Environ. Sci. Technol. 2008, 42, 8465–8470
Simultaneous Determination of Nitrate and Dissolved Oxygen under Neutral Conditions Using a Novel Silver-Deposited Gold Microelectrode YOU-PENG CHEN,† SHAO-YANG LIU,† FANG FANG,† SHU-HONG LI,† GANG LIU,‡ YANG-CHAO TIAN,‡ Y I N G X I O N G , ‡ A N D H A N - Q I N G Y U * ,† Department of Chemistry and National Synchrotron Radiation Laboratory, University of Science & Technology of China, Hefei, 230026, China
Received April 11, 2008. Revised manuscript received July 28, 2008. Accepted August 14, 2008.
In this work a novel gold-based microelectrode was successfully fabricated using photolithographic techniques and electrochemical deposition for a simultaneous determination of nitrate and dissolved oxygen (DO) under neutral conditions. Three-dimensional tree-shaped silver nanorods were formed on the gold surface through electrochemical deposition and they had an electrochemical catalytic reductive activity for both nitrate and oxygen under neutral conditions. Thus, the silver nanorods served as the active center of the microelectrode. The microelectrode could be renewed over five times. Linear sweep voltammetry was employed to quantitatively analyze the nitrate and DO in solution. The microelectrode was used to measure the nitrate and DO microprofiles in a nitrifying aerobic granule from a sequencing batch reactor, which shows that denitrification did not occur in the tested granule. The measurement results demonstrate that the microelectrode was able to simultaneously determine the nitrate and DO levels in the granules under neutral conditions accurately and precisely.
Introduction Nitrogenous substances have been extensively used and widely present in the ecosystem, resulting in threatening human health and disquieting pollution to the environment (1, 2). Nitrate accumulated in the human body may cause methemoglobinemia, and furthermore, may become a source of carcinogenic N-nitrosamines (3). The excess presence of nitrate in water bodies can cause algal blooms and eutrophication. In current wastewater treatment systems nitrification and denitrification are the two essential processes for biological nitrogen removal (4). Dissolved oxygen (DO) is the electron acceptor in the nitrification process, which is described as follows: + NH+ 4 +2O2 ) 2H +H2O + NO3
(1)
* Corresponding author fax: +86 551 3601592; e-mail: hqyu@ ustc.edu.cn. † Department of Chemistry. ‡ National Synchrotron Radiation Laboratory. 10.1021/es8010157 CCC: $40.75
Published on Web 10/15/2008
2008 American Chemical Society
In the absence of DO and the presence of denitrifiers and an electron donor, nitrate will become the electron acceptor in the following denitrification reaction (4): + 2NO3 +10e +12H ) N2(g) + 6H2O
(2)
Thus, nitrate and DO are the two key components in the nitrification and denitrification processes for biological wastewater treatment. Biofilms (5), activated sludge (6), and aerobic granules (7) have been widely used for nitrification and denitrification in wastewater treatment. In this case, oxygen diffusion limitation may cause both aerobic and anoxic microenvironments in biofilms or granules, allowing a sequential utilization of oxygen and nitrate as electron accepters. Thus, accurate and rapid determination of nitrate and DO distributions inside biofilms or granules has attracted increasing interest (8, 9). As a powerful tool for microscale measurement, microelectrodes have been widely used in various fields, including physiology, medicine, neuroscience, microbial ecology, and environmental research (10). Compared with macroscale electrodes, microelectrodes have several advantageous properties, including smaller double-layer capacitance, faster response, higher sensitivity, and lower ohmic losses that result in a higher signal-to-noise ratio (11). Furthermore, microelectrodes, attributing to their micrometer scale tip, can be used for some special analyses with a high spatial resolution. Thus, microelectrodes have been regarded as the most appropriate tool for in situ measurements of nitrate and DO because of its size in micrometer scale. Copper electrodes have been employed for the quantitative analysis of nitrate with voltammetric and amperometric techniques based on electrochemical reduction (9, 12, 13). So far, a large number of other electrode materials, including nickel (14, 15), copper-nickel alloys (16), cadmium (17, 18), platinum (14, 19), glassy carbon (20, 21), and boron-doped diamond (22-24), have been investigated. However, a direct determination of nitrate by using these bare electrodes is difficult, mainly attributed to the slow charge-transferring speed (25). Moreover, a pH value as low as 3.0 is required for nitrate measurement with these electrodes (12, 26). On the other hand, many in situ analyses, such as determination of nitrification/denitrification, or in vivo detection, have to be performed under neutral conditions. Therefore, the in situ analysis of nitrate and DO under neutral conditions could not be achieved with these bare