Au Nanoparticle-Embedded Carbon Films for ... - ACS Publications

Feb 1, 2016 - National Institute of Advanced Industrial Science and Technology (AIST) ... Chiba Institute of Technology, 2-17-1 Tsudanuma, Narashino, ...
0 downloads 0 Views 4MB Size
Download PDF
Article pubs.acs.org/ac

Au Nanoparticle-Embedded Carbon Films for Electrochemical As3+ Detection with High Sensitivity and Stability Daiki Kato,†,‡ Tomoyuki Kamata,†,§ Dai Kato,† Hiroyuki Yanagisawa,‡ and Osamu Niwa*,†,∥ †

National Institute of Advanced Industrial Science and Technology (AIST), 1-1-1 Higashi, Tsukuba 305-8566, Japan Institute of Materials Science, Graduate School of Pure and Applied Sciences, University of Tsukuba, 1-1-1 Tennodai, Tsukuba 305-8573, Japan § Chiba Institute of Technology, 2-17-1 Tsudanuma, Narashino, Chiba 275-0016, Japan ∥ Saitama Institute of Technology, 1690 Fusaiji, Fukaya, Saitama 369-0293, Japan ‡

S Supporting Information *

ABSTRACT: Au nanoparticle (AuNP)-embedded carbon films were formed with a one-step reproducible process by using unbalanced magnetron (UBM) cosputtering to make it possible to detect As3+ in water. The sputtered Au components formed NPs (typically 5 nm in diameter) spontaneously in the carbon films, owing to the poor intermiscibility of Au with carbon. The surface contents of embedded AuNPs in the carbon film were widely controllable (Au = 13−21 at %) by regulating the target powers of Au and carbon individually. The obtained film had a flat surface (Ra = 0.1 nm) despite the fact the AuNPs were partially exposed at the surface. By anodic stripping voltammetry (ASV) As3+ detection, a limit of detection of 0.55 ppb and linear dynamic range of 1−100 ppb were obtained with our electrode. These values meet the requirements imposed by international regulation. Moreover, our electrode structure realized good electrode stability for repetitive ASV measurements (relative standard deviation (RSD) = 11.7%, n = 15) because the partially embedded AuNP structures prevented the AuNPs from detaching from the surface. This result was achieved by the electrode recovery only by a potential scan from 0.1 to 1.5 V. Our electrodes can be stocked for a long time (2 years) with maintaining the electrode performance, which is very attractive for practical electrode. Selectivity test by using Tsukuba tap water added 10 ppb As3+ and 1000 ppb Cu2+ was successfully achieved with existence of 0.1 M EDTA (RSD = 2.6%, n = 3). The ASV results with tap water samples agreed well with those by the conventional ICPMS method.

A

for the ASV of As because these metal-based electrode materials influence the measurement performance including sensitivity and detection limit. Among these electrodes, Au electrodes have been extensively used for As3+ detection, since Au is not toxic and interacts with As during the deposition and stripping steps.21,22 In particular, Au-coated carbon electrodes fabricated by electrodepositing Au particles on the surface of carbon electrodes including graphite, glassy carbon (GC),12−14 diamond thin film,15,16 boron doped diamond (BDD),17,20 carbon-paste,18 and carbon nanotubes (CNT)19 have been reported. This is because Au particles deposited on such surfaces exhibit higher electrocatalytic activity for As preconcentration than a bulk Au electrode,13 and it is difficult to preconcentrate As3+ ions on a pure carbon electrode surface. As a result, these Au deposited carbon electrodes achieve lower detection limits that satisfy the requirements imposed by the WHO guidelines (10 ppb). Au-electrodeposited GC electrodes, which mainly consist of sp2 bonded carbon, exhibited unstable

rsenic (As) is a toxic substance that is widely distributed in soil and rivers. Exposure to As, mainly through drinking water, can lead to various health problems involving, for example, the skin, lungs, urinary bladder, and kidneys.1,2 Therefore, the allowable As concentration in drinking water is strictly regulated. According to the drinking water guidelines published by the World Health Organization (WHO)2 and the Environmental Protection Agency (EPA),3 the As concentration in drinking water must be below 10 ppb (10 ng/mL). Currently, As is detected by using inductively coupled plasma mass spectrometry (ICPMS), inductively coupled plasmaatomic emission spectrometry (ICP-AES), and graphite furnace atomic adsorption spectrometry (GFAAS).3 These methods are highly sensitive, but not suitable for on-site measurements because of their large equipment and long measurement time. In addition, the analysis cost using these equipment is high. In contrast, anodic stripping voltammetry (ASV) is a well-known electrochemical technique for easily detecting various heavy metal ions. ASV has advantages compared with other methods because of its high sensitivity and low equipment cost. It is also easy to develop portable systems for on-site analysis.4,5 Various electrode materials such as Hg,6 Pt,7 Au,8−11 and Au coated carbon12−20 have been employed as working electrodes © XXXX American Chemical Society

