Molecularly Imprinted Electrochemical Luminescence Sensor Based ...

83 downloads 2096 Views 871KB Size Report
Oct 26, 2012 - range, good accuracy, and fast response. Beer samples were .... Sinopharm Chemical Reagent Company, Ltd.; A 0.01 mol/L luminol stock ...
Article pubs.acs.org/ac

Molecularly Imprinted Electrochemical Luminescence Sensor Based On Signal Amplification for Selective Determination of Trace Gibberellin A3 Jianping Li,* Shuhuai Li, Xiaoping Wei, Huilin Tao, and Hongcheng Pan College of Chemistry and Bioengineering, Guilin University of Technology, Guangxi 541004, China S Supporting Information *

ABSTRACT: A new molecularly imprinted electrochemical luminescence (MIP-ECL) sensor was developed for Gibberellin A3 (GA3) determination. This sensor is based on competitive binding between the GA3 and the Rhodamine B (RhB)-labeled GA3 (RhB-GA3) to the MIP film. After the competitive binding, the residual RhB-GA3 on the MIP was electro-oxidized to produce RhB oxide, which could greatly amplify the weak electrochemiluminescence (ECL) signal of luminol. The ECL intensity decreased when the RhB-GA3 was replaced by GA3 molecules in the samples. Accordingly, GA3 was determined in the concentration range from 1 × 10−11 to 3 × 10−9 mol/L with a detection limit of 3.45 × 10−12 mol/L. The sensor shows high sensitivity and selectivity, wide response range, good accuracy, and fast response. Beer samples were assayed by using the sensors, and the recoveries ranging from 96.0% to 103.2% were obtained.

G

In this paper, a MIP-ECL sensor was developed, and the GA3 was selected as the template. Similar to the approach described in our previous work,21,22 the MIP-ECL sensor fabrication involves the steps of elution, blocking, incubation, and competition (Scheme 1). Competitive binding between GA3 in the sample and RhB-GA3 in the MIP film were observed. Luminol can be electro-oxidized on a gold electrode to produce a weak ECL signal.23 It was found that RhB could be electrochemically oxidized to produce the oxidative intermediate product (RhBox), and RhBox might react with dissolved oxygen in the solution to yield a superoxide anion radical (O2−).24 In this study, the GA3-labeled RhB-derived superoxide anion radical reacted with a free radical of luminol to generate the excited 3-aminophtalate (AP2−)*. Then, an amplifying chemiluminescence signal emits. To the best of our knowledge, this is a new mode of ECL produced by the coupling of the electrochemical reactions and chemical reaction with the subsequent CL reaction. With the decrease of the RhB-GA3 after competition, the intensity of the ECL signal of the luminol had an obvious decrease, and the GA3 concentrations can be determined. So, a highly sensitive MIPECL sensor has been developed for the determination of GA3, and a new ECL type was described. The proposed sensor was successfully applied to the determination of GA3 in beer. The scheme of analysis of an MIP-ECL sensor based on the RhB amplification is shown in Scheme 1.

ibberellins, a phytohormone involved in many processes of plant growth, have been widely used in agriculture. The biological activity of various gibberellins varies with the individual, and the activity of Gibberellin A3 (GA3), which is studied the most, is the highest. As a stimulator of leaf and bud growth, GA3 was used to increase the bud ratio of malt which is the main raw material for brewing beer.1 Recent studies show that the accumulation of GAs in human body can cause chronic toxicity, cancer, and other diseases. Consequently, strict limits have been set on GA3 residues in beer2 and other agricultural products.3 This remains a great challenge to detect trace amount of GA3 residues. So far, various methods such as highperformance liquid chromatography-tandem mass spectrometry (HPLC-MS)4,5 and immunoassays6 have been developed for GA3 detection. But HPLC-MS has some deficiencies such as costly equipment and tedious and time-consuming procedures. The immunoassay exhibits relatively low detection limits. Thus, a simple, highly sensitive, and selective method for trace GA3 detection is highly desirable. Molecular imprinting polymer (MIP)-based sensors with predetermination, specific recognition, and practicability are widely used in bioanalysis,7−12 drug analysis,13 and pesticide or veterinary drug residue analysis.14,15 Recent works have demonstrated that molecular imprinted electrochemical sensors (MIECS) can combine the advantageous properties of MIP and electrochemical sensors to improve sensor performance.16−20 However, the MIECS typically suffer from a low current efficiency. To further improve the sensitivity, electrochemiluminescence (ECL) has been applied to fabricate MIP-ECL sensors. © 2012 American Chemical Society