electrodes. Recently, a biosensor was fabricated to determine nitrate under neutral conditions. It was prepared by immobilizing nitrate reductase derived from yeast by using a polymer (poly(vinyl alcohol)) entrapment method (27). However, for this biosensor, DO was an interference to the measurement, and dosing of an oxygen scavenger, sulfite, prior to each test was essentially required as a pretreatment (27). In this work, a needle-type microelectrode was fabricated using photolithography and electrochemical deposition of three-dimensional tree-shaped silver nanorods. The microelectrode had a needle of 70 µm width, 100 µm thickness, and 5 mm length, on which three individual working microelectrodes existed. Under neutral conditions, nitrate and DO could be electrochemically reduced and quantitatively measured by the microelectrode. With this microelectrode, nitrate and DO distributions in aerobic nitrifying granules from a biological wastewater treatment reactor were measured to demonstrate its applicability. VOL. 42, NO. 22, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
8465
Experimental Section Chemicals. Potassium ferricyanide, potassium ferrocyanide, potassium chloride, sodium nitrite, sodium nitrate, sodium hydroxide, ferric trichloride, calcium chloride, ammonium chloride, sodium fluoride, sodium phosphate, sodium hydrogenphosphate, sodium dihydrogenphosphate, and silver nitrate, all purchased from Shanghai Chemical Reagent Company, were of analytical grade and used as received. Standard gases of various oxygen/nitrogen ratios were purchased from Nanjing Tianyuan Chemical Co., China. Deionized water was used throughout the experiments. Apparatus. Electrochemical measurements, including chronoamperometry, cyclic voltammetry (CV), and linear sweep voltammetry (LSV), were performed using an electrochemical detector (CHI 800, CH Instruments Inc., Austin, TX). All electrochemical experiments were carried out at 25 °C. Three-electrode configuration was employed. A commercial Ag/AgCl electrode (3 M KCl, CHI111, CH Instruments Inc.) was used as the reference electrode, while a platinum wire (CHI115, CH Instruments Inc.) was used as the counter electrode. The microelectrodes and a silver disk electrode with a diameter of 1 mm were used as the working electrodes. All potentials in this paper are referenced to the Ag/AgCl electrode. The scanning electron microscope (SEM) image of the microelectrode was obtained with a nanoengineering workstation (Sirion 200, FEI Ltd.). After electrochemical deposition and rinsing, the tip of the microelectrode was fixed on the sample platform of workstation. Later, the morphology of the surface was imaged. Preparation of the Microelectrode. The detailed fabrication procedures of the microelectrode array using photolithography were reported previously (28). The microelectrode arrays had a small needle of 70 µm width, 100 µm thickness, and 5 mm length, on which three individual microelectrodes existed. The needle with this dimension and hardness is sufficient for being inserted into the granules for measurement (28, 29). Unlike the microelectrodes fabricated in our previous work, the counter electrode reported here was not integrated in the microelectrode array. Instead, the gold tracks of the working electrodes (10 µm width) were embedded in cross-linked SU-8, and separated from each other by 10 µm. As a result, the three individual working microelectrodes could be used as a multisensor after appropriate functionalization. Before being functionalized as a nitrate and DO sensor, the needle tip of the microelectrode array was severed using a sharp knife to reveal the cross section of the gold tracks. After packing and renewing of the tip, one of the microelectrode was electrochemically modified. Chronoamperometry was employed for the electrochemical deposition on one of the gold surfaces on the cross section of the tip. An Ag/AgCl electrode was used as reference electrode, and a platinum wire was used as counter electrode. The electrochemical deposition was performed at -0.15 V (V vs Ag/ AgCl) for 150 s in a solution of 10 mM AgNO3 and 10 mM NaNO3. Later, the Ag-deposited microelectrode was rinsed with and stored in deionized water. Electrochemical Analysis. Cyclic voltammetry was employed to characterize the electrochemical behavior of the microelectrode before and after deposition. A Ag/AgCl electrode was used as reference electrode, and a platinum wire was used as counter electrode. The cyclic voltammograms were obtained in 1 mM K4Fe(CN)6/K3Fe(CN)6 solution with 0.1 M KCl (scan rate 50 mV/s). Three circles were recorded. Linear sweep voltammetry was used to characterize the electrochemical responses of the three electrodes (bare Ag, bare Au, and Ag-deposited) in four solutions (background, only DO, only nitrate, and both DO and nitrate). The 8466
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 22, 2008