Received: January 12, 2016 Accepted: February 1, 2016

A

DOI: 10.1021/acs.analchem.6b00136 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

Ultra (Al Kα 1486.6 eV) spectrometer to determine the elemental composition and the quantity of chemical bonds in the Au-embedded UBM carbon film surface. The Au concentrations in the films were estimated from the peak area ratio of the Au 4f corrected using instrumental sensitivity factors. The C 1s spectrum was fitted by Gaussian−Lorentzian sum components after the background had been subtracted according to Shirley’s methods.28 We prepared a sample for transmission electron microscopy (TEM) observation by scratching the film from the Si substrate surface with a diamond pencil. TEM and high-angle annular dark field scanning TEM (HADDF-STEM) images were collected on an EM-002B high-resolution TEM (Topcon) with a point-topoint resolution of 0.23 nm. The elemental mapping in the scanning TEM was operated by using STEM-energy dispersive spectroscopy (EDS). Electrochemical Measurements. Cyclic voltammograms (CVs) and square-wave voltammograms (SWVs) were obtained using a three-electrode configuration with an electrochemical analyzer model ALS/CHI 730C (CH Instruments, Inc.). In all the experiments, Pt wire and an Ag/AgCl electrode were used as auxiliary and reference electrodes, respectively. An Au-embedded UBM carbon film electrode (diameter = 2 mm) was used as the working electrode. A bulk Au electrode (diameter = 3 mm, BAS, Tokyo, Japan) were used as control electrodes in the experiment. All the SWV measurements were performed with a ΔE of 30 mV at 50 Hz. The detection limit was decided when the ASV peak current is 3 times higher than background noise level. Chemicals. All the chemicals were analytical grade and were used as received. The chemicals and materials used were As2O3 (Aldrich, 99.8−100.0%), AuCl3, (Aldrich, 99.9%), Cu(NO3)2 (Wako Pure Chemical, 99.9%), Na2HPO4 (Kanto Chemical, 99.9%), EDTA (Wako Pure Chemical), HCl (Wako Pure Chemical, 99.9%), and H2SO4 (Wako Pure Chemical, 99.9%). Ultrapure water (Milli-Q) was used in all the experiments. The application in real sample analysis was done by using Tsukuba tap water from Tsukuba Waterworks Bureau.

signals and poor reproducibility due to the passivation of the electrode surface.12 On the other hand, Au-electrodeposited diamond electrodes, which mainly consist of sp3 bonded carbon, displayed better stability than other carbon electrodes because a diamond electrode has advantages including a wide potential window in aqueous solution, low background current, and weak adsorption of polar molecules.15−17,20 Einaga et al. indicated that the responses obtained on Au-BDD carbon electrodes gradually decreased with repetitive measurements because the electrodeposited Au particles were detached from the electrode surface by potential scan particularly in high potential region.20 Therefore, they investigated As detection with BDD electrode added gold ions (100 ppm) into the solution as catalysis and achieved a highly sensitive and stable method. However, this method was not ideal for practical applications where a portable arsenic sensor was required. In this paper, we report a new electrode consisting of Au nanoparticles embedded in carbon film (Au-embedded carbon film) with which we realized the sufficiently sensitive and stable measurement of As3+ in water. This electrode can be formed by unbalanced magnetron (UBM) sputtering in a simple one-step process with high reproducibility. One of the authors has previously reported metal NPs-embedded carbon film electrodes realized by the cosputtering of carbon and various metals (Pt, Ni, Cu, Pd, Ir).24−26 These sputtered metal components spontaneously formed NPs (typically several nanometers in diameter) in the carbon films, owing to their low intermiscibility with carbon. However, one target was used for cosputtering by placing metal disks on the carbon target. This made it difficult to control the individual atom content because the sputtering rates of carbon and metals differ greatly. We recently reported how we developed an sp2 and sp3 hybrid carbon film electrode, which has high conductivity and stability, by using UBM sputtering.23,27 A carbon film matrix with sufficient conductivity and a wide potential window can be obtained by controlling the sp2 and sp3 ratio. In this study, the cosputtering of Au and carbon were controlled by regulating individual target powers to form fine AuNPs with high electrocatalytic activity embedded in the chemically stable carbon film. This film structure suppresses the detachment of AuNPs from the electrode surface since parts of the NPs were embedded in the chemically stable hybrid carbon film. These characteristics will enable us to electrochemically detect As3+ ions while maintaining high sensitivity and stability. We optimized the ASV measurement condition for As3+ and evaluated the limit of detection (LOD) and linear dynamic range (LDR), at the Au-embedded UBM carbon film electrode. Stability test for repetitive measurements and selectivity performance in the presence of interfering Cu2+ (1000 ppb) in the tap water samples of our electrode were also studied. The ASV results were compared with those by conventional ICPMS measurement.

