-ray diffractometer using Cu Kα radiation. Raman spectra were recorded on a Raman spectrometer (InVia, Renishaw Co., UK). A 320 W low-pressure mercury lamp (Philips, 254 nm) was used as UV irradiation source.
2.2. Fabrication of the urea sensor Titanium thin films of 1 mm thickness were deposited onto ITOcoated glass substrates by RF sputtering. RF power of 200 W was applied during the deposition process. The sputtering was conducted under an Ar gas pressure of 2.7 10 3 Torr (base pressure 4.5 10 7 Torr) at 250 °C with a deposition rate of 8 nm/min. Prior to electrochemical treatment, the obtained Ti/ITO/glass electrodes were ultrasonically cleaned in ethanol and DI water for several minutes. Subsequently, Anodization was performed at 20 V in the high-purity glycerol containing 0.7 M NH4F and 2.5 vol% water using a platinum cathode. The sputtered titanium films on the glass substrates were transformed to nanotubular layers after three hours anodization at room temperature. Finally, the obtained nanotubular arrays were annealed at 500 °C for 2 h to transform completely into an anatase structure. Thus, TiO2 nanotube/ITO/ glass electrode was fabricated. The formation of MIT layer on TiO2 nanotube/ITO/glass electrode was performed by sol–gel hydrolysis of Ti(O-nBu)4 followed by calcination treatment. Specifically, Ti(O-nBu)4 (0.5 mmol) and urea (0.2 mmol) were dissolved in 8 mL of toluene-ethanol (v/v, 2:1) mixture, and then vigorously stirred at room temperature. Subsequently, the solution was diluted by 20 times with watersaturated toluene and continuously stirred for 5 h to get a sol–gel solution which was used as a dipping solution. The obtained TiO2 nanotube/ITO/glass electrode was immersed into the dipping solution for 6 min at room temperature, rinsed with toluene, hydrolyzed in water and dried in N2 atmosphere. The “dipping-rinsehydrolyzation-drying”cycle was repeated three times, and then the TiO2 nanotube/ITO/glass electrode whose surface was covered with a TiO2 layer containing the template molecule of urea was obtained. Finally, the TiO2 nanotube/ITO/glass electrode modified by TiO2 layer was dried at 80 °C for 2 h, and then calcined in air at 450 °C for 2 h to remove template molecules and crystalline the TiO2, resulting in the formation of MIT/TiO2 nanotube/ITO/glass electrode. Thus, the urea sensor was fabricated. For comparison, the non-imprinted TiO2 modified TiO2 nanotube/ITO/glass electrode was were prepared using the same procedure without template molecules, and was denoted as NIT/TiO2 nanotube/ITO/ glass electrode.
Fig. 1. Schematic illustration for setup and detection mechanism of the ESPS/FIA system.
Z.-p. Yang et al. / Biosensors and Bioelectronics 74 (2015) 85–90
2.3. Piezoelectric measurements The piezoelectric measurements of the obtained urea sensor were carried out in 0.01 mM phosphate buffer solution (PBS, pH 7.4) at 35 °C. After mounting MIT/TiO2 nanotube/ITO/glass electrode and quartz crystal disc into the flow cell, urea solution was injected into the detection system. The UV illumination was performed after incubation for 15 min, and the time-dependent change in the frequency was recorded by a frequency counter and stored in a microcomputer.
3. Results and discussion 3.1. Characterization of the piezoelectric urea sensor Fig. 2A shows the digital image of the obtained MIT/TiO2 nanotube/ITO/glass electrode. A high definition of the printed image underneath the layer was observed, indicating an excellent transparency of the electrode. The transparency was further investigated by UV-visible spectroscopy. Fig. 2B shows the UV–vis transmission spectra of the ITO/glass and MIT/TiO2 nanotube/ITO/
Fig. 2. (A) Digital image of the MIT/TiO2 nanotube coated on ITO/glass substrate. The text “HPU” is written underneath on paper to demonstrate the degree of transparency. (B) Transmittance spectra of (a) ITO/glass and (b) MIT/TiO2 nanotube/ ITO/glass.