background sample was a N2-saturated 0.01 M KCl solution (exposed to N2 for 20 min); the “only DO” sample was a 0.01 M KCl solution exposed to air over 30 min; the “only nitrate” sample was a N2-saturated 0.01 M KCl solution with 0.01 M NaNO3; the “both DO and nitrate” sample was a 0.01 M KCl and 0.01 M NaNO3 solution exposed to air over 30 min. Potassium chloride was the only supporting electrolyte in the solution for electrochemical measurements (scan rate 50 mV/s, potential -1.5-0 V vs Ag/AgCl). Linear sweep voltammetry was also employed to quantitatively analyze the DO levels and the nitrate concentrations in the solutions. The sigmoid peak at approximately -0.4 V (vs Ag/AgCl) in the linear sweep voltammograms was related to DO, while the peak at about -1.1 V (vs Ag/AgCl) was assigned to nitrate. The value of the current peak was proportional to the concentrations of DO or nitrate (scan rate 50 mV/s, potential: -1.5-0 V vs Ag/AgCl). The DO level was controlled by saturating solution with standard gases containing 0, 5, 10, 15, or 20% oxygen. For the DO calibration experiments the gases were introduced to the reaction cell for 20 min to reach an equilibration between the gases and liquid. To avoid bubble agitation, the gas supply was terminated prior to each measurement, but was resumed immediately after the measurement. For the nitrate calibration experiments, 0.001, 0.005, 0.01, 0.05, and 0.1 M NaNO3 solutions were used to prepare the different nitrate concentrations in the solutions. The calibration data are the average of five measurements. Various foreign ions, including Ca2+, K+, NH4+, Cl-, F-, H2PO4-, HPO42-, PO43-, Fe3+, and NO2- were added into 0.01 M NaNO3 solution to evaluate their interferences to the determination of DO and nitrate. Doses of 0.05, 0.1, and 0.2 M of the foreign ions were used. Varieties of the peak currents were evaluated. Measurements of Nitrate and DO in an Aerobic Nitrifying Granule. The measurement apparatus is illustrated in Figure 1A. The functionalized working electrode at the microelectrode tip was used to measure the DO and nitrate distributions in granules by using linear sweep voltammetry. The corresponding reductive peak current was proportional to the DO level and nitrate concentration respectively. A column-type sequencing batch reactor (SBR) was used for cultivating microbial granules, which were able to convert ammonium to nitrate accompanied with the consumption of DO. The synthetic wastewater was composed as follows (in mg/L): NH4Cl 200; K2HPO4 40; CaCl2 10; MgCl2 · 6H2O 5; FeSO4 · 7H2O 4.2. The pH was kept in a range of 7.0-8.5 through dosing of NaHCO3. The nitrifying granules (Figure 1B) with a diameter of 3-4 mm and solution in the SBR were sampled at the end of the operating cycle for measurements. The granules were held on the nylon net for the measurements in a stagnant system (Figure 1A). A micromanipulator was used to adjust the fine positions of the microelectrode tip at a spatial resolution of better than 5 µm, and a microscope was used for precisely locating the granule surface. The movement of the tip was perpendicular to the granule surface and the tip could be readily inserted into the granule. No buckling of the needle and no significant change of the granule were observed in the measurements. The data reported were an average of three measurements. More than three granules were analyzed with this nitrate and DO sensor and similar results were obtained.
Results and Discussion Microelectrode Functionalization. One of the working microelectrodes in the array, fabricated by using photolithographic techniques, was functionalized through electrochemical deposition method. Figure 2 shows the typical current vs time curve as the electrochemical deposition was performed at -0.15 V for 150 s in a solution of 10 mM AgNO3
FIGURE 1. (A) Photograph of the measurement apparatus for the aerobic granule and (B) microphotograph of the aerobic granule tested in this work.
FIGURE 2. Current as a function of time for the electrochemical deposition of silver on gold surface of the microelectrode in 10 mM AgNO3 and 10 mM NaNO3 solution at -0.15 V for 150 s. and 10 mM NaNO3. Silver ion was electrochemically reduced from the solution, transferred to silver atom, and then deposited on the gold surface at the negative potential. Silver atoms accumulated on the gold surface as time elapsed. The increase in current in the deposition process indicates that the active area of the silver coating on the gold surface increased continuously. Adhesion between the two SU-8 layers was sufficiently tight, as no detachment was observed in the fabrication, measurement, and renewal processes. Characterization of the Functionalized Microelectrode. Figure 3A shows the microelectrode arrays which had three individual working electrodes after and before packed respectively. The morphology of the active surface on the functionalized microelectrode was observed using SEM. Many tree-shaped silver nanorods were observed to be
FIGURE 3. (A) Photograph of the unpacked and packed microelectrodes, silver electrode with a diameter of 1 mm and (B) three-dimensional tree-shaped silver nanorods on the microelectrode surface.