■

RESULTS AND DISCUSSION Characterization of Au-Embedded UBM Carbon Film. Au-electrodeposited carbon electrodes have been employed for detecting As3+ because these electrodes have advantages including easy electrode fabrication and high sensitivity of As detection. Some groups reported that electrodeposited Au particles with small size (10−20 nm) exhibited high electrocatalytic activity of As predeposition.13,16 The smaller AuNP we fabricate, the more efficient performance of As detection we can expect. However, electrodeposition method is particularly difficult to control the size of Au particles less than 10 nm. In fact, to the best of our knowledge, no group has yet reported to use the electrode deposited with AuNPs smaller than 10 nm. In contrast, our Au-embedded UBM carbon films can be formed with uniformly and few nanometers in size of AuNPs only by controlling the sputtering conditions including the independently controllable sputtering powers of Au and carbon targets, gas pressure, and the bias voltage between the target and substrate. Indeed, a linear increase in the surface Au content (Au = 13, 17, 21 at %) estimated by XPS was observed when the Au sputtering power was increased (100, 160, 200 W) while maintaining a constant carbon sputtering power (Table S1). Moreover, C 1s XPS results indicated that the carbon film matrix consisted of sp2 and sp3 bonds (sp2/sp3 = 80:20). As we

■

EXPERIMENTAL SECTION Au-Embedded UBM Carbon Film Fabrication. An Auembedded UBM carbon film electrode was deposited on a silicon substrate by UBM cosputtering at room temperature. Argon (Ar) gas was used and the deposition pressure was maintained at 0.6 Pa. The Au target power ranged from 0 to 200 W, and the carbon target power was set at 400 W. The bias voltage between the substrate and the target was −20 V. Film Characterization. X-ray photoelectron spectroscopy (XPS) was performed using a Shimadzu/Kratos model AXIS B

DOI: 10.1021/acs.analchem.6b00136 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

Figure 1. TEM images of the Au-embedded UBM carbon film (Au = 17 at %) electrode: (A) plan and (B) cross-section view, (C) HAADF-STEM image, and EDS mappings of (D) Au and (E) carbon.

have already reported, carbon film containing sp3 bonds showed a wide potential window in aqueous solution, low background current, and weak adsorption of polar molecules than GC.23 In addition, the sp3 bond concentration can be increased by increasing the bias voltage between the target and substrate according to the requirements. This controllable sp3/ sp2 ratio is significant in terms of choosing the optimum structure for ASV measurement, particularly in regards to reducing the background current, despite the fact that carbon regions do not work as electroactive sites for As preconcentration. Other significant factors with respect to improving the sensitivity and detection limit of As detection are Au particle size and electrode surface flatness. Figure 1 shows typical TEM and HAADF-STEM images of the Au-embedded UBM carbon film (Au = 17 at %) electrode. The AuNPs in the films were about 5 nm in diameter and uniformly dispersed without aggregation in the carbon film as shown in the TEM (Figure 1A) and EDS mapping of each element (Figure 1C−E). Moreover, a cross-sectional image of the film electrode clearly demonstrates that the AuNPs are also uniformly distributed inside the film (Figure 1B). This suggests that repetitive measurements are possible even if the surface electrode layer is accidentally etched during measurement. Figure S1 shows AFM images of the Au-embedded UBM carbon film (Au = 17 at %) surface. The Au-embedded UBM carbon film electrode has a very flat surface with an average surface roughness (Ra) of 0.1 nm, which is similar to that of a pure UBM carbon film electrode (Ra = 0.3 nm).27 This suggests that AuNPs are embedded in the carbon film. Those are highly advantageous as regards achieving AuNP and carbon film with a stable structure because Au particles electrodeposited on the carbon electrodes detach easily from the electrode surface during ASV measurements.10,14 Moreover, Au particles electrodeposited on the carbon electrode surface had