87
glass. In the visible region, the average transmittance of MIT/TiO2 nanotube/ITO/glass reached about 67% to be compared with 79% of ITO/glass. A sharp decline in transmittance for wavelengths of less than 400 nm was clearly observed for the two substrates due to the strong absorption of glass and TiO2 in the ultraviolet region. Compared with the ITO/glass, a small decrease in transmittance of MIT/TiO2 nanotube/ITO/glass under the same wavelength can be attributed to the presence of MIT/TiO2 nanotube on ITO/glass. The surface morphology of the MIT/TiO2 nanotube/ITO/glass electrode was investigated by SEM. As presented in Fig. 3A, the electrode surface was actually rough, and was composed of highdensity, well-ordered and uniform nanopores. From the enlarged SEM images (the inset in Fig. 3A), the average pore size was about 40 nm. For comparison, the SEM observation of TiO2 nanotube/ ITO/glass electrode was also performed (not shown in the graph). The pore size was increased from 40 to 60 nm, indicating a successful modification of MIT on the surface of TiO2 nanotube arrays. The crystal structure and composition of electrode were identified by XRD. As seen in Fig. 3B, the diffraction peaks at 2θ ¼30.5°, 35.5°, 37.4°, 45.5°, 50.8° and 60.4° can be assigned to those XRD patterns of the remaining ITO coating (Berger et al., 2009). An amorphous shoulder can be related to the glass substrate. Several relatively narrow diffraction peaks at 2θ ¼ 25.4°, 47.8°, 53.7°, 54.8° and 62.4° can be perfectly indexed to the typical crystal planes of anatase, revealing the formation of anatase TiO2 on the ITO/glass substrate. Additionally, No diffraction peaks of titanium could be observed in the spectra of electrode, which suggested that the sputtered titanium thin film had been consumed and completely transformed into an anatase structure after the anodization and annealing treatment. Fig. 3C shows the Raman spectra taken for the MIT/TiO2 nanotube samples after annealing treatment. Ti–O–Ti signals in the range of 400–700 cm 1 are the characteristic of anatase and rutile structures (Bersani et al., 1998). Typically, several relatively strong vibration peaks were clearly observed at 145, 196, 398, 510 and 632 cm 1, which can be attributed to the Raman active modes of anatase phase (Hardwick et al., 2007). Thus, The Raman analysis further confirmed the existence of anatase TiO2 on the electrode surface. The formation of MIT film on the TiO2 nanotube/ITO/glass substrate was further evaluated using piezoelectric measurements. Fig. 3D shows the frequency responses of ESPS/FIA detection system with TiO2 nanotube/ITO/glass electrode, NIT/TiO2 nanotube/ ITO/glass electrode and MIT/TiO2 nanotube/ITO/glass electrode in 60 mM urea solution. The frequency change on MIT/TiO2 nanotube/ ITO/glass electrode reached 60 Hz, which was 4.6 times that on TiO2 nanotube/ITO/glass electrode and 3.0 times that on NIT/TiO2 nanotube/ITO/glass electrode. Significantly, the imprinted electrode exhibited much larger response compared to the TiO2 nanotube and nonimprinted electrodes. The larger frequency change can be attributed to the presence of MIT film on the electrode surface. The imprinted sites on the surface of MIT/TiO2 nanotube/ ITO/glass electrode can specifically rebind and accumulate urea molecules on the electrode surface. As a result, more urea molecules were photocatalytically degradated on the MIT/TiO2 nanotube/ITO/glass electrode. According to the report in literature (Kondo and Jardim, 1991; Zhang, 2008), the photodegradation products of urea on anatase TiO2 are mainly NO−3 and CO2− 3 . Thus, the initial uncharged urea molecules are transformed into charged ions, causing the increase of conductivity (namely the decrease in frequency response) in the flow cell. In addition, our study indicated that a steady-state response could be reached within 2 s. The response time is shorter than that reported in literature (Yang and Zhang, 2013; Srivastava et al., 2012).
88
Z.-p. Yang et al. / Biosensors and Bioelectronics 74 (2015) 85–90
Fig. 3. (A) Surface-view SEM image of the MIT/TiO2 nanotube coated on ITO/glass. The inset shows the enlarged view. (B) XRD spectra of MIT/TiO2 nanotube coated on ITO/ glass. (C) Raman spectra of MIT/TiO2 nanotube. (D) Frequency responses of ESPS/FIA detection system with TiO2 nanotube/ITO/glass electrode, NIT/TiO2 nanotube/ITO/glass electrode and MIT/TiO2 nanotube/ITO/glass electrode. 60 mM urea, pH 7.4, 35 °C. Average of three measurements (mean 7S.D.).
3.2. Biosensing performance The dynamic range of the urea sensor was studied by monitoring the urea solution in the concentration range from 0.01 to 150 mM. As seen in Fig. 4, the frequency response of the urea sensor increased with increasing urea concentration, and was linear in the concentration range of 0.04 to 120 μM (see the inset in Fig. 4). The linear regression equation was y (Hz) ¼ 12.05–0.76 [urea] (μM) with a correlation coefficient of 0.9979. The limit of detection (LOD) was found to be about 0.01 μM (S/N ¼3), which was lower than that reported by other researchers (Siqueira et al., 2014; Mondal and Sangaranarayanan, 2013; Syu and Chang, 2009). Obviously, the proposed electrode shows promise for application in the monitoring of urea with its low LOD and rapid response. Anti-interference properties are important considerations for sensors. In this work, glucose, uric acid, creatinine, thiourea, lactic acid, glycine and alanine were selected as interfering substances. After incubation for 15 min, the frequency response was recorded under UV illumination. The selectivity of the urea sensor was evaluated by calculating the frequency change ratio (ΔF/ΔF0). Herein, ΔF0 was the frequency change for 60 μM urea (S0) and was set as 100%, while ΔF referred to the frequency change of the two-component system containing 60 μM UA and 60 μM interfering substances (S1–S7). As seen in Fig. 5, these interferents
Fig. 4. Calibration graph obtained for the urea sensor in a urea concentration range from 0.01 to 150 mM. The inset shows the linear part of the calibration graph.
showed a slight effect on the frequency change of urea, and the deviations were all less than 4.5%, which adequately revealed that the urea sensor favored urea much more than other molecules and
Z.-p. Yang et al. / Biosensors and Bioelectronics 74 (2015) 85–90
89
practical applications, the proposed sensor was used to measure urea in urine samples. Prior to analysis, the urine samples were diluted 1000 times with PBS (pH 7.4) for spiked recovery tests, without other pretreatment. The recovery test was performed by spiking with different concentrations of urea to diluted urine samples. The parallel experiment was carried out three times. As showed in Table 1, with six different additions of urea to the urine samples containing 4.8 mM urea determined by the sensor, the obtained recoveries ranged from 95 to 102%. The good recovery rate indicates that the urea sensor is feasible for practical analysis.
4. Conclusions
Fig. 5. Frequency change ratio of urea sensor in 60 μM UA in the absence (ΔF0) and presence (ΔF) of 60 μM glucose (S1), uric acid (S2), creatinine (S3), thiourea (S4), lactic acid (S5), glycine (S6) and alanine (S7), respectively.
