Anal. Chem. 2003, 75, 7020-7025

Photoelectrocatalytic Oxidation of o-Phenols on Copper-Plated Screen-Printed Electrodes Jyh-Myng Zen,* Hsieh-Hsun Chung, Hsueh-Hui Yang, Mei-Hsin Chiu, and Jun-Wei Sue

Department of Chemistry, National Chung Hsing University, Taichung 402, Taiwan

A novel and sensitive detection method based on photoelectrocatalytic oxidation of o-diphenols was demonstrated on a copper-plated screen-printed carbon electrode (designated CuSPE) in pH 8 phosphate buffer solution. The o-diphenols can be detected amperometrically through electrochemical oxidation at a low applied potential of -0.1 V versus Ag/AgCl, where the CuSPE is much less subject to interfering reactions. The mechanism that induces good selectivity of the CuSPE is explained in terms of the formation of a cyclic five-member complex intermediate (CuII-o-quinolate). A prototype homemade flow through cell design is described for incorporating the photoelectrode and light source. Electrode irradiation results in a large increase in anodic current. The oxidative photocurrents produced by irradiation increase with light intensity presumably because of the formation of semiconductor Cu2O. The principle used in this study has an opportunity to extend into various research applications. In a previous article, we described a simple and disposable copper-plated screen-printed carbon electrode (designated CuSPE) for the selective determination of o-diphenols (e.g., catechol, dopamine, and pyrogallol) in the presence of other isomers and phenols.1 The CuSPE shows an unusual catalytic response at -0.05 V versus Ag/AgCl to the oxidation of o-diphenolic compounds. The mechanism of good selectivity is explained in terms of the formation of cyclic five-member complex intermediate (CuII-o-quinolate). A highly stable system that operates at a low detection potential under hydrodynamic conditions is attractive in analytical chemistry. Most important, the common drawback of electrode fouling through polymerization is completely overcome in this system. Recently our group further developed an efficient photocatalytic amperometric sensor for the determination of dissolved oxygen using the same CuSPE.2 The photoelectrochemical activity toward dissolved oxygen is related to the formation of Cu2O. It is well known that Cu2O is a p-type semiconductor and has good photochemical behavior.3-6 Because * To whom correspondence should be addressed. Phone: (+886) 4-22854007. Fax: (+886) 4-2286-2547. E-mail: [email protected]. (1) Zen, J.-M.; Chung, H.-H.; Senthil Kumar, A. Anal. Chem. 2002, 74, 12021206. (2) Zen, J.-M.; Song, Y.-S.; Chung, H.-H.; Hsu, C.-T.; Senthil Kumar, A. Anal. Chem. 2002, 74, 6126-6130. (3) de Jongh, P. E.; Vamaekelbergh, D.; Kelly, J. J. Chem. Commun. 1999, 1069-1067. (4) Tennakone, K.; Kumarsinghe, A. R.; Sirimanne, P. M. J. Photochem. Photobiol. A 1995, 88, 39-41.