various diameters and their minimum size was nearly 10−20 nm.13,16 It is noteworthy that the Au-embedded UBM carbon film can be fabricated by one-step sputtering, which provides a high throughput and an inexpensive process for fabricating metal and carbon hybrid film electrodes. Furthermore, since it has been reported that AuNP with a smaller size exhibits high catalytic activity,29,30 the AuNPs with an average diameter of 3−5 nm embedded in the UBM carbon film are expected to show a high S/N ratio and a low detection limit for the ASV measurement of As3+ thanks to their higher electrocatalytic activity for the As3+ deposition reaction and the relatively flat surface compared with that of the Au-electrodeposited carbon electrodes. Electrochemical Properties of Au-Embedded UBM Carbon Film Electrode. The electrochemical properties of the Au-embedded UBM carbon film electrodes with different surface Au contents were studied. The voltammograms of the Au-embedded UBM carbon film (Au = 17 at %) electrodes were also compared with that of a bulk Au electrode. Figure 2A shows the shapes of the oxidation and reduction peaks assigned to Au surface oxidation and reduction at both electrodes. This indicates the embedded Au particles were exposed from the surface of the UBM carbon film electrode. The reduction peak currents were 56.1 μA cm−2 for the Au-embedded carbon film and 341.8 μA cm−2 for the bulk Au electrodes, which reveal the effective Au surface areas of the electrodes. In fact, the effective Au surface ratio of Au-embedded UBM carbon film to bulk Au (16.4:100) agreed well with the surface Au content ratio estimated by XPS analysis (17.3:100, Table S1). We therefore corrected the voltammograms of the Au-embedded UBM carbon film electrode (Figure 2A,b) to multiply 100/17 (Figure 2A,b′; red dotted line) for comparison with the result obtained for the Au bulk electrode (Figure 2A,a; black line). As a result, the reduction peak currents of the Au-embedded UBM carbon C

DOI: 10.1021/acs.analchem.6b00136 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

Figure 3. ASV curves of 1000 ppb As3+ in 0.1 M Na2HPO4 with the Au-embedded UBM carbon film (red solid line, Au = 17 at %) and the bulk Au electrodes (black line). The red dotted line is ASV result (red solid line) corrected by Au concentration. ASV parameter: deposition at −0.8 V vs Ag/AgCl for 20 s, potential scan rate of 1.5 V s−1.

(422 μA cm−2, black line). Thus, small size Au particles play a key role in achieving high electrocatalytic activity for As preconcentration with only a small concentration of Au, as previously reported.29,30 We optimized the ASV measurement conditions (deposition potential and scan rate calculated by frequency (50 Hz) × various amplitudes (25−500 mV) to obtain a much clearer peak. Figure S2 shows the variation in the peak current that was realized by changing the deposition potentials. The peak currents increased as the deposition potential was decreased to −0.8 V but started to decrease when the deposition potential became more negative than −0.8 V. This suggests that less As is deposited due to the competitive reaction of H2 generation. Thus, an optimum deposition potential of −0.8 V was selected. We also studied the scan rate dependence of stripping peaks as shown in Figure S3. At a low scan rate, a well-defined As stripping peak could be observed but the peak current was also small. In contrast, the peak current increases as the scan rate increases because the magnitude of the stripping current is proportional to the scan rate. However, when the scan rate exceeds 2 V s−1, the stripping reaction cannot catch up with it, resulting in a peak potential shift and distortion of the peak shape. Thus, a scan rate of 1.5 V s−1 was determined as the optimum As3+ measurement conditions. Using the above conditions, we evaluated the LOD and LDR of As3+. LOD and LDR. Figure 4A shows stripping voltammograms for various As3+ concentrations obtained with the Auembedded UBM carbon film (Au = 17 at %) electrode. Figure 4B is a calibration curve obtained from the results in Figure 4A. The Δpeak current density increased linearly with increasing As3+ concentration. As a result, an LOD of 0.55 ppb (S/N = 3) and an LDR of 1−100 ppb were obtained. This LOD is much lower than the concentration recommended by the WHO guideline (10 ppb). We also investigated the effect of the surface Au content for As detection performance by preparing Au-embedded UBM carbon film electrodes containing 13 and 21 at % of Au. Table 1 summarizes these results. With both electrodes, we observed linear relationships between the stripping current and As3+ concentration, which was similar to the electrode with 17 at % of Au. As noted earlier, it has been reported that smaller size AuNPs showed high catalytic activity.29,30 If the AuNP size is increased, the sensitivity will tend to become saturated as the

Figure 2. (A) CV curves in 1 M H2SO4 for bulk Au (a, black line) and the Au-embedded UBM carbon film (b, red line, Au = 17 at %) electrodes. The red dotted line (b′) was corrected voltammograms of (b) to uniform the amounts of Au. (B) CVs of the Au-embedded UBM carbon film electrodes with different surface Au contents (Au = 21 at % (c, blue line), 17 at % (b, red line), 13 at % (d, green line), and 0 at % (e, dotted black line)). CV parameters: potential step from 0.1 to 1.5 V vs Ag/AgCl, potential scan rate of 0.1 V s−1.