Table 1 Recovery test of the urea sensor. Urea concentration (μM) Added
0.2 1.0 5.0 20.0 40.0 80.0 a b
A novel nonenzymatic piezoelectric urea sensor was fabricated by a convenient attachment of MIT on highly ordered transparent TiO2 nanotube arrays. The constructed sensor exhibited high sensitivity, rapid response (o 2 s), wide linear range (0.04– 120 mM), low detection limit (0.01 mM), good selectivity and stability in the ESPS/FIA system. Detection of urea in urine sample was also performed with satisfactory results. The proposed sensor shows a promising application in monitoring of biomolecules and offers a new avenue to broaden the applications of transparent TiO2 nanotube arrays in sensors.
Recovery (%)
Acknowledgments
Found Meana
R.S.D.b (%)
0.19 1.0 5.1 19.9 38.8 80.4
3.1 2.6 4.0 2.9 5.7 2.8
95.0 100.0 102.0 99.5 97.0 100.5
Average of three determinations. Relative standard deviation.
therefore presented outstanding selectivity towards urea. The superior selectivity can be attributed to the recognition function of surface molecularly imprinted modification. The imprinted cavities formed in MIT film can distinguish urea from other species through molecular size and functional group distribution, and rebind urea selectively. Thus, urea molecules were specifically accumulated on the sensor surface, whereas other coexistent molecules not complementary to the cavities mainly remained in the bulk solution. The reproducibility of the same urea sensor was evaluated by measuring the response signal twenty times in a continuous manner. After each injection of 60 mM urea solution and the attainment of equilibrium, the urea sensor was recovered by washing with DI water under UV illumination until the frequency response reached a steady value. The sensor can be stored dry at room temperature and reused for next experiments. The obtained results indicated that the relative standard deviation of frequency response was only 1.31%, indicating a good reproducibility of the sensor. The long-term stability revealed that after storage of five months at room temperature, the frequency change decreased by only 2.07% compared to the initial frequency response. The obtained results indicated the high stability of the MIT/TiO2 nanotube/ITO/glass electrode for urea sensing. The rigid structure of MIT, the strong binding of MIT on TiO2 nanotube/ITO/glass may be responsible for the excellent reproducibility and stability. In order to evaluate the reliability of the developed sensor in
This work was supported by the National Natural Science Foundation of China (No. 51202059), the Research Foundation for Youth Scholars of Higher Education of Henan Province (No. 2013GGJS-047), the Fundamental Research Funds for the Universities of Henan Province (NSFRF140201, NSFRF140601), the Key Foundation of He'nan Educational Committee (No. 14A430014) and Foundation for Distinguished Young Scientists of Henan Polytechnic University (J2014-05).
References Ahmad, R., Tripathy, N., Hahn, Y.B., 2014. Sens. Actuators B 194, 290–295. Berger, S., Ghicov, A., Nah, Y.C., Schmuki, P., 2009. Langmuir 25, 4841–4844. Bersani, D., Antonioli, G., Lottici, P.P., Lopez, T.J., 1998. Non-Cryst. Solids 232, 175–181. Ciana, L.D., Caputo, G., 1996. Clin. Chem. 42, 1079–1085. de Jesus, D.P., Neves, C.A., do Lago, C.L., 2001. J. Brazil. Chem. Soc. 12, 123–128. Dhawan, G., Sumana, G., Malhotra, B.D., 2009. Biochem. Eng. J. 44, 42–52. Francis, P.S., Lewis, S.W., Lim, K.F., 2002. Trends Anal. Chem. 21, 389–400. Gutierrez, M., Alegret, S., Valle, M., 2007. Biosens. Bioelectron. 22, 2171–2178. Hamilton, A., Breslin, C.B., 2014. Electrochim. Acta 145, 19–26. Hardwick, L.J., Holzapfel, M., Novak, P., Dupont, L., Baudrin, E., 2007. Electrochim. Acta 52, 5357–5367. Ji, Y.J., Lin, K.C., Zheng, H.G., Liu, C.C., Dudik, L., Zhu, J.J., Burda, C., 2010. ACS Appl. Mater. Inter. 2, 3075–3082. Kessler, A., Siekmann, L., 1999. Clin. Chem. 45, 1523–1529. Kondo, M.M., Jardim, W.F., 1991. Water Res. 25, 823–827. Lamas-Ardisana, P.J., Costa-Garcia, A., 2006. Sens. Actuators B 115, 567–574. Li, S.Q., Zhang, G.M., Guo, D.Z., Yu, L.G., Zhang, W.J., 2009. Phys. Chem. C 113, 12759–12765. Lin, Y.H., Wang, S.H., Wu, M.H., Pan, T.M., Lai, C.S., Luo, J.D., Chiou, C.C., 2013. Biosens. Bioelectron. 43, 328–335. Liu, Y.T., Liu, R.H., Liu, C.B., Luo, S.L., Yang, L.X., Sui, F., Teng, Y.R., Yang, R.B., Cai, Q.Y., 2010. J. Hazard. Mater. 182, 912–918. Mondal, S., Sangaranarayanan, M.V., 2013. Sens. Actuators B 177, 478–486. Mor, G.K., Varghese, O.K., Paulose, M., Grimes, C.A., 2005. Adv. Funct. Mater. 15, 1291–1296. Roch-Ramel, F., 1967. Anal. Biochem. 21, 372–381. Shen, D.Z., Zhu, W.H., Nie, L.H., Yao, S.Z., 1993. Anal. Chim. Acta 276, 87–97. Siqueira, J.R., Molinnus, D., Beging, S., Schoning, M.J., 2014. Anal. Chem. 86, 5370–5375. Soldatkin, O.O., Kucherenko, I.S., Marchenko, S.V., Ozansoy Kasap, B., Akata, B., Soldatkin, A.P., Dzyadevych, S.V., 2014. Mater. Sci. Eng. C 42, 155–160. Srivastava, R.K., Srivastava, S., Narayanan, T.N., Mahlotra, B.D., Vajtai, R., Ajayan, P.