7020 Analytical Chemistry, Vol. 75, No. 24, December 15, 2003

the preparation of the CuSPE is simple and inexpensive, it offers an easy route for the quantitative detection of dissolved oxygen. By combining these two interesting concepts, we report here a new approach to the sensitive detection of o-diphenols using the photoelectrochemical method. To our knowledge, there has been no report for the photochemical quantitative assays of o-diphenols using a semiconducting system so far. The application of semiconductor electrodes to electrochemical sensor systems offers advantages such as selectivity and high signal-to-noise ratio stemming from the peculiarities of electron exchange between semiconductor bands and electroactive species from solution. Because of the role of catecholamines of o-diphenol derivatives in neurochemistry, their qualitative and quantitative assays are of clinical and biochemical importance.7-11 Since these chemicals often exist as a mixture of polyphenols in the working matrix,12-15 the development of sensitive and selective assays for phenolic groups, especially o-diphenols, deserves further studies. In this report, a prototype homemade flow-through cell design is described for the purpose of incorporating the photoelectrode and light source. Meanwhile, catechol was chosen as a model compound to demonstrate the analytical utility of this method. The results of carefully designed experiments reveal a possible mechanism for the good selectivity. This approach not only keeps the advantages of high selectivity, low interference, and freedom from passivation in our previous report1 but also provides better sensitivity. EXPERIMENTAL SECTION Chemicals and Reagents. Hydroquinone (HQ, 1,4-dihydroxyphenol), resorcinal (RS, 1,3-dihydroxyphenol), catechol (CA), dopamine (DA), 3,4-dihydroxyhenylaetic acid (DOPAC), norepi(5) Richardson, T. J.; Slack, J. L.; Rubin, M. D. Electrochim. Acta 2001, 46, 2281-2284. (6) Nair, M. T. S.; Guerrero, L.; Arenas, O. L.; Nair, P. K. Appl. Surf. Sci. 1999, 150, 143-151. (7) Lobbes, J. M.; Fitznar, H. P.; Kattner, G. Anal. Chem. 1999, 71, 30083102. (8) Morales, S.; Cela, R. J. Chromatogr., A 2000, 896, 95-104. (9) Sainthorant, C.; Morin, P.; Dreux, M.; Baudry, A.; Goetz, N. J. Chromatogr., A 1995, 717, 167-179. (10) Herberer, T.; Stan, H.-J. Anal. Chim. Acta 1997, 341, 21-34. (11) Uchimura, T.; Imasaka, T. Anal. Chem. 2000, 72, 2648-2652. (12) Murga, R.; Ruiz, R.; Beltran, S.; Cabezas, J. L. J. Agric. Food Chem. 2000, 48, 3408-3412. (13) Koganow, M. M.; Dueva, O. V.; Tsorin, B. L. J. Nat. Prod. 1999, 62, 481483. (14) de Armas, R.; Martinez, M.; Vicente, C.; Legaz, M.-E. J. Agric. Food Chem. 1999, 47, 3086-3092. (15) Canizares, P.; Dominguez, J. A.; Rodrigo, M. A.; Villasenor, J.; Rodriguez, J. Ind. Eng. Chem. Res. 1999, 38, 3779-3785. 10.1021/ac030183i CCC: $25.00

© 2003 American Chemical Society Published on Web 11/15/2003

Figure 1. Light-passable flow-through cell: (a) light source, (b) sheet glass, (c) silicon gel mode, (d) clamp, (e) Ag/AgCl reference electrode, (f) CuSPE, (g) PVC plate, (h) outlet tube, and (i) inlet stainless tube as auxiliary electrode.

nephrine (NE), and homovanillic acid (HVA) were purchased from Sigma (St. Louis, MO). All of the other compounds used in this work were ACS-certified reagent grade. Distilled, deionized water (E-pure water purification system, Barnstead) was used to prepare the standard solutions. A 200 ppm Cu(II) solution in 0.1 M HNO3 was used for the platting experiments. Unless otherwise stated, the base electrolyte solution was pH 8 phosphate buffer solution (PBS, I ) 0.1 M). Phenol standard solutions in pH 8 PBS were prepared daily and used for electrochemical experiments. Apparatus. Cyclic voltammetric and chronoamperometric (i-t) experiments were carried out with a CH 900 electrochemical workstation (CH instruments, Austin, TX). The three-electrode system consists of a CuSPE working electrode (∼0.2 cm2), an Ag/AgCl reference electrode, and a platinum or stainless auxiliary electrode. The flow injection system consists of a Cole-Parmer (Vernon Hills, IN) microprocessor pump drive, a Rehodyne

(Cotati, CA) model 7125 sample injection valve (20-µL loop) with interconnecting Teflon tube, and a BAS electrochemical thin-layer detecting system (West Lafayette, IN). The disposable SPEs were purchased from Zensor R&D (Taichung, Taiwan). The incorporation of the CuSPE into a thin-layer cell was similar to our earlier procedure.16,17 Electrode irradiation was performed by using an LSH-100F industrial illuminator (Taiwan Fiber Optics, Taiwan) equipped with an Osram 150-W lamp as the light source and an optical fiber as the light guide (active area, 6-mm i.d.). A prototype homemade flow-through cell (∼1 mL of volume) was designed for analytical measurements as illustrated in Figure 1. It consists of a CuSPE (16) Zen, J.-M.; Chung, H.-H.; Senthil Kumar, A. Analyst 2000, 125, 16331637. (17) Zen, J.-M.; Chen, H.-P.; Senthil Kumar, A. Anal. Chim. Acta 2001, 449, 95-102.