Received: August 19, 2012 Accepted: October 26, 2012 Published: October 26, 2012 9951

dx.doi.org/10.1021/ac302401s | Anal. Chem. 2012, 84, 9951−9955

Analytical Chemistry

Article

Scheme 1. The Procedure to Construct a MIP-ECL Sensor and Determination of GA3

After electropolymerization, the MIP and nMIP sensors were washed with carbinol/acetic acid (8:1 in volume) for 10 min to remove the imprinting molecules or the adsorbates within or on the surface of the imprinted membrane, and then the MIP sensors with stereo cavities in imprinted membranes were obtained. An nMIP sensor was prepared in accordance with the same method only without GA3. After each usage, the MIP sensors were washed using carbinol/acetic acid (8:1 in volume) for 10 min to remove the imprinting molecules. Masking, Incubation, and Competition. The MIP sensor was immersed in 10 mL of 4 × 10−3 mol/L GA3 solution for 12 min to mask all the vacant binding cavities in MIP. Then, it was incubated in 2 mL of 5 g/mL RhB-GA3 solution for 15 min to allow RhB-GA3 to replace the GA3. Finally, the sensor was competed in 10 mL of 1 × 10−11 to 3 × 10−9 mol/L GA3 for 12 min. All measurements were carried out at 25 °C. The Electrochemical and ECL Measuring Methods. The electrochemical experiments were performed in 3 × 10−4 mol/L K3[Fe(CN)6]/K4[Fe(CN)6] solution containing 0.5 mol/L KCl. Cyclic voltammograms (CV) were performed over a potential range from −0.2 to +0.6 V, at a scan rate of 50 mV/ s, while differential pulse voltammetric (DPV) measurements were performed over a potential range from −0.2 to +0.8 V, at a scan rate of 50 mV/s and pulse amplitude of 50 mV. Alternating current impedance (AC) was performed at a potential of 0.19 V, over the frequency range from 100 mHz to 100 kHz, using an alternating voltage of 5 mV. The ECL test was conducted in a 10 mL 0.05 mol/L TrisHCl buffer (pH =7.8) containing 1.2 × 10−3 mol/L luminol. The ECL measurement was performed by CV from −0.2 to +0.6 V with the scan rate of 100 mV/s. The voltage of the photomultiplier tube (PMT) was set at 900 V. The ECL signal−time curve under continuous potential scanning was performed for 5 cycles with the magnification of 3. ECL signals related to the GA3 concentrations were recorded.

The application of the method for the determination of GA3 in the beer sample was performed, and the results indicated that the MIP-ECL sensor met the needs of practical analysis.



EXPERIMENTAL SECTION Apparatus. The electrochemiluminescence measurements were carried out on a model MPI-E ECL analyzer system (Xi’an Remex Instrument Co. Ltd.) using a three-electrode system. Electroanalytical measurements such as differential pulse voltammetry (DPV), cyclic voltammetry (CV), and alternating current (AC) were performed on an Autolab Electrochemical Workstation with a standard three-electrode cell (Metrohm China Co., Ltd.). The three-electrode system consisted of a Ag/ AgCl electrode containing a saturated KCl solution as the reference electrode, a platinum wire electrode as the auxiliary electrode, and an MIP-modified gold electrode (d = 2 mm) as the working electrode. A pHS-2C model pH meter (Shanghai Leici Instruments) and a DK-8B Electrothermal Constant Temperature Incubator (Shanghai Jinghong Instruments) were also used. All measurements were carried out at 25 °C. Reagents and Chemicals. Gibberellin A1, A2, A3, A4, and A7 were obtained from Aladdin reagent Inc.; RhB-GA3 was obtained from Zhengzhou Biocell Biotechnology Co., Ltd.; Luminol (>98%, Fluka) was obtained from Yingrun Biotechnologies Inc.; o-phenylenediamine was obtained from Sinopharm Chemical Reagent Company, Ltd.; A 0.01 mol/L luminol stock solution was prepared by dissolving 0.0886 g luminol in 0.1 mol/L sodium borate buffer; 0.05 mol/L TrisHCl buffer solution (pH = 7.8) was prepared by 0.05 mol/L tris(hydroxymethyl) aminomethane and 0.1 mol/L HCl; 3 × 10−4 mol/L K3[Fe(CN)6]/K4[Fe(CN)6] solution containing 0.5 mol/L KCl. All the reagents used in the experiment were analytical grade reagent, and all solutions were prepared with ultra pure water from a high purity water system (Youpu Super Water Company, Ltd., Chengdu, China). Preparation of MIP and Nonmolecular Imprinted Polymer (nMIP)-Modified Electrodes. The gold electrodes were polished using a microcloth (chamois leather) with 1.0, 0.3, and 0.05 μm of an aqueous slurry of alumina before alternately being washed with water, alcohol, and HNO3 (50% in volume). Then, MIP and nMIP were prepared on a gold electrode by the electro-polymerization method. The ophenylenediamine was dissolved in 0.05 mol/L NaAc-HAc buffer (pH = 5.2), and then 4 × 10−3 mol/L GA3 solution was added in and mixed. Thirty cycles of CVs in the potential range from 0 to +0.8 V at 50 mV/s in the above solution were performed on a gold electrode.