90
Z.-p. Yang et al. / Biosensors and Bioelectronics 74 (2015) 85–90
M., Srivastava, A., 2012. ACS Nano 6, 168–175. Syu, M.J., Chang, Y.S., 2009. Biosens. Bioelectron. 24, 2671–2677. TranT., T., Li, J.Z., Feng, H., Cai, J., Yuan, L.J., Wang, N.Y., Cai, Q.Y., 2014. Sens. Actuators B 190, 745–751. Verma, R., Gupta, B.D., 2014. Analyst 139, 1449–1455. Yang, Z.P., Shang, X.G., Zhang, C.J., Zhu, J.P., 2014. Sens. Actuators B 201, 167–172. Yang, Z.P., Si, S.H., Dai, H.J., Zhang, C.J., 2007. Biosens. Bioelectron. 22, 3283–3287. Yang, Z.P., Yan, J.L., Zhang, C.J., Luo, S.Q., 2011. Colloid. Surf. B 87, 187–191.
Yang, Z.P., Zhang, C.J., 2013. Sens. Actuators B 188, 313–317. Zhang, C.J., Si, S.H., Yang, Z.P., 2015. Biosens. Bioelectron. 65, 115–120. Zhang, J., Liu, G.X., Zhang, Y., Gao, Q., Wang, D.P., Liu, H., 2014. J. Agric. Food Chem. 62, 2797–2802. Zhang, S.Y., 2008. Preliminary Study on Urea Photolysis (Master's Degree Dissertation). Huazhong Agricultural University, Wuhan.
Contents lists available at ScienceDirect
Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios
A high-performance nonenzymatic piezoelectric sensor based on molecularly imprinted transparent TiO2 film for detection of urea Zheng-peng Yang a, Xuan Liu a, Chun-jing Zhang a,b,n, Bao-zhong Liu a,nn a b
Institute of Materials Science and Engineering, Henan Polytechnic University, Jiaozuo 454000, China College of Chemistry and Chemical Engineering, Central South University, Changsha 410083, China
art ic l e i nf o
a b s t r a c t
Article history: Received 18 March 2015 Received in revised form 21 May 2015 Accepted 9 June 2015 Available online 19 June 2015
Transparent photocatalytic surfaces are of ever increasing importance for the enhancement of the photocatalytic efficiency. Here, the highly ordered transparent TiO2 nanotube arrays were prepared by the anodization and thermal annealing of titanium layer deposited onto the glass substrate, and a novel nonenzymatic piezoelectric sensor was developed for urea detection based on the modification of molecularly imprinted TiO2 thin film onto transparent TiO2 nanotube arrays. The performance of the fabricated sensor was evaluated and the results indicated that the sensor exhibited high sensitivity in urea detection, with a linear range from 0.04 to 120 μM and a limit of detection of 0.01 μM. Moreover, the sensor presented outstanding selectivity while used in coexisting systems containing various interferents with high concentration. The analytical application of the urea sensor confirmed the feasibility of urea detection in urine sample. & 2015 Elsevier B.V. All rights reserved.
Keywords: Urea sensor Molecular imprinting Transparent TiO2 nanotube Piezoelectric sensing Photocatalysis
1. Introduction Urea, a main end product of protein metabolism, can act as an important indicator of liver and kidney function (Gutierrez et al., 2007; Ciana and Caputo, 1996). A physiological urea level is in a range from 2.5 to 7.5 mM, depending on the build and relative health of the body (Dhawan et al., 2009). A decreased urea level can be associated with severe liver disease or insufficient protein intake, and a high urea concentration can cause renal dysfunction, urinary tract obstruction, dehydration, shock and gastrointestinal bleeding (Soldatkin et al., 2014; Ahmad et al., 2014). Therefore, it is very important to monitor the level of urea to determine the health of the livers and kidneys in the human body. In the past years, some analytical methods have been developed for urea detection, such as enzymatic assay (Hamilton and Breslin, 2014; Yang and Zhang, 2013; Siqueira et al., 2014; Lin et al., 2013), gas chromatography (Kessler and Siekmann, 1999), calorimetry (Francis et al., 2002), high-performance liquid chromatography (Zhang et al., 2014), fluorimetry (Roch-Ramel, 1967), surface plasmon resonance (Verma and Gupta, 2014) and electrochemistry (Mondal and Sangaranarayanan, 2013; Srivastava et al., 2012). Despite many advances in urea detection, many of these methods n Corresponding author at: Institute of Materials Science and Engineering, Henan Polytechnic University, Jiaozuo 454000, China. nn Corresponding author. E-mail addresses: [email protected] (C.-j. Zhang), [email protected] (B.-z. Liu).
http://dx.doi.org/10.1016/j.bios.2015.06.022 0956-5663/& 2015 Elsevier B.V. All rights reserved.