electrode were the same as those of the bulk Au electrode. These results suggest that the effective Au surface area per total electrode area agreed well with surface Au content of the Auembedded UBM carbon electrode. Figure 2B shows CVs of the Au-embedded UBM carbon film electrodes with a range of surface Au content (Au = 0−21 at %). The obtained reduction peak currents were increased linearly with increasing surface Au content. Thus, we can control the effective Au surface area in the carbon film by changing the sputtering conditions. Detection of As3+ by ASV Measurement. An ASV measurement of As3+ was performed using the Au-embedded UBM carbon film (Au = 17 at %) and the bulk Au electrodes. As shown in Figure 3, a well-defined peak caused by the oxidation of As to As3+ was observed near 0 V vs Ag/AgCl in the presence of As3+ (1000 ppb) with Au-embedded UBM carbon film (red line) and the bulk Au electrode (black line), which was also similar to other reports.8−20 The obtained peak current density was 267 μA cm−2 for the Au-embedded carbon film and 422 μA cm−2 for the bulk Au electrode. By taking account of the actual Au electrocatalytic activity for As preconcentration, the electrocatalytic activity was defined as follows; the activity = peak current density/surface Au content. From Figure 3, we found that the electrocatalytic activity of the Au-embedded carbon film electrode (1540 μA cm−2, dotted red line) was 3.6 times higher than that of the bulk Au electrode D

DOI: 10.1021/acs.analchem.6b00136 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

almost the same LOD and LDR were obtained in spite of different surface Au content. The performance of the Au-embedded UBM carbon film electrode was compared with that of other previously reported electrodes including bulk Au and Au electrodeposited carbon electrodes (Table 1). The sensitivity (0.01−0.039 μA/ppb) and LOD (0.21−0.83 ppb) for the Au-embedded UBM carbon film electrodes were comparable to those of other methods reported in the literature.8,9,11,13−19 With the sensitivity and signal reproducibility of the obtained responses, the Au-embedded UBM carbon film (Au = 17 at %) electrode exhibited better performance than other Au-embedded UBM carbon film electrodes (Au = 13 and 21 at %). We therefore used the Au-embedded UBM carbon film (Au = 17 at %) electrode for stability and selectivity study. Electrode Stability. Since the working electrode for ASV measurements must be used repeatedly in practical applications, the signal reproducibility and electrode stability are very important. To compare those electrode performances, Auelectrodeposited UBM (sp2/ sp3 = 80:20) electrode prepared by the procedure adapted from previous reports.17,27 Figure 5A shows the variation in the stripping current for repetitive measurements (3 measurements per day) of As3+ (100 ppb) when using the Au-embedded UBM carbon film (Au = 17% at %) and the Au-electrodeposited UBM carbon film electrodes. Each electrode has almost the same electrochemical activity due to their similar magnitudes of Au surface of the reduction currents (Figure 5B,C). These results indicate that the almost same amounts of Au were exposed on both electrodes. The RSD for 3 measurements (first day) were 6.9% of the Auembedded UBM carbon film and 15.2% of the Au-electrodeposited UBM carbon film electrodes, respectively, which indicated the Au-embedded UBM carbon film electrode has superior reproducibility. After the experiment, the electrodes were stored in air. Next day, both electrodes were scanned from 0.1 to 1.5 V in 1 M H2SO4 to remove adsorption substances clean electrodes surface. Interestingly, highly stable current peaks were obtained at the Au-embedded UBM carbon film electrode during 5 days with an RSD of 11.7% (15 measurements/5 days). In contrast, the Au-electrodeposited UBM carbon film electrode initially exhibited a very large stripping peak current of 46 μAcm−2, but the peak current decreased rapidly and almost disappeared after 3 days of measurements.

Figure 4. (A) ASV curves for various As3+ concentrations in 0.1 M Na2HPO4 with the Au-embedded UBM carbon film (Au = 17 at %) electrode. ASV parameters: deposition at −0.8 V vs Ag/AgCl for 120 s, potential scan rate of 1.5 V s−1. (B) Calibration of As3+ concentrations obtained from part A. The values of the Δpeak current density was background current density (As3+ concentration = 0 ppb) subtracted peak current density.

surface Au content increases. However, in our case, the sensitivity increased linearly as the surface Au content increased (Table 1), indicating that individual AuNPs embedded in the carbon electrodes exhibited similar electrocatalytic activity for As3+ detection. In other words, this also suggests that the individual AuNPs are dispersed in the carbon film without aggregation in this surface Au content region. In addition,

Table 1. Comparison of As3+ Stripping Response of Au-Embedded Carbon Film Electrode Containing Various at % of Au with Au-Only and Au-on the Carbon Electrodes electrode Au-only