Analytical Chemistry, Vol. 75, No. 24, December 15, 2003

7021

RESULTS AND DISCUSSIONS Photoelectrochemical Response. The surface of a copper electrode is complex and contains both soluble and insoluble layers. It is known that copper forms passivating oxide films that consist of either a simple Cu2O film or a duplex Cu2O/CuO film.18 Different pretreated conditions such as preanodization or electrode irradiation may induce a different effect from the CuSPE. As illustrated in Figure 2, without any pretreatment, the cyclic voltammogram (CV) shows a distinct peak C1a (reduction of CuO) at -0.17 V and a small peak C2a (reduction of Cu2O) at -0.24 V, indicating that CuO is the major component in the oxide composition of a fresh-prepared CuSPE. Interestingly, after the electrode pretreatment of light illumination for 300 s, the peak potential of C1a shifts to a more positive potential (C1b) together with a decrease in peak current. On the contrary, the peak C2a shifts to a more negative potential (C2b) accompanied by an increase in peak current. The quantitative evaluation of the peak potential (E), charge (Q), and current (I) at both the C1 and C2 peaks is summarized in Table 1. It is noteworthy that the decrease in the charge of CuO (QC1a) is approximately half that of the increase of Cu2O (QC2b). Considering that double the electrons are needed for Cu2O to reduce equal moles of Cu2O and CuO, a transition of CuO to Cu2O by a synproportionation of CuO and Cu (that is CuO + Cu f Cu2O) through electrode irradiation is believed to occur.

A similar observation of the cathodic photocurrent leading to reduction of the CuO overlayer rather than to self-reduction of Cu2O to Cu was also reported in an earlier photoelectrochemical study regarding passive copper that contains a CuO layer of the duplex film.18 Our previous study of an oxygen sensor using the CuSPE also indicates that the photoelectrochemical activity is related to the formation of Cu2O.2 It is thus anticipated that the formation rate of Cu2O can be further increased through electrode preanodization. As expected, the increase for Cu2O formed was enormous when applying an anodic potential (+0.1 V versus Ag/AgCl) together with electrode irradiation for 300 s, as shown in peak C2c. To investigate whether the Cu2O component of the CuSPE can be used for photoelectrochemical analysis, the photoelectrocatalytic response of the CuSPE was then tested with various compounds. These compounds can be classified into two groups, consisting of (I) dihydroxyphenols and (II) neurotransmitters. More specifically, only HQ, RS, and HVA are not in the category of o-diphenolic compounds. It has shown that a stable bilayer film forms on the Cu electrode surface at slightly positive potentials.19 The first layer is reported to be a highly insoluble layer of Cu2O that shields the Cu surface from direct contact with the analyte. The second layer is believed to contain a highly soluble layer of CuO and cupric salts. Complexation with the Cu(II) ions results in an enhancement in the solubility of the outer layer and thus an increase in the steady-state current. A prototype homemade flow-through cell design was used here for the purpose of in situ electrode irradiation by FIA. As expected, the CuSPEs show a catalytic photochemical response at -0.1 V versus Ag/AgCl selectively to the oxidation of o-diphenolic compounds as shown in Table 2. It is likely that the anodic photocurrent, presumably a result of the formation of semiconductor Cu2O, can be useful for sensor application. Detailed analytical optimization will be discussed in a later section. Photoelectrocatalytic Mechanism. Catechol was chosen as a model compound for the mechanistic study because of its simple molecular structure and good photoelectrocatalytic behavior at the CuSPE. Figure 3 shows the effect of in situ electrode irradiation on the CV of the CuSPE in the absence and presence of CA. It is interesting to compare the CV for the case of electrode irradiation as shown in Figure 2. In the absence of CA, the only difference between these two experiments is preillumination (Figure 2) and in situ electrode irradiation (Figure 3). As can be seen, the peak C2 was found to increase at the expense of the peak C1, except that the anodic current at -0.1 V also increased in consequence to the formation of semiconductor Cu2O. This enhanced photoelectrochemical current (IPEC) was caused by backformation of CuO at an anodic potential (Ep). When CA was added to the solution, an obvious catalytic photocurrent was observed (Figure 3, line iii). Based on the observation, a possible catalytic oxidation mechanism on the CuSPE is proposed as shown in Figure 4. First, the semiconductor Cu2O is photochemically excited to promote electrons from valance band (VB) to conduction band (CB), creating an electronic vacancy or a hole at the VB. This electronic vacancy is capable of oxidizing molecules through path b under an oxidation potential that is less positive than the energy level

(18) Collisi, U.; Strehblow, H.-H. J. Electroanal. Chem. 1986, 210, 213-227.

