Investigation on Electrochemically Cathodic Polarization of Boron-Doped Diamond Electrodes and Its Inﬂuence on Lead Ions Analysis

The purpose of this paper is to further understand the dependence of the electrochemical activity of a boron-doped diamond (BDD) ﬁlm electrode on its surface boron doping concentration and the pretreatment applied to its surface. An electrochemically cathodic polarization ( − 3 V vs. SCE) with a varying polarization duration (5–50 min) was carried separately out to three types of BDDs with a boron doping level of 700, 2500, 8000 ppm in 0.5 M H 2 SO 4 solution. Electrochemical impedance spectroscopy (EIS) and cyclic voltammetry (CV) were used to evaluate surface activity of the BDDs with and without cathodic polarizations by using 1.0 mM Fe(CN) 63 − /4 − . Moreover, the surface activity of the polarized BDD electrodes was examined further by using anodic stripping voltammetry to detect a series of known concentrations of Pb 2 + . Our results indicated that the applied cathodic polarization signiﬁcantly improved the surface activity of the as-received BDD, and exhibited more active effect on the surface of a higher boron doping level BDD compared to a lower boron doping level BDD, thereby improving the ability of the polarized BDDs to detect trace Pb(II) in aqueous solutions by anodic stripping voltammetry.

Boron-doped diamond (BDD) electrodes not only exhibit similar chemical-stability to that obtained in graphite and glassy carbon electrodes, but also possess a rather wide electrochemical window (>3 V) and very low background current. Thus, it can be a promising replacement for dropping mercury electrode and classic other electrodes used in electrochemical sensors, 1 electrochemical synthesis, 2 electrooxidization decomposition of pollutants, 3 and electroanalysis. 4 BDD is characterized by p-type semiconductor [5][6][7][8][9] and its electrochemical property is dependent strongly on boron doping level, 10 purity of the diamond, crystal orientation of grain 11 and the kind of atom/group bonded chemically to carbon atoms on the surface of diamond. Generally, the electrochemical property of BDD electrodes is primarily determined by the terminal type of the carbon atoms on the surface. The surface termination of an as-deposited BDD electrode is H-termination, 9,[12][13][14][15] but the H-terminal surface is also converted to a O-terminated surface as it is exposed to air, 8 or an aqueous solution containing strong oxidizers, or subjected to an electrochemically anodic treatment. [16][17][18][19][20] The change in surface carbon atom termination alters undoubtedly the kinetics of redox reactions occurring on the BDD surface due to a distinct difference in electronic structure and surface energy between H-terminated and O-terminated surface. [21][22][23] Therefore, in order to make full use of the merits of BDD electrodes, it is important to make the type of surface carbon termination of a BDD electrode match up with application purposes.
The O-termination surface of a BDD presents high surface energy and positive electron affinity, 22,[24][25][26] thereby favoring hydroxyl radicals ( · OH) formation during anodic oxidation. 27 Thus, this feature has been widely used in oxidation degradation of organic pollutants in waste water. However, the O-termination is not suitable for electroanalysis because of its low electron transfer rate [28][29][30] and high surface pollution derived from its hydrophilicity. In contrast, H-terminated BDD electrodes are suitable especially to electroanalysis because its low surface energy and a negative electron affinity [31][32][33] that can increase the rate of interface charge transfer, 28,34 thereby improving the sensitivity and detection limit of heavy metal ions analysis. Therefore, a BDD electrode with H-terminated surface should be a preferred electrode used for an electroanalysis. However, because of the possible oxidation from air or anodic polarization, it is very difficult to retain H-terminated surface of a BDD unchanged. Therefore, finding an efz E-mail: a2028607@hotmail.com fective method to convert a O-terminated surface to a H-terminated is believed to be of practical interest for electroanalysis.
Basically, hydrogen terminalization of a BDD surface can be achieved by hydrogen plasma treatment [35][36][37] and electrochemically cathodic polarization. 8,15,29,[38][39][40][41][42] In comparison to hydrogen plasma treatment where the elevated-temperature vacuum environment can damage the BDD encapsulated by epoxy resin, an electrochemically cathodic pretreatment shows significant advantages such as simplicity and cost-effectiveness. Thus, cathodic polarization has been employed to pre-treat BDD electrodes. 8,41,42 However, because the resulting Htermination of a BDD electrode is determined largely by both cathodic polarization applied and the BDD itself, a given BDD should have its own optimal cathodic polarization.