Au-on carbon

Au-embedeed carbon

electrode

LOD/ppb

sensitivity/μA μM−1

linearity range/ppb

ref

Au-UMEA sonically assisted gold microdisk electrode Au(111)-like polycrystalline gold electrode AuNPs/GC electrode AuNPs/GC electrode Au-coated diamond thin-film electrode AuNPs/BDD electrode gold−carbon composite electrode Au-CNTs AuNPs-embedded UBM carbon film electrode l3 atom % AuNPs-embedded UBM carbon film electrode 17 atom % AuNPs-embedded UBM carbon film electrode 21 atom %

0.01 0.28 0.28 0.01 l.8 0.01 1 0.38 0.6 0.21 0.55 0.83

0.044 0.363 0.3636 0.24 4.27 0.0097 2.13 0.133 0.129 0.020 0.026 0.039

0−500 7.5−75 0−1125 0−7.5 0−87 0.01−40 0−5 1.5−16.5

8 9 11 13 14 15 17 18 19 this work this work this work

E

1−100 1−100 1−100

DOI: 10.1021/acs.analchem.6b00136 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

measurements cycles and a longer term stability test (5 days). This indicates that the electrodeposited Au particles were detached from the electrode surface as a result of the repetitive ASV measurements particularly in the high potential region. In contrast, the Au embedded structure of our electrode prevented the detachment of AuNPs, which maintained electrode electrocatalytic activity over 5 days of measurement. To confirm this, we obtained CVs of both electrodes (Figure 5B,C) before (solid lines) and after 2 (dashed lines) and 5 days (dotted lines) ASV measurements, as in Figure 2. As a result, we observed almost the same peaks at the Au-embedded UBM carbon film electrodes (Figure 5B). In contrast, the peaks decreased at the Au-electrodeposited UBM electrodes (Figure 5C). These results clearly indicated that the AuNP structures partially embedded in the carbon film provide superior stability for repetitive ASV measurements while maintaining high electrocatalytic activity for As3+ reduction to the Au-electrodeposited UBM carbon film electrode. To confirm this fact, other carbon (GC, BDD) electrodes were also used for electrodeposition of Au particle by a previously reported method,17 and the same stability tests were performed, as in Figure 5B,C. Likewise, the peak currents of these Auelectrodeposited carbon electrodes were decreased by days (Figure S4). Therefore, we found that AuNP-embedded structure plays an important role in contributing not only high electrocatalytic activity but also electrode stability. Moreover, our electrodes can be stocked for a long time (2 years) with maintaining the electrode performance, which is very attractive for practical electrodes. Selectivity Measurements in Real Sample. The selective detection of As3+ in real samples is a challenging task because such samples typically contain many kinds of interfering substances such as other heavy metal ions. There are two mechanisms by which these metal ions interfere: competition with As during the deposition step and the formation of an intermetallic complex with As.11,13−16,19 Among the heavy metal ions in drinking water, Cu2+ mainly interferes with the ASV measurement of As and causes a decrease in the As stripping peak current. Therefore, Cu2+ ions must be removed from the real samples for the accurate detection of As3+. To evaluate the adverse effect of Cu2+ on the detection of As3+ (10 ppb), ASV was carried out in the presence of a large excess of Cu2+ (1000 ppb). In the presence of Cu2+ (Figure 6A, dotted lines), the peak current of As3+ (0.0 V, dotted red line) was unclear due to interference caused by the stripping competition of Cu2+ (0.2 V, dotted black line), unlike the result for As3+ alone. To eliminate this interference, EDTA was added because EDTA is known to strongly bind Cu2+ ions (log K(EDTACu2+) = 18.80).31 ASV measurements of As3+ were performed in the absence (solid black line) and presence of Cu2+ (solid red line) in 0.1 M Na2HPO4 containing 0.1 M EDTA. The two voltammograms we obtained were almost the same as seen by the solid lines in Figure 6A. This indicates that EDTA can bind just Cu2+ without masking As3+ despite the fact that the Cu2+ concentration is 100 times that of As3+. Organic matter such as humic and fulvic acids are known to interfere with the stripping detection of heavy metal ions due to adsorption on the Au electrode.15,16 However, any effect of the substances on the As response is likely to be minor because drinking water is thought to contain much lower levels of such organic substances.17,20 In addition, we have already reported UV irradiation pretreatment, which efficiently removes organic molecules from water samples without any reagents.23

Figure 5. (A) Stability of As measurement at the Au-embedded UBM carbon film (circle, Au = 17 at %) and the Au-electrodeposited UBM carbon film (triangle) electrodes. The average peak current density was obtained from repetitive As measurements (3 measurements per day). ASV parameters: deposition at −0.8 V vs Ag/AgCl for 60 s, potential scan rate of 1.5 V s−1. (B, C) CV curves in 1 M H2SO4 for the Auembedded UBM carbon film electrode (red lines, Au = 17 at %) and the Au-electrodeposited UBM carbon film (blue lines) before measurements (solid lines) and after 2 days (dashed lines) and 5 days (dotted lines) of measurement. The CV conditions are the same as those in Figure 2.