(19) Brazill, S. A.; Singhal, P.; Kuhr, W. G. Anal. Chem. 2000, 72, 5542-5548.

Figure 2. Cyclic voltammograms of the CuSPE in pH 8 PBS: (a) without any pretreatment, (b) after electrode irradiation for 300 s, and (c) after preanodization at +0.1 V vs Ag/AgCl under electrode irradiation for 300 s.

working electrode, a stainless tube counter electrode, an Ag/AgCl reference electrode, a silicone gel gasket (2 mm in thickness), a PVC plate, and a glass plate (2 mm in thickness). Procedure. The CuSPE was first washed thoroughly with deionized water and then dipped into working solution that contains pH 8 PBS for subsequent static experiments. For flow injection analysis (FIA), the CuSPE was equilibrated in the carrier solution of pH 8 PBS at an optimized potential of -0.1 V versus Ag/AgCl until the current became constant. The quantification of interesting compounds was achieved by measuring the oxidation current from chronoamperometric signals. All experiments were performed at room temperature (25 °C).

7022

Analytical Chemistry, Vol. 75, No. 24, December 15, 2003

Table 1. Peak Potential (E), Charge (Q), and Current (I) Obtained under Various Pretreated Conditions (Data from Figure 2) conditions

C1

C2

expt

preanodizationa

irradiationb

E (V)

Q (µC)

I (µA)

E (V)

Q (µC)

I (µA)

a b c

no no yes

no yes yes

-0.17 -0.16 -0.16

79.59 54.09 60.57

32.91 20.18 18.77

-0.24 -0.26 -0.29

19.37 77.29 174.70

3.95 13.74 27.96

a

Preanodization time, 300 s at 0.1 V versus Ag/AgCl. b Electrode irradiation time, 300 s.

Table 2. Oxidation Electrochemical Currenta (Ioff) and Photoelectrochemical Current (Ion) Measured on the CuSPE for Various Compounds

Figure 3. Cyclic voltammograms of the CuSPE in (i) blank pH 8 PBS without electrode irradiation, (ii) blank pH 8 PBS with electrode irradiation, and (iii) pH 8 PBS contained 2 mM CA with electrode irradiation. CV conditions: scan from +0.1 to -0.5 V at a scan rate of 5 mV/s.

a Analyte concentration, 100 µM. FIA parameters: flow rate 100 mL/ min and Ep ) -0.1 V versus Ag/AgCl. b IIncr ) (Ion - Ioff)/Ioff.

of the VB. As a result of the flow of electrons in the CB via path a, a photocurrent is detected upon an applied Ep. The catalytic cycle is formed under a continuous electrode irradiation, and thus, a catalytic photocurrent is observed. Note that, because the recombination of the photogenerated electron and hole is rapid, interfacial electron transfer is kinetically competitive only when the electron donor or acceptor is preadsorbed.20 The formation of a substrate-catalyst five-member rf complex intermediate is considered as an essential factor not only for the selectivity but also for the interfacial electron transfer. Such a substrate-catalyst five-member complex (CuII-catecholate) was reported previously in the homogeneous solution-phase catalysis between the copper salts and catechol.21,22 Analytical Optimization. To optimize the analytical signal, the effects of flow rate and light intensity in FIA were studied

Figure 4. (A) Proposed mechanism for the photoelectrocatalytic oxidation of o-phenols. (B) An electron flow diagram based on the energy state of semiconductor Cu2O layer on the CuSPE.

(20) Fox, M. A.; Dulay, M. T. Chem. Rev. 1993, 93, 341-357. (21) Maumy, M.; Capdevielle, P. J. Mol. Catal. 1996, 113, 159-166. (22) Sayre, L. M.; Nadkarni, D. V. J. Am. Chem. Soc. 1994, 116, 3157-3158.

next. The FIA responses show an increase up to a flow rate of 100 mL/min for 100 µM CA at Ep ) -0.1 V under in situ electrode Analytical Chemistry, Vol. 75, No. 24, December 15, 2003

7023

Figure 5. (A) Amperometric responses for 100 µM CA under electrode irradiation at various light intensities. (B) Percentage increase in peak current and background photocurrent against light intensity. Experimental conditions: flow rate 100 mL/min and Ep ) -0.1 V (vs Ag/AgCl).