still do not meet the growing demand for more sensitive and selective detection of urea. Thus, the further development of simple, facile and stable methods with high sensitivity and selectivity for urea detection is highly desirable. TiO2 is an ever growing research field due to its remarkable photocatalytic activity, chemical stability, structural rigidity, excellent biocompatibility and relatively low price (Yang et al., 2011). It is wellknown that the properties and performance of TiO2 partly depend on its crystallinity and morphology (Li et al., 2009). Currently, the highly ordered transparent TiO2 nanotube arrays have attracted considerable scientific interest in such fields as solar energy conversion, electrochromic device, antireflection coating and selfcleaning applications because the arrays possess highly active surfaces with a large surface to volume ratio, aligned tubular nature, high transparency and photocatalytic activity (Ji et al., 2010; Berger et al., 2009; Mor et al., 2005). Still, the photocatalysis of transparent TiO2 nanotube lacks selectivity due to the nature of hydroxyl radicals or holes generated in photocatalytic oxidation, which will limit its application (TranT. et al., 2014). Molecular imprinting technique, the design and construction of mimetic receptor system with predetermined recognition for target molecule, has been proposed and developed rapidly (Yang et al., 2014). Recently, TiO2 sol–gel materials have been used to prepare molecularly imprinted thin film, which was used for the detection and separation of organisms (Zhang et al., 2015; Liu et al., 2010). The electrode-separated piezoelectric sensor (ESPS) is sensitive sensor and can monitor the frequency shift caused by conductive change of solution (Lamas-Ardisana and Costa-Garcia, 2006; de Jesus
86
Z.-p. Yang et al. / Biosensors and Bioelectronics 74 (2015) 85–90
et al., 2001). The frequency response of ESPS depends on the highfrequency excitation electricfield, which is applied to quartz crystal via solution conduction and closely related to the conductivity of solution. The ESPS does not produce thermal effects for the biological materials and exhibits high sensitivity and accuracy (Shen et al., 1993). In this work, molecularly imprinted TiO2 (MIT) was selected as sensing film and the ESPS/flow injection analysis (FIA) system was employed to monitor the photocatalytic reaction. A novel nonenzymatic piezoelectric sensor has been developed for the detection of urea based on the modification of molecularly imprinted TiO2 thin film onto transparent TiO2 nanotube arrays. The constructed ESPS sensor with high sensitivity, selectivity and stability showed excellent performance for urea detection.
2. Experimental 2.1. Materials and apparatus High purity titanium (99.99%) was purchased from Sigma-Aldrich. Indium tin oxide (ITO) glass was obtained from Shenzhen Jianshengda Technology Co. Ltd. Glycerol, NH4F and Ti(O-nBu)4 were purchased from Sinopharm Co. Urea, glucose, uric acid, creatinine, thiourea, lactic acid, glycine and alanine were purchased from Shanghai Aladdin Chemical Co., and used without further purification. All other chemicals were of analytical grade and deionized (DI) water was used throughout the experiments. Piezoelectric measurements were performed in an ESPS/FIA sensing system which was described in our earlier report (Yang et al., 2007). The setup of the detection system is illustrated in Fig. 1. A MIT/TiO2 nanotube/ITO/glass electrode with an effective working area of 2.0 2.0 cm2 was used as the separated electrode. UV–vis transmission spectra were recorded by an UV/Vis spectrophotometer (UV-1601, Shimadzu, Japan). Scanning electron micrograph (SEM) was carried out on a Hitachi S-520 scanning electron microscope (Hitachi Ltd., Tokyo, Japan) working at 20 kV. X-ray diffraction (XRD) was performed on a Rigaku D/MAX-RC X-ray diffractometer using Cu Kα radiation. Raman spectra were recorded on a Raman spectrometer (InVia, Renishaw Co., UK). A 320 W low-pressure mercury lamp (Philips, 254 nm) was used as UV irradiation source.
2.2. Fabrication of the urea sensor Titanium thin films of 1 mm thickness were deposited onto ITOcoated glass substrates by RF sputtering. RF power of 200 W was applied during the deposition process. The sputtering was conducted under an Ar gas pressure of 2.7 10 3 Torr (base pressure 4.5 10 7 Torr) at 250 °C with a deposition rate of 8 nm/min. Prior to electrochemical treatment, the obtained Ti/ITO/glass electrodes were ultrasonically cleaned in ethanol and DI water for several minutes. Subsequently, Anodization was performed at 20 V in the high-purity glycerol containing 0.7 M NH4F and 2.5 vol% water using a platinum cathode. The sputtered titanium films on the glass substrates were transformed to nanotubular layers after three hours anodization at room temperature. Finally, the obtained nanotubular arrays were annealed at 500 °C for 2 h to transform completely into an anatase structure. Thus, TiO2 nanotube/ITO/ glass electrode was fabricated. The formation of MIT layer on TiO2 nanotube/ITO/glass electrode was performed by sol–gel hydrolysis of Ti(O-nBu)4 followed by calcination treatment. Specifically, Ti(O-nBu)4 (0.5 mmol) and urea (0.2 mmol) were dissolved in 8 mL of toluene-ethanol (v/v, 2:1) mixture, and then vigorously stirred at room temperature. Subsequently, the solution was diluted by 20 times with watersaturated toluene and continuously stirred for 5 h to get a sol–gel solution which was used as a dipping solution. The obtained TiO2 nanotube/ITO/glass electrode was immersed into the dipping solution for 6 min at room temperature, rinsed with toluene, hydrolyzed in water and dried in N2 atmosphere. The “dipping-rinsehydrolyzation-drying”cycle was repeated three times, and then the TiO2 nanotube/ITO/glass electrode whose surface was covered with a TiO2 layer containing the template molecule of urea was obtained. Finally, the TiO2 nanotube/ITO/glass electrode modified by TiO2 layer was dried at 80 °C for 2 h, and then calcined in air at 450 °C for 2 h to remove template molecules and crystalline the TiO2, resulting in the formation of MIT/TiO2 nanotube/ITO/glass electrode. Thus, the urea sensor was fabricated. For comparison, the non-imprinted TiO2 modified TiO2 nanotube/ITO/glass electrode was were prepared using the same procedure without template molecules, and was denoted as NIT/TiO2 nanotube/ITO/ glass electrode.