This article reports the investigation on electrochemically cathodic activation of three kinds of boron doping level BDD electrodes. The properties of BDD electrodes after electrochemically cathodic polarization in different conditions had been characterized by electrochemical impedance spectroscopy (EIS) and cyclic voltammogram (CV). Furthermore, the effect of cathodic polarization was assessed by the determination limit and sensitivity of analyzing lead ions with LSSV.

Experimental
The BDD disks with final boron content of 700, 2500 and 8000 ppm, respectively, were purchased from a commercial company (Neo-Coat, Switzerland), and according to the manufacturer's instruction, boron doped polycrystalline diamond films (1.5-2 μm) were deposited on one side of a silicon wafer by using a hot-filament-assisted chemical vapor deposition technique, and a gold layer used for electrical contact was also deposited on the other side of the Si wafer. The SEM images of three as-received BDD disks showed that the asreceived BDD surfaces consist of randomly orientated polycrystalline diamond film, and the grain size decreased with an increase in boron doping level ( Figure 1).
All BDD disks tested were 8 mm in diameter, and a copper wire as an electric lead was individually adhered onto the gold-coated side of a BDD disk, and then the BDD disks were sealed separately except for a center area (0.07 cm 2 ) on the front face of a BDD, by heat-melting polytetrafluoro ethylene resin. The BDD electrodes were ultrasonically cleaned with ethanol, nitric acid and deionized water in turn prior to the electrochemical experiments. All electrochemical experiments were performed using CHI660E electrochemical workstation (CH Instruments, USA), in a 3-electrode cell where the BDD served as a working electrode, 1 cm 2 -Pt plate as a counter electrode, and saturated calomel as reference electrode.
All solutions were prepared using analytical grade reagents without further purification in deionized water. Unless otherwise stated, experiments were performed at room temperature (∼25 • C).
Electrochemically cathodic pretreatments were carried out at −3 V versus SCE for 5, 10, 20, 30, and 50 min, respectively. Electrochemical impedance tests were performed by applying a 10 mV AC signal in the 1 MHz to 0.1 Hz frequency range. The data were obtained at the opencircuit potential of the working electrode. Cyclic voltammograms were obtained at a scan rate of 50 mV/s over potential range of −0.2 V to +0.6 V. A 0.2 M KCl aqueous solution containing 1.0 mM Fe(CN) 6 3−/4− redox couple was used for EIS and CV tests. Linear sweeping stripping voltammetry (LSSV) was employed to determine the concentrations of Pb 2+ in 0.1 M HCl support electrolyte under the conditions in which the deposition time was set 5 min with 3 min rest period before anodic stripping voltammetry from −1.2 V to +0.6 V with a sweeping rate of 50 mV/s.

Results
Electrochemical impedance spectroscopy.-In order to estimate whether the electrode surface was suitable for the redox reaction, EIS was employed to appraise the electrochemical activity of BDD electrodes by using potassium ferricyanide as a redox probe. Figure 2a to 2c show the typical AC EIS of an as-received BDD electrode in 0.2 M KCl aqueous solution containing 1.0 mM Fe(CN) 6 3−/4− . All of these EIS are made of a semi-circle at the high frequency part and a straight oblique line at the low frequency part, which means that the electrode reaction occurring on the electrode surface is controlled by a dynamic hybrid process containing a charge transfer and a mass transport step. The semi-circle reflects the charge transfer process and the oblique straight segment reflects the process of mass transfer (diffusion). The rate of a total electrochemical reaction is the result of competition between the charge transfer and mass transport. In general, the diameter of the semi-circle in the impedance spectrum curve reflects the rate of the charge transfer directly, and the bigger the diameter of the semi-circle is, the higher resistance of the charge transfer step has. When the semi-circle shrinks into a small circle, and even disappears, it means that the charge transfer rate is much higher than the diffusion rate (typically at least 1000 times), then the rate of the electrode process is almost controlled by the diffusion step, and the electrode surface exhibits a very high redox activity. On the contrary, the greater diameter of the semi-circle means a slow electron transfer step controlling the total electrode process. Therefore, the diameter of the EIS semicircle reflects qualitatively the difficulty of the electron transfer process occurring on the electrode surface.