Swain et al. investigated the ASV detection of As with Aucoated (electrodeposited) diamond and glassy carbon electrodes and obtained RSDs of 1.5 and 4.4% for repetitive measurements (10 measurements), and 9.1 and 15.5% for long-term response stability test (10 measurements/10 h), respectively.15 Our results shows comparable RSDs to previous studies (10 h) in spite of employing 1.5 times the number of F

DOI: 10.1021/acs.analchem.6b00136 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

We confirmed the accuracy of our ASV results by comparison with the conventional ICPMS method. Tsukuba tap water with added As3+ (10, 50, 100 ppb) and Cu2+ (1000 ppb) was analyzed using both methods. Figure S5 shows the correlation between our ASV results and those obtained with the ICPMS method. This result revealed a high correlation between the results with a correlation coefficient of 0.999 even with tap water samples containing a high concentration of interfering Cu2+ ions.

■

CONCLUSION In this study, we achieved the simple fabrication of AuNP embedded carbon film electrodes formed by UBM cosputtering. The AuNPs (typically 5 nm diameter) were uniformly dispersed in the carbon films while maintaining a relativity flat surface (Ra = 0.1 nm) despite the fact that the AuNPs were partially exposed at the surface, which is unlike the Auelectrodeposited carbon electrodes. With the Au-embedded UBM carbon film (Au = 17 at %) electrodes, we obtained superior analytical performance for As3+ detection including an LOD of 0.55 ppb, an LDR of 1−100 ppb, and a sensitivity of 0.026 μA/ppb, which complies with WHO regulations. The reproducibility and stability of the Au-embedded UBM carbon film electrode are also superior to those of previously reported Au-electrodeposited carbon electrodes because the AuNPs embedded in the carbon film are resistant to detachment and aggregation and maintain high electrocatalytic activity for As3+ reduction. The Au-embedded UBM carbon film electrode also realized excellent selectivity of As3+ detection in interference (Cu2+) even with tap water samples by adding EDTA. The ASV results agreed well with those by ICPMS with tap water samples containing high concentration of interfering Cu2+ ions (r = 0.999).

Figure 6. (A, dotted lines) Effect of Cu2+ on As3+ detection at the Auembedded UBM carbon film electrode (Au = 17 at %) measured in 0.1 M Na2HPO4 solution in the absence (dotted black line) and presence (dotted red line) of Cu2+. (A, solid lines) Effect of EDTA addition on selective As3+ measurements at the Au-embedded UBM carbon film electrode measured in various solutions; As3+ in the 0.1 M Na2HPO4 solution containing 0.1 M EDTA (a, black solid line), As3+ + Cu2+ in the 0.1 M Na2HPO4 solution containing 0.1 M EDTA (b, red solid line), and As3+ + Cu2+ in Tsukuba tap water containing 0.1 M EDTA (c, blue solid line). (B) The average peak current density (three measurements) obtained from As3+ measurements under the same conditions as in part A, solid lines. The final concentrations of As3+ and Cu2+ were 10 and 1000 ppb, respectively. ASV parameters: deposition at −0.6 V vs Ag/AgCl for 300 s, potential scan rate of 1.5 V s−1.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.6b00136. Surface Au concentrations; AFM measurement; ICPMS analysis (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*Phone: +81-29-861-6158. Fax: +81-29-861-6177. E-mail: [email protected].

To test the feasibility of the more practical application of the Au-embedded UBM carbon film, we also undertook the detection of As3+ (10 ppb) in Tsukuba tap water, because tap water is a realistically complex matrix containing various metals (in ppm, 0.02 Al, 0.01 Fe, 0.001 Mn), organic matter (in ppm, 0.8), and other contaminants. Note that there was no Cu, As, Zn, Cr, or Pt (under 1 ppb). The tap water sample was prepared by diluting it (10 times) with the electrolyte solution, which contained 0.1 M EDTA and Cu2+ (1000 ppb). With the tap water sample (Figure 6A,c), we observed a similar current peak for As3+ to that in Figure 6A,b. This result indicates that this EDTA treatment is suitable for detecting As3+ from real samples. Moreover, the measurements seen in Figure 6A (three solid lines) demonstrated high reproducibility (RSD < 5.4% (n = 3)), as shown in Figure 6B. These facts indicated that it will be possible to detect As3+ directly from drinking water without the need to remove metal ions.