RESULTS AND DISCUSSION

Electropolymerization of o-Phenylenediamine and GA3. The electropolymerization of o-phenylenediamine on the gold electrode was an irreversible process, and there was a distinct and irreversible oxidation peak at 0.35 V. As the number of cycles increased, the oxidation peak current of ophenylenediamine continuously decreased due to a compact nonconducting polymer film formed and covered onto the surface of the electrode. When the number of cycles increased to 30, the current density of the oxidation peak became smaller; 9952

dx.doi.org/10.1021/ac302401s | Anal. Chem. 2012, 84, 9951−9955

Analytical Chemistry

Article

this indicated that the MIP film had been formed onto the surface of the electrode. The ECL of MIP Sensor. The amplification effects of RhB on the weak ECL signal of luminol were studied and conformed. As shown in Figure 1, the bare gold electrode

Figure 2. CVs of the MIP sensor and nonmolecular imprinted polymer (nMIP) in different condition. (a) Bare gold electrode, (b) MIP-electrode, (c) MIP-electrode after template removal, (d) MIPelectrode after rebinding, (e) nMIP-electrode, and (f) nMIP after washing. The scan rate was 50 mV/s.

template molecules to enter the cavities. Therefore, it is evident that the molecularly imprinted membrane yields a response to GA3. The insert map in Figure 2 was the CVs of nMIP. When the nMIP film of poly-o-phenylenediamine was formed on the surface of the electrode, the current of the probe decreased, which indicated that the poly-o-phenylenediamine film was nonconductive for the electrochemical reaction (curve e). After the elution step, the peak current in curve f was almost unchanged compared with that in curve e. It demonstrated that there were no template molecules doped in the MIP film and no cavities formed as the channels for electron transportation after elution. Alternating Current (AC) Characterized by MIPModified Film and nMIP-Modified Film. Alternating current impedances were also used to show the impedance change while preparing the MIP electrode (Figure 3). An MIP

Figure 1. The effect of Rhodamine B amplification for ECL of the MIP senor. (a) The ECL of the bare electrode in luminol. (b) The ECL of MIP after removing the template molecule in luminol. (c) The ECL of MIP after masking by GA3 in luminol. (d) The ECL of MIP after removing the template molecule in RhB. (e) The ECL of MIP after GA3 competed with RhB-GA3 without luminol. (f) The ECL of MIP after GA3 competed with RhB-GA3 in luminol.

has a weak ECL signal in the presence of luminol (curve a). The ECL intensity of MIP after removing the template molecule (curve b) is smaller since the MIP was formed on the electrode surface. After blocking by GA3, the ECL of MIP in luminol reduces (curve c). It indicates that GA3 cannot increase the ECL intensity of luminol. The RhB has no ECL signal in the base solution without luminol (curve e), as the same MIP after removing the template molecule has no ECL signal in the base solution with RhB (curve d). Then, the ECL intensity of luminol increases greatly when RhB meets luminol (curve f). It indicates that the ECL intensity of luminol is amplified by RhB. Characterization of MIP-Modified Film. Since GAs could not electrochemically react on the Au electrode and no redox peaks were observed in the potential range from −0.2 to + 0.6 V, 0.01 mol/L K3[Fe(CN)6] solution (containing 0.5 mol/L KCl) was used as the redox probe between imprinted electrodes and substrate solutions to characterize the imprinting processes of MIP film by cyclic voltammetry (CV) and alternating current (AC) with Autolab Electrochemical Workstation (Metrohm Company, Ltd.). Cyclic Voltammograms (CV) Characterized for MIPModified Film and nMIP-Modified Film. Since cavities in MIP can be used as the channels for electron transportation, the blocking of channels causes change in the redox peaks, correspondingly. As shown in Figure 2, from curve a to b, a nonconductive MIP was formed on the Au-electrode surface: there was difficulty getting the redox probe to the Au electrode surface and the peak current of the probe decreased. After the removal of the template molecules, some cavities in MIP appeared and the peak current of the probe increased (curve c). When the MIP electrode continuously rebounded with GA3, the cavities were blocked again and the peak current decreased (curve d). But it was much higher than that of curve b, because some cavities might be deformed, making it difficult for the