Fig. 1. Schematic illustration for setup and detection mechanism of the ESPS/FIA system.
Z.-p. Yang et al. / Biosensors and Bioelectronics 74 (2015) 85–90
2.3. Piezoelectric measurements The piezoelectric measurements of the obtained urea sensor were carried out in 0.01 mM phosphate buffer solution (PBS, pH 7.4) at 35 °C. After mounting MIT/TiO2 nanotube/ITO/glass electrode and quartz crystal disc into the flow cell, urea solution was injected into the detection system. The UV illumination was performed after incubation for 15 min, and the time-dependent change in the frequency was recorded by a frequency counter and stored in a microcomputer.
3. Results and discussion 3.1. Characterization of the piezoelectric urea sensor Fig. 2A shows the digital image of the obtained MIT/TiO2 nanotube/ITO/glass electrode. A high definition of the printed image underneath the layer was observed, indicating an excellent transparency of the electrode. The transparency was further investigated by UV-visible spectroscopy. Fig. 2B shows the UV–vis transmission spectra of the ITO/glass and MIT/TiO2 nanotube/ITO/
Fig. 2. (A) Digital image of the MIT/TiO2 nanotube coated on ITO/glass substrate. The text “HPU” is written underneath on paper to demonstrate the degree of transparency. (B) Transmittance spectra of (a) ITO/glass and (b) MIT/TiO2 nanotube/ ITO/glass.
87
glass. In the visible region, the average transmittance of MIT/TiO2 nanotube/ITO/glass reached about 67% to be compared with 79% of ITO/glass. A sharp decline in transmittance for wavelengths of less than 400 nm was clearly observed for the two substrates due to the strong absorption of glass and TiO2 in the ultraviolet region. Compared with the ITO/glass, a small decrease in transmittance of MIT/TiO2 nanotube/ITO/glass under the same wavelength can be attributed to the presence of MIT/TiO2 nanotube on ITO/glass. The surface morphology of the MIT/TiO2 nanotube/ITO/glass electrode was investigated by SEM. As presented in Fig. 3A, the electrode surface was actually rough, and was composed of highdensity, well-ordered and uniform nanopores. From the enlarged SEM images (the inset in Fig. 3A), the average pore size was about 40 nm. For comparison, the SEM observation of TiO2 nanotube/ ITO/glass electrode was also performed (not shown in the graph). The pore size was increased from 40 to 60 nm, indicating a successful modification of MIT on the surface of TiO2 nanotube arrays. The crystal structure and composition of electrode were identified by XRD. As seen in Fig. 3B, the diffraction peaks at 2θ ¼30.5°, 35.5°, 37.4°, 45.5°, 50.8° and 60.4° can be assigned to those XRD patterns of the remaining ITO coating (Berger et al., 2009). An amorphous shoulder can be related to the glass substrate. Several relatively narrow diffraction peaks at 2θ ¼ 25.4°, 47.8°, 53.7°, 54.8° and 62.4° can be perfectly indexed to the typical crystal planes of anatase, revealing the formation of anatase TiO2 on the ITO/glass substrate. Additionally, No diffraction peaks of titanium could be observed in the spectra of electrode, which suggested that the sputtered titanium thin film had been consumed and completely transformed into an anatase structure after the anodization and annealing treatment. Fig. 3C shows the Raman spectra taken for the MIT/TiO2 nanotube samples after annealing treatment. Ti–O–Ti signals in the range of 400–700 cm 1 are the characteristic of anatase and rutile structures (Bersani et al., 1998). Typically, several relatively strong vibration peaks were clearly observed at 145, 196, 398, 510 and 632 cm 1, which can be attributed to the Raman active modes of anatase phase (Hardwick et al., 2007). Thus, The Raman analysis further confirmed the existence of anatase TiO2 on the electrode surface. The formation of MIT film on the TiO2 nanotube/ITO/glass substrate was further evaluated using piezoelectric measurements. Fig. 3D shows the frequency responses of ESPS/FIA detection system with TiO2 nanotube/ITO/glass electrode, NIT/TiO2 nanotube/ ITO/glass electrode and MIT/TiO2 nanotube/ITO/glass electrode in 60 mM urea solution. The frequency change on MIT/TiO2 nanotube/ ITO/glass electrode reached 60 Hz, which was 4.6 times that on TiO2 nanotube/ITO/glass electrode and 3.0 times that on NIT/TiO2 nanotube/ITO/glass electrode. Significantly, the imprinted electrode exhibited much larger response compared to the TiO2 nanotube and nonimprinted electrodes. The larger frequency change can be attributed to the presence of MIT film on the electrode surface. The imprinted sites on the surface of MIT/TiO2 nanotube/ ITO/glass electrode can specifically rebind and accumulate urea molecules on the electrode surface. As a result, more urea molecules were photocatalytically degradated on the MIT/TiO2 nanotube/ITO/glass electrode. According to the report in literature (Kondo and Jardim, 1991; Zhang, 2008), the photodegradation products of urea on anatase TiO2 are mainly NO−3 and CO2− 3 . Thus, the initial uncharged urea molecules are transformed into charged ions, causing the increase of conductivity (namely the decrease in frequency response) in the flow cell. In addition, our study indicated that a steady-state response could be reached within 2 s. The response time is shorter than that reported in literature (Yang and Zhang, 2013; Srivastava et al., 2012).