From Figure 2a-2c, it is clearly seen that each as-received BDDs has a semicircle with a diameter bigger than that of oneself being cathodically pretreated, indicating that the as-received BDD surface showed an obvious inhibition to the charge transfer. Furthermore, the diameter of the semi-circles decreased significantly with the cathodic polarization time, and the higher the boron doping level of the BDD was, the smaller the diameter of the semi-circle became, indicating that the cathodic polarization was more effective to a higher boron doping level BDD electrode, which is likely to be associated with the increased electron transfer rate derived from the higher surface conductivity of the BDD with high B-doping level. 21,41 In order to compare the effect of cathodic polarization quantitatively, we chose an equivalent circuit (Figure 2d) to simulate the redox process of Fe(CN) 6 3−/4− on the BDD surface. In the equivalent circuit, Rss is the sum of the solution resistance and the BDD surface resistance; Cdl is electric double layer capacitor; Rct is charge transfer resistance; Zd is diffusion resistance. Among the parameters, Rct is the most important parameter as its change in magnitude directly reflects the change in surface activity of the BDD tested. In principle, the smaller the Rct is, the higher the surface redox activity of the BDD possess. All of the data shown in Figure 2a  with a B-dopant level of 2500 and 8000 ppm should be suitable for electroanalysis due to their relative low electron transfer resistance.
Cyclic voltammetry.-In order to further examine the role of electrochemically cathodic polarization in improving the surface activity of BDD electrodes, a CV test was carried out on all the BDD electrodes after the EIS tests, respectively.  6 3−/4− . For 700 ppm BDD, a redox current peak appeared only in case of 50 min polarization (Figure 4a), and the potential difference between the oxidization and reduction peaks reached 367 mV, much larger than the value (∼70 mV) that an ideal electrode surface should possess. The large peak-peak potential difference indicated that the 700 ppm BDD exhibited a very poor surface activity even if it had been cathodically polarized for 50 min. This might be a typical character of all BDD with low B-dopant. 21 Unlike 700 ppm BDD where no reduction peaks appeared in cases of cathodic pretreatment durations from 5 to 30 min, both 2500 and 8000 ppm BDD presented their redox peaks (Figure 4b and 4c), regardless of the polarization duration. To compare the influence of the applied cathodic pretreatments on the surface activity of BDDs, Figure 4d shows their peak-peak potential differences varying with the cathodic polarization duration. It is worth noting that, in Figure 4d, 700 ppm BDD just had one value from 50 min cathodic pretreatment due to only the pretreatment causing a reduction peak occurring. As can be seen in Figure 4d, the peak-peak potential difference of the BDDs except 700 ppm BDD decreased with cathodic polarization duration, but there were significant differences in decline rate and the absolute value of potential difference. For 2500 ppm BDD, a sharp decline in potential difference occurred at its initial polarization duration (10 min), followed by a slow decline and approach to 110 mV or so. For 8000 ppm BDD, a sharp decline occurred within 30 min initial polarization duration, followed by a mild decline and approached to 75 mV or so. A comparison of the least value of peak-peak potential difference obtained under their own optimal polarization duration, where 10 min polarization led to 110 mV for 2500 ppm BDD, and 30 min led to 75 mV for 8000 ppm BDD, was made to assess the surface activity of the polarized BDD electrodes, and the result shows that 8000 ppm BDD might be the best one among the three kinds of BDDs, being in line with the results from EIS tests.