Figure 3. AC impedances characterization of MIP films. (a) Bare gold electrode, (b) MIP-electrode, (c) MIP-electrode after template removal, (d) MIP-electrode after rebinding. Perform potential: 0.19 V; frequency range: from 100 mHz to 100 kHz; alternating voltage: 5 mV.

was formed and covered on the surface of the Au electrode, which made the electron transfer difficult between the bulk solution and electrode surface, so the resistance increased (curve a to b). When the template molecule was removed, cavities were exposed as the channels for electron transfer, and the resistance decreased (curve b to c). While the template 9953

dx.doi.org/10.1021/ac302401s | Anal. Chem. 2012, 84, 9951−9955

Analytical Chemistry

Article

electrochemical luminescence intensity of the luminol system obviously reduced. The decreasing ECL intensities were linear to the concentrations (C) of GA3 in the range from 1 × 10−11 to 3 × 10−9 mol/L (Figure 5). The linear regression equation

molecules had been recombined on the MIP electrode again, the channels were blocked, and the resistance increased (curve c to d). Optimization of Experimental Conditions of Masking, Incubation, and Competition. The masking step was carried out in 10 mL of 4.5 × 10−5 mol/L GA3 solution. DPV was performed every 3 min in the experiment. The currents decreased gradually and then remained constant after 15 min. Thus, the optimal masking time was selected as 15 min. The time for RhB-GA3 reaching the imprinted membrane and replacing the GA3 to combine with cavities in the MIP was defined as the incubation time. The electrode was incubated in 5 μg/mL RhB-GA3 solution. Different ECL intensity was recorded every 3 min. The increase of the concentration of RhB-GA3 resulted in the rise of ECL intensity. The intensity reached the maximum when the incubation time reached 18 min (Figure 4, curve a). This demonstrates that the incubation reaction was complete within 18 min. As a result, 18 min was selected as the incubation time in all of the following assays.

Figure 5. ECL of the MIP-sensor after incubation in different concentrations of GA3 (a→m): (0, 1, 3, 6, 12, 25, 50, 80, 120, 160, 200, 250, and 300) × 10−11 mol/L GA3, respectively. Scan rate: 100 mV/s; scan range: from −0.2 to +0.6 V; and PMT voltage: 900 V.

was ΔI = 11.51 C (10−11 mol/L) + 134.44, with a coefficient of correlation, r = 0.999, and the detection limit was 3.45 × 10−12 mol/L (DL = 3δb/K). Compared with other methods in literatures,4−6 the MIP-ECL sensor with an enzymatic effect is more sensitive with a lower detection limit. Selectivity of the MIP Sensor. In order to evaluate the recognition sites in the MIP which have a specific selection to the template molecules, measurements of the ECL intensity were performed after the competition procedure of 1.0 × 10−9 mol/L GA3 in the coexistance of other gibberellins with similar structure and homothetic function, such as GA1, GA2, GA4, and GA7. The ECL intensity of the MIP sensors had almost no change after the samples were mingled 50 times with GA1, GA2, GA4, and GA7, respectively. The results (Figure 6) verified the selectivity of the MIP sensor for GA3. On the contrary, when these analogues were analyzed using the bare electrode, the ECL intensity responses were similar to those obtained from GA3 detection, suggesting that the developed MIP-ECL sensor possesses satisfactory analytical selectivity.

Figure 4. Effects of (a) incubation reaction time and (b) competition reaction time on response signals. (a) MIP was incubated in 5 μg/mL RhB-GA3 after blocking. (b) MIP competed in 2 × 10−9 mol/L GA3 after incubation.