88
Z.-p. Yang et al. / Biosensors and Bioelectronics 74 (2015) 85–90
Fig. 3. (A) Surface-view SEM image of the MIT/TiO2 nanotube coated on ITO/glass. The inset shows the enlarged view. (B) XRD spectra of MIT/TiO2 nanotube coated on ITO/ glass. (C) Raman spectra of MIT/TiO2 nanotube. (D) Frequency responses of ESPS/FIA detection system with TiO2 nanotube/ITO/glass electrode, NIT/TiO2 nanotube/ITO/glass electrode and MIT/TiO2 nanotube/ITO/glass electrode. 60 mM urea, pH 7.4, 35 °C. Average of three measurements (mean 7S.D.).
3.2. Biosensing performance The dynamic range of the urea sensor was studied by monitoring the urea solution in the concentration range from 0.01 to 150 mM. As seen in Fig. 4, the frequency response of the urea sensor increased with increasing urea concentration, and was linear in the concentration range of 0.04 to 120 μM (see the inset in Fig. 4). The linear regression equation was y (Hz) ¼ 12.05–0.76 [urea] (μM) with a correlation coefficient of 0.9979. The limit of detection (LOD) was found to be about 0.01 μM (S/N ¼3), which was lower than that reported by other researchers (Siqueira et al., 2014; Mondal and Sangaranarayanan, 2013; Syu and Chang, 2009). Obviously, the proposed electrode shows promise for application in the monitoring of urea with its low LOD and rapid response. Anti-interference properties are important considerations for sensors. In this work, glucose, uric acid, creatinine, thiourea, lactic acid, glycine and alanine were selected as interfering substances. After incubation for 15 min, the frequency response was recorded under UV illumination. The selectivity of the urea sensor was evaluated by calculating the frequency change ratio (ΔF/ΔF0). Herein, ΔF0 was the frequency change for 60 μM urea (S0) and was set as 100%, while ΔF referred to the frequency change of the two-component system containing 60 μM UA and 60 μM interfering substances (S1–S7). As seen in Fig. 5, these interferents
Fig. 4. Calibration graph obtained for the urea sensor in a urea concentration range from 0.01 to 150 mM. The inset shows the linear part of the calibration graph.
showed a slight effect on the frequency change of urea, and the deviations were all less than 4.5%, which adequately revealed that the urea sensor favored urea much more than other molecules and
Z.-p. Yang et al. / Biosensors and Bioelectronics 74 (2015) 85–90
89
practical applications, the proposed sensor was used to measure urea in urine samples. Prior to analysis, the urine samples were diluted 1000 times with PBS (pH 7.4) for spiked recovery tests, without other pretreatment. The recovery test was performed by spiking with different concentrations of urea to diluted urine samples. The parallel experiment was carried out three times. As showed in Table 1, with six different additions of urea to the urine samples containing 4.8 mM urea determined by the sensor, the obtained recoveries ranged from 95 to 102%. The good recovery rate indicates that the urea sensor is feasible for practical analysis.
4. Conclusions
Fig. 5. Frequency change ratio of urea sensor in 60 μM UA in the absence (ΔF0) and presence (ΔF) of 60 μM glucose (S1), uric acid (S2), creatinine (S3), thiourea (S4), lactic acid (S5), glycine (S6) and alanine (S7), respectively.
Table 1 Recovery test of the urea sensor. Urea concentration (μM) Added
0.2 1.0 5.0 20.0 40.0 80.0 a b
A novel nonenzymatic piezoelectric urea sensor was fabricated by a convenient attachment of MIT on highly ordered transparent TiO2 nanotube arrays. The constructed sensor exhibited high sensitivity, rapid response (o 2 s), wide linear range (0.04– 120 mM), low detection limit (0.01 mM), good selectivity and stability in the ESPS/FIA system. Detection of urea in urine sample was also performed with satisfactory results. The proposed sensor shows a promising application in monitoring of biomolecules and offers a new avenue to broaden the applications of transparent TiO2 nanotube arrays in sensors.
Recovery (%)
Acknowledgments
Found Meana
R.S.D.b (%)
0.19 1.0 5.1 19.9 38.8 80.4
3.1 2.6 4.0 2.9 5.7 2.8
95.0 100.0 102.0 99.5 97.0 100.5
Average of three determinations. Relative standard deviation.
therefore presented outstanding selectivity towards urea. The superior selectivity can be attributed to the recognition function of surface molecularly imprinted modification. The imprinted cavities formed in MIT film can distinguish urea from other species through molecular size and functional group distribution, and rebind urea selectively. Thus, urea molecules were specifically accumulated on the sensor surface, whereas other coexistent molecules not complementary to the cavities mainly remained in the bulk solution. The reproducibility of the same urea sensor was evaluated by measuring the response signal twenty times in a continuous manner. After each injection of 60 mM urea solution and the attainment of equilibrium, the urea sensor was recovered by washing with DI water under UV illumination until the frequency response reached a steady value. The sensor can be stored dry at room temperature and reused for next experiments. The obtained results indicated that the relative standard deviation of frequency response was only 1.31%, indicating a good reproducibility of the sensor. The long-term stability revealed that after storage of five months at room temperature, the frequency change decreased by only 2.07% compared to the initial frequency response. The obtained results indicated the high stability of the MIT/TiO2 nanotube/ITO/glass electrode for urea sensing. The rigid structure of MIT, the strong binding of MIT on TiO2 nanotube/ITO/glass may be responsible for the excellent reproducibility and stability. In order to evaluate the reliability of the developed sensor in
This work was supported by the National Natural Science Foundation of China (No. 51202059), the Research Foundation for Youth Scholars of Higher Education of Henan Province (No. 2013GGJS-047), the Fundamental Research Funds for the Universities of Henan Province (NSFRF140201, NSFRF140601), the Key Foundation of He'nan Educational Committee (No. 14A430014) and Foundation for Distinguished Young Scientists of Henan Polytechnic University (J2014-05).