Figure 6. (A) Amperometric responses for the analysis of 10, 50, and 100 µM CA. (B) Calibration curve for CA. Experimental conditions: flow rate 100 mL/min, Ep ) -0.1 V (vs Ag/AgCl), and light power 120 W.

irradiation. The FIA response dropped rather quickly, however, when the flow rate was faster than 100 mL/min. Higher flow rate would be unfavorable to the formation of the substrate-catalyst five-member complex and hence causes a sudden decrease in the observed signals. This result further supports the proposed mechanism mentioned above, and an optimal flow rate of 100 mL/min was used in the subsequent experiments. Figure 5 shows the effect of light intensity on the amperometrically response of 100 µM CA by FIA. As can be seen, the increase in anodic photocurrent is proportional to the light intensity (through path a in Figure 4). Under optimal experimental conditions, ∼265% increase in current was obtained with a light power of 120 W. Furthermore, background photocurrent was also found to increase as the light intensity increases (through path b in Figure 4). Once again, these results are inconsistent with the proposed reaction mechanism. In brief, higher light intensity increases the amount of Cu2O formed and in turn the photocurrent produced regardless of going through either path a or path b. Figure 6 shows the FIA signals under optimized hydrodynamic conditions with/without light illumination. The FIA signals show good linearity up to 1 mM for both cases except with a big difference in slope, i.e., 9.86 × 10-3 µA/µM (R2 ) 0.9959) and

1.44 × 10-3 µA/µM (R2 ) 0.9945) with/without in situ electrode irradiation, respectively). An approximately 7 times increase in slope was obtained under light illumination. The measured detection limit (S/N ) 3) also improves from 10.00 to 0.84 µM under light illumination. Note that the detection limit without electrode irradiation is not as sensitive as our previous result because of the relatively larger cell volume in this study.1 There is still plenty of room to further improve the sensitivity as long as a stronger light source is used. The relative standard deviation was less than 3% for the measuring FIA signals, indicating good reproducibility of the present system. Experiments on the performance of the CuSPE in the presence of likely interferences were further studied. The interference of ascorbic acid, an electrochemically easily oxidizable reactant that is contained in biological liquids, was first evaluated. The interference due to ascorbic acid was completely arrested on the CuSPE at this low detection potential. To test the interference effect of other diphenols and phenols on the detection of o-diphenols, 10 times higher concentrations of Ph, Re, and HQ were added into CA. There was only a ∼3% decrease in the FIA signal of CA, validating the selective photoelectroanalytical assays of o-diphenols in the presence of other diphenolic compounds. Overall, these observa-

7024 Analytical Chemistry, Vol. 75, No. 24, December 15, 2003

tions demonstrate the capability of the CuSPE in sensitive and selective detection of o-diphenols at very low potentials. Future development of this research is to lower the detection limit by decreasing the cell volume and increasing the light intensity. CONCLUSION A CuSPE was successfully applied to the selective detection of o-diphenols in pH 8 PBS at a low oxidation potential of -0.1 V versus Ag/AgCl. The catalytically photoelectrochemical oxidation of o-diphenols was observed by using a homemade flow-through cell. The sensitivity increases in proportion to the light intensity as a result of the formation of semiconductor Cu2O. Meanwhile, it is the generation of a five-member complex intermediate (CuII-o-quinolate) being considered as an essential step for the selectivity toward o-diphenols. To further increase the sensitivity

to meet the requirement in biomedical study, the combination of a stronger light source and a better sensor design is necessary. Research along this line of direction by incorporating an all-inone unit comprising both the photoelectrode and light source is currently underway in this laboratory. ACKNOWLEDGMENT The authors gratefully acknowledge financial support from the National Science Council of the Republic of China.

Received for review May 12, 2003. Accepted August 26, 2003. AC030183I

Analytical Chemistry, Vol. 75, No. 24, December 15, 2003

7025