The process in which GA3 in samples reached the imprinted film and replaced the RhB-GA3 combined with cavities in the MIP was defined as competition. It was operated in a 2 × 10−9 mol/L GA3 solution. The ECL response was recorded every 3 min. The ECL intensity decreased and remained constant after 15 min (Figure 4, curve b). This suggested that the interaction reached equilibrium at 15 min. As a result, 15 min was selected as the competition time. Effects of Buffer Solution and Luminol Concentration on ECL Intensity. Several buffer solutions including 0.05 mol/ L borax solution, 0.05 mol/L Tris-HCl, and 0.05 mol/L PBS were tested. The result demonstrated that the maximal ECL intensity could be obtained in Tris-HCl. Then, the effect of pH values of Tris-HCl on ECL intensity was examined in the range of 7.0−9.0. The results showed that the ECL signal increased with the pH value increase and reached the maximum at pH 7.8, and then it decreased with the pH up to 9.0. The effect of concentrations of luminol on ECL intensity was also evaluated. The results showed that the maximal ECL intensity could be obtained in 1.2 × 10−3 mol/L luminol. ECL Response to GA3. Under the optimized conditions, the incubated sensor competed in the GA3 solution of different concentrations, and the ECL intensity was recorded. With the decrease of the RhB-labeled GA3 in the MIP film, the

Figure 6. ECLs of the MIP-sensor after rebinding in different GAs. (a) MIP-sensor after incubater. (b) MIP-sensor after rebinding 1.2 × 10−9 mol/L GA3. (c−f) MIP-sensor after rebinding 1.2 × 10−9 mol/L GA3 mingled with 6 × 10−8 mol/L GA1, GA2, GA4, and GA7. Scan rate: 100 mV/s; scan range: from −0.2 to +0.6 V; PMT voltage: 900 V. 9954

dx.doi.org/10.1021/ac302401s | Anal. Chem. 2012, 84, 9951−9955

Analytical Chemistry

Article

Table 1. The Results of Sample Assay and Recoveries sample

found (10−11 mol/L)

RSD (%) (n = 5)

found (HPLC method) (10−11 mol/L)

added (10−11 mol/L)

total found (10−11 mol/L)

RSD (%) (n = 5)

recovery (%)

beer 1 beer 2 beer 3

78.33 46.41 51.34

2.29 2.14 1.87

78.11 47.12 50.71

50 75 100

123.16 118.35 151.11

1.97 1.66 1.57

96.0 97.5 103.2



ACKNOWLEDGMENTS The authors gratefully acknowledge the financial support received from the National Nature Science Foundation of China (21165007), the Natural Science Foundation of the Guangxi Province (2012jjAA20076), and the Innovation Project of Guangxi Graduate Education (2011105960703M19).

The influence of other interferences on the determination of 1.0 × 10−9 mol/L GA3 was also examined. The results showed that Na+, Ca2+, Mg2+, SO42−, Cl−, and CO32− in 1000-fold; glucose, sucrose, glutamic acid, Fe2+, Al3+, and Zn2+ in 500-fold; Cu2+ and Hg2+ in 300-fold of GA3 concentration did not affect the ECL signals produced by GA3. Reproducibility and Stability Experiment. The reproducibility of the sensor was examined by the measurement of the ECL intensities when five different sensors fabricated under the same condition mentioned above competed in an 8 × 10−10 mol/L GA3 solution. The ECL intensities of the five sensors obtained were as follows: 2980.4, 2988.4, 2955.4, 2978.5, and 2988.1. It resulted in a relative standard deviation (RSD) of 0.23%, which indicated good sensor-to-sensor reproducibility. Moreover, the reproducibility of detection was also evaluated by measurements of 2 × 10−9 mol/L GA3 for five successive times by a sensor; the RSD obtained was 0.98%. To ensure the stability, the sensor was dipped in doubly distilled water at 4 °C when not in use. 2 × 10−9 mol/L GA3 was tested periodically after competition. No apparent decrease was observed in 7 days. After 15 days, the ECL intensity decreased about 5.7%, and after a month, it decreased about 16.9% compared with the initial response. The decrease of the signal might cause the degradation of RhB. The UV spectra and HPLC-MS spectra showed that RhB evidently degraded after 4 weeks (see the Supporting Information). Samples Ddetection. The marketable beer samples were analyzed and the standard addition method was performed to assess the accuracy. The result showed that the recoveries of this sensor ranged from 96.0% to 103.2%, with the RSDs less than 2.3%. The analytical results of this method are consistent with that of HPLC, which can be an indicator of good accuracy and practicability (Table 1).