References Ahmad, R., Tripathy, N., Hahn, Y.B., 2014. Sens. Actuators B 194, 290–295. Berger, S., Ghicov, A., Nah, Y.C., Schmuki, P., 2009. Langmuir 25, 4841–4844. Bersani, D., Antonioli, G., Lottici, P.P., Lopez, T.J., 1998. Non-Cryst. Solids 232, 175–181. Ciana, L.D., Caputo, G., 1996. Clin. Chem. 42, 1079–1085. de Jesus, D.P., Neves, C.A., do Lago, C.L., 2001. J. Brazil. Chem. Soc. 12, 123–128. Dhawan, G., Sumana, G., Malhotra, B.D., 2009. Biochem. Eng. J. 44, 42–52. Francis, P.S., Lewis, S.W., Lim, K.F., 2002. Trends Anal. Chem. 21, 389–400. Gutierrez, M., Alegret, S., Valle, M., 2007. Biosens. Bioelectron. 22, 2171–2178. Hamilton, A., Breslin, C.B., 2014. Electrochim. Acta 145, 19–26. Hardwick, L.J., Holzapfel, M., Novak, P., Dupont, L., Baudrin, E., 2007. Electrochim. Acta 52, 5357–5367. Ji, Y.J., Lin, K.C., Zheng, H.G., Liu, C.C., Dudik, L., Zhu, J.J., Burda, C., 2010. ACS Appl. Mater. Inter. 2, 3075–3082. Kessler, A., Siekmann, L., 1999. Clin. Chem. 45, 1523–1529. Kondo, M.M., Jardim, W.F., 1991. Water Res. 25, 823–827. Lamas-Ardisana, P.J., Costa-Garcia, A., 2006. Sens. Actuators B 115, 567–574. Li, S.Q., Zhang, G.M., Guo, D.Z., Yu, L.G., Zhang, W.J., 2009. Phys. Chem. C 113, 12759–12765. Lin, Y.H., Wang, S.H., Wu, M.H., Pan, T.M., Lai, C.S., Luo, J.D., Chiou, C.C., 2013. Biosens. Bioelectron. 43, 328–335. Liu, Y.T., Liu, R.H., Liu, C.B., Luo, S.L., Yang, L.X., Sui, F., Teng, Y.R., Yang, R.B., Cai, Q.Y., 2010. J. Hazard. Mater. 182, 912–918. Mondal, S., Sangaranarayanan, M.V., 2013. Sens. Actuators B 177, 478–486. Mor, G.K., Varghese, O.K., Paulose, M., Grimes, C.A., 2005. Adv. Funct. Mater. 15, 1291–1296. Roch-Ramel, F., 1967. Anal. Biochem. 21, 372–381. Shen, D.Z., Zhu, W.H., Nie, L.H., Yao, S.Z., 1993. Anal. Chim. Acta 276, 87–97. Siqueira, J.R., Molinnus, D., Beging, S., Schoning, M.J., 2014. Anal. Chem. 86, 5370–5375. Soldatkin, O.O., Kucherenko, I.S., Marchenko, S.V., Ozansoy Kasap, B., Akata, B., Soldatkin, A.P., Dzyadevych, S.V., 2014. Mater. Sci. Eng. C 42, 155–160. Srivastava, R.K., Srivastava, S., Narayanan, T.N., Mahlotra, B.D., Vajtai, R., Ajayan, P.
90
Z.-p. Yang et al. / Biosensors and Bioelectronics 74 (2015) 85–90
M., Srivastava, A., 2012. ACS Nano 6, 168–175. Syu, M.J., Chang, Y.S., 2009. Biosens. Bioelectron. 24, 2671–2677. TranT., T., Li, J.Z., Feng, H., Cai, J., Yuan, L.J., Wang, N.Y., Cai, Q.Y., 2014. Sens. Actuators B 190, 745–751. Verma, R., Gupta, B.D., 2014. Analyst 139, 1449–1455. Yang, Z.P., Shang, X.G., Zhang, C.J., Zhu, J.P., 2014. Sens. Actuators B 201, 167–172. Yang, Z.P., Si, S.H., Dai, H.J., Zhang, C.J., 2007. Biosens. Bioelectron. 22, 3283–3287. Yang, Z.P., Yan, J.L., Zhang, C.J., Luo, S.Q., 2011. Colloid. Surf. B 87, 187–191.
Yang, Z.P., Zhang, C.J., 2013. Sens. Actuators B 188, 313–317. Zhang, C.J., Si, S.H., Yang, Z.P., 2015. Biosens. Bioelectron. 65, 115–120. Zhang, J., Liu, G.X., Zhang, Y., Gao, Q., Wang, D.P., Liu, H., 2014. J. Agric. Food Chem. 62, 2797–2802. Zhang, S.Y., 2008. Preliminary Study on Urea Photolysis (Master's Degree Dissertation). Huazhong Agricultural University, Wuhan.