Author Contributions

All authors contributed to this manuscript and approved the final version. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported in part by a Grant-in-Aid for Scientific Research (O.N. No. 26288074) from MEXT, Japan. The work was also conducted in part at the Nano-Processing Facility, AIST, Japan.

■

REFERENCES

(1) Li, H.; Smart, R. B. Anal. Chim. Acta 1996, 325, 25. (2) World Health Organization. Fact Sheet No. 210, May 2001. (3) EPA. Report EPA-815-R-00-010, December 1991.

G

DOI: 10.1021/acs.analchem.6b00136 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry (4) Injang, U.; Noyrod, P.; Siangproh, W.; Dungchai, W.; Motomizu, S.; Chailapakul, O. Anal. Chim. Acta 2010, 668, 54. (5) Nie, Z.; Nijhuis, C. A.; Gong, J.; Chen, X.; Kumachev, A.; Martinez, A. W.; Narovlyansky, M.; Whitesides, G. M. Lab Chip 2010, 10, 477. (6) Holak, W. Anal. Chem. 1980, 52, 2189. (7) Wei, Z.; Somasundaran, P. J. Appl. Electrochem. 2004, 34, 241. (8) Feeney, R.; Kounaves, P. S. Anal. Chem. 2000, 72, 2222. (9) Simm, A. O.; Banks, C. E.; Compton, R. G. Anal. Chem. 2004, 76, 5051. (10) Prakash, R.; Srivastava, R. C.; Seth, P. K. Electroanalysis 2003, 15, 1410. (11) Rahman, R. M.; Okajima, T.; Ohsaka, T. Anal. Chem. 2010, 82, 9169. (12) Viltchinskaia, E. A.; Zeigman, L. L.; Garcia, D. M.; Santos, P. F. Electroanalysis 1997, 9, 633. (13) Dai, X.; Nekrassova, O.; Hyde, M. E.; Compton, R. G. Anal. Chem. 2004, 76, 5924. (14) Hossain, M. M.; Islam, M. M.; Ferdousi, S.; Okajima, T.; Ohsaka, T. Electroanalysis 2008, 20, 2435. (15) Song, Y.; Swain, G. M. Anal. Chem. 2007, 79, 2412. (16) Song, Y.; Swain, G. M. Anal. Chim. Acta 2007, 593, 7. (17) Yamada, D.; Ivandini, T. A.; Komatsu, M.; Fujishima, A.; Einaga, Y. J. Electroanal. Chem. 2008, 615, 145. (18) Simm, A. O.; Banks, C. E.; Wilkins, S. J.; Karousos, N. G.; Davis, J.; Compton, R. G. Anal. Bioanal. Chem. 2005, 381, 979. (19) Xiao, L.; Wildgoose, G. G.; Compton, R. G. Anal. Chim. Acta 2008, 620, 44. (20) Nagaoka, Y.; Ivandini, T. A.; Yamada, D.; Fujita, S.; Yamanuki, M.; Einaga, Y. Chem. Lett. 2010, 39, 1055. (21) Akhmetdzhanova, G. M.; Tikhomirova, V. I. Exp. Geosci. 1995, 4, 26. (22) Tikhomirova, V. I.; Akhmedzhanova, G. M.; Nekrasov, I. Y. Exp. Geosci. 1994, 3, 1. (23) Yanagisawa, H.; Kurita, R.; Kamata, T.; Kato, D.; Niwa, O. Electrochemistry 2014, 82 (11), 949. (24) You, T.; Niwa, O.; Tomita, M.; Hirono, S. Anal. Chem. 2003, 75, 2080. (25) You, T.; Niwa, O.; Chen, Z.; Hayashi, K.; Tomita, M.; Hirono, S. Anal. Chem. 2003, 75, 5191. (26) Kamata, T.; Kato, D.; Hirono, S.; Niwa, O. Anal. Chem. 2013, 85, 9845. (27) Kamata, T.; Kato, D.; Ida, H.; Niwa, O. Diamond Relat. Mater. 2014, 49, 25. (28) Shirley, D. A. Phys. Rev. B 1972, 5, 4709. (29) Haruta, M. Catal. Today 1997, 36, 153. (30) Haruta, M. Appl. Catal., A 2001, 222, 427. (31) Di Palma, L.; Mecozzi, R. J. Hazard. Mater. 2007, 147, 768.

H

DOI: 10.1021/acs.analchem.6b00136 Anal. Chem. XXXX, XXX, XXX−XXX