Analyzing lead ions by linear sweeping anodic stripping voltammetry.-To evaluate the effectiveness of the cathodic pretreatments for trace metal analysis of BDD, we detected a series of known Pb 2+ concentration solutions using LSSV. Lead being chosen as an example of heavily metal ions was based on the fact that lead contamination from air, food, soil and drink water has long been a major  global problem, and a rapid and accurate detection method for lead is thus essential to evaluate environment pollution. Additionally, 700 ppm BDD electrode was not selected to be used for Pb 2+ analysis due to its too poor surface activity. The 2500 and 8000 ppm BDD that have been cathodically polarized for different durations, were applied to a set of known Pb 2+ concentration solutions where C Pb 2+ was 0, 20, 45, 94, 143, 191 and 238 ppb, respectively. Figure 5 shows a typical LSSV curve of the 8000 ppm BDD that was cathodically pretreated for 30 min. From Figure 5, it can be clearly seen that the Pb 2+ stripping peak current values increased with increasing C Pb 2+ from 0 to 238 ppb, and the peak current values were plotted versus the corresponding C Pb 2+ and is also shown in the insert of Figure 5. Apparently, the curve in the insert was strictly linear within C Pb 2+ from 20-238 ppb, with a regression equation I p = 0.189C Pb 2+ + 0.035, and R 2 was 0.99662.
The identical measurement was made three times for the same BDD electrode, and the linear coefficients of variance were found not to exceed 5%, respectively, meaning that the detections with the pretreated BDD were very reproducible.
To examine the effect of the cathodic polarization duration on the performance of the pretreated BDDs in the determination of Pb(II) in solution, like the tests applied to the 8000 ppm BDD electrode pretreated for 30 min described above, a series of Pb(II) concentration tests were also applied to the 2500 ppm BDD and the other 8000 ppm BDD electrodes that were cathodically polarized for different durations. The sensitivity and detection limit for Pb(II) of each BDD electrode pretreated were obtained by fitting a series of linear curves of the stripping peak current values versus their Pb(II) concentrations, and the resulting sensitivities and detection limits for Pb(II) were also plotted as a function of cathodic polarization durations, as shown in Figure 6.
The cathodic pretreatment time significantly influenced the Pb(II)detecting capacity of the BDDs. With increasing pretreatment time, for both 2500 ppm BDD and 8000 ppm BDD, the sensitivity for Pb(II) increased rapidly and reached a maximum value, followed by a plateau, but significant differences are observed from Figure 6a in the maximum value of the sensitivity and in polarization time to reach the maximum value. The differences are likely due to the significant difference in boron doping level between them, because a higher boron doping concentration always generates a BDD surface with a higher density of active sites, thereby taking a longer time to hydrogenate such active sites. Likewise, such difference in B-dopant level may also contribute to the changing trend in the detection limits for Pb(II) (Figure 6b), in which 8000 ppm BDD presents the lowest detection limit (0.96 ppb) compared to 2500 ppm BDD (2.74 ppb), but the former needed a longer polarization time (30 min) than did the later (10 min).
The electrochemical property of a BDD electrode is strongly dependent on its surface boron doping level and surface termination. In general, a high boron doping level always renders the BDD electrode highly conductive, 21,41 thereby promoting redox reactions occurring on its surface. Our tests on BDDs with boron doping level from 700, 2500 to 8000 ppm also confirmed this point.
The surface termination has a profound effect on the electrochemical properties of the BDD electrode because the rate for electron transfer is primarily dependent on whether the surface is H-terminated or O-terminated. In our experiments, the as-received BDDs were believed to be primarily H-terminated, but their surface activities were found to increase significantly with increasing cathodic pretreatment time, as evidenced by the change in their charge transfer resistances ( Figure 3) and in the peak potential separation values (Figure 4d), indicating that the applied cathodic pretreatments led to surface activation of the BDDs, in agreement with the results reported by other groups. 8,22,[41][42][43] The cathodic pretreatment has been proven to play the same role in hydrogenating BDD's O-terminated surface as hydrogen plasma treatment. 14,37,40 As a result, the hydrogenation increases the surface state of the H-terminated BDD film, thereby promoting the charge transfer occurring at BDD electrode-electrolyte interface. 15,21 Conclusions Under our experimental conditions, the electrochemically cathodic polarization was demonstrated to be an effective way to increase surface activity of the as-received BDD electrodes, and the increase in the surface activity of the BDD was more effective for the BDD with a high boron doping level compared with the BDD with a low boron doing level. Moreover, the cathodically pretreated BDD electrodes exhibited a powerful ability to increase the detection sensitivity and decrease the detection limit for Pb(II). The results obtained in our tests can be helpful to select BDD type and enhance the surface activity of a BDD electrode used for electroanalysis.