CONCLUSIONS In this paper, a novel MIP sensor based on ECL and amplification effects of RhB on the ECL signal of luminol for the detection of GA3 was proposed. It provided a new analytical method with excellent performance in high sensitivity, selectivity, and low cost, which could be expected in the trace analysis of environment, food, and drugs.



ASSOCIATED CONTENT

S Supporting Information *

This material is available free of charge via the Internet at http://pubs.acs.org.



REFERENCES

(1) Mander, L. N. Chem. Rev. (Washington, DC, U.S.) 1992, 92, 573− 612. (2) Code of Federal Regulations, Title 40, Part 180.1098; U.S. Government Printing Office: Washington, DC, 2012. (3) Food and Drug Administration (FDA) 21 CFR, Part 172.725; U.S. Government Printing Office: Washington, DC, 2000. (4) Zhang, Z. M.; Tan, W.; Hu, Y. L.; Li, G. K.; Zan, S. Analyst 2012, 137, 968−977. (5) Han, Z.; Liu, G.; Rao, Q. X.; Bai, B.; Zhao, Z. H.; Liu, H.; Wu, A. B. J. Chromatogr., B 2012, 881−882, 83−89. (6) Li, J.; Xiao, L. T.; Zeng, G. M.; Huang, G. H.; Shen, G. L.; Yu, R. Q. J. Agric. Food Chem. 2005, 53, 1348−1353. (7) Greene, N. T.; Morgan, S. L.; Shimizu, K. D. Chem. Commun. 2004, 10, 1172−1173. (8) Haupt, K.; Mosbach, K. Chem. Rev. (Washington, DC, U.S.) 2000, 100, 2495−2504. (9) Sergeyeva, T. A.; Piletsky, S. A.; Panasyuk, T. L.; El’skaya, A. V.; Brovko, A. A.; Slinchenko, E. A.; Sergeeva, L. M. Analyst 1999, 124, 331−334. (10) Ye, L.; Mosbach, K. Chem. Mater. 2008, 20, 859−868. (11) Whitcombe, M. J.; Chianella, I.; Larcombe, L.; Piletsky, S. A.; Noble, J.; Porter, R.; Horgan, A. Chem. Soc. Rev. 2011, 40, 1547−1571. (12) Gräfea, A.; Haupt, K.; Mohr, G. Anal. Chim. Acta 2006, 565, 42−47. (13) Tokareva, I.; Tokarev, I.; Minko, S.; Hutter, E.; Fendler, J. H. Chem. Commun. 2006, 3343−3345. (14) Li, C.; Wang, C.; Wang, C.; Hu, S. Sens. Actuators, B 2006, 117, 166−171. (15) Mojica, E. R. E.; Autschbach, J.; Bright, F. V.; Aga, D. S. Analyst 2011, 136, 749−755. (16) Alizadeh, T.; Akhoundian, M. Electrochim. Acta 2010, 55, 3479− 3485. (17) Jin, G.; Li, W.; Yu, S.; Peng, Y.; Kong, J. Analyst 2008, 133, 1367−1372. (18) Suedee, R.; Intakong, W.; Dickert, F. L. Anal. Chim. Acta 2006, 569, 66−75. (19) Wang, Y. T.; Zhou, Y. X.; Sokolov, J.; Rigas, B.; Levon, K.; Rafailovich, M. H. Biosens. Bioelectron. 2008, 24, 162−166. (20) Kan, X. W.; Zhou, H.; Li, C.; Zhu, A. H.; Xing, Z. L.; Zhao, Z. Electrochim. Acta 2012, 63, 69−75. (21) Li, J. P.; Jiang, F. Y.; Wei, X. P. Anal. Chem. 2010, 82, 6074− 6078. (22) Li, J. P.; Li, Y. P.; Zhang, Y.; Wei, G. Anal. Chem. 2012, 84, 1888−1893. (23) Kuwana, T.; Epstein, B.; Seo, E. T. J. Phys. Chem. 1963, 67, 2243−2244. (24) Austin, J. M.; Harrison, I. R.; Quickenden, T. I. J. Phys. Chem. 1986, 90, 1839−1843.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest. 9955

dx.doi.org/10.1021/ac302401s | Anal. Chem. 2012, 84, 9951−9955