FIGURE 4. Comparison of the cyclic voltammograms in 1 mM K4Fe(CN)6/K3Fe(CN)6 and 0.1 M KCl solution with bare gold microelectrodes and silver-deposited microelectrode (scan rate 50 mV/s, 3 cycles). located on the gold surface after the electrochemical deposition (Figure 3B). The active silver surface, with a large number of three-dimensional tree-shaped nanorods and large surface area, is responsible for the electrochemical reactions. The large active surface area is also confirmed by the substantially increased current as shown in Figure 2. Furthermore, the silver nanorods had electrochemical catalytic activity for the reduction of oxygen and nitrate. The electrochemical behavior of the microelectrode was evaluated. Figure 4 shows a comparison of CVs of the microelectrodes before and after electrochemical deposition of silver in a solution with 1 mM K4Fe(CN)6/K3Fe(CN)6 and 0.1 M KCl. The microelectrode had a higher current after silver deposition, indicating that it has a larger electrochemical active area than that of the bare gold microelectrode. VOL. 42, NO. 22, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
8467
FIGURE 6. DO and nitrate current response calibration curves at a scan rate of 30 mV/s.
FIGURE 5. Linear sweep voltammograms of (A) bare Ag with a diameter of 1 mm; (B) bare Au; and (C) Ag-deposited electrodes in the four solution samples at a scan rate of 30 mV/s. (Background: black curves, N2-saturated 0.01 M KCl solution; DO: red curves, 0.01 M KCl solution exposed to air; nitrate: blue curves, N2-saturated 0.01 M KCl and 0.01 M NaNO3 solution; DO + nitrate: purple curves, 0.01 M KCl and 0.01 M NaNO3 solution exposed to air. Potassium chloride was only the supporting electrolyte in the solution.) The two CVs were stable and reversible, suggesting that the working electrode was reliable. Performance of the Functionalized Microelectrode as a Nitrate and DO Sensor. The electrochemical response of the microelectrode to nitrate and DO was investigated. Figure 5 shows the linear sweep voltammograms of the bare silver electrode, bare gold microelectrode, and silver-deposited microelectrode in four solutions (background, only DO, only nitrate, both DO and nitrate). In Figure 5A, the DO reductive peak was observed at about -0.7 V (vs Ag/AgCl), while no nitrate reductive peak could be found. Similar phenomena could be also seen in Figure 5B. Therefore, the bare silver electrode or gold microelectrode could electrochemically reduce DO but not nitrate. In Figure 5C, both DO and nitrate could be electrochemically reduced by the silver-deposited microelectrode. The reductive peaks of DO and nitrate were 8468
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 22, 2008
at about -0.4 and -1.0 V, respectively. Furthermore, the currents of the two peaks could be used to quantitatively analyze the concentrations of DO and nitrate. The three linear sweep voltammograms with the three different electrodes in the DO and nitrate solutions are illustrated in Figure 5. The reductive peak of DO was observed at about -0.7 V for the bare gold microelectrode and the bare silver electrode, but the corresponding reductive peak of DO was obtained at about -0.4 V for the silver-deposited microelectrode. The oxygen reductive peak was shifted from -0.7 to -0.4 V, suggesting that the silver-deposited microelectrode has a better catalytic activity to the electrochemical reduction of oxygen than the bare gold or silver electrodes. Furthermore, the current of the DO peak for the silverdeposited microelectrode was much larger than that of the bare gold microelectrode. The increased DO peak current should be attributed to the catalytic effects and the larger active surface area of the three-dimensional silver nanorods on the microelectrode surface. The electrochemical reduction of nitrate was also observed for the silver-deposited microelectrode, but not for the bare gold or silver electrode, demonstrating that the reduction of nitrate might be associated with the tree-shaped silver nanorods on the microelectrode surface. Linear sweep voltammetry was employed to quantitatively determine the DO and nitrate concentrations in solutions (purple curve in Figure 5C). The reduction peak current represented a good linear response in a DO range from 0 to 7.94 mg O2/L at a scan rate of 30 mV/s in 0.01 M NaNO3 solution (Figure 6A). The correlation coefficient of the regression curves was 0.9983, and the largest relative standard deviation (N ) 5) was below 2% at 1.89 mg O2/L. The detection limit of DO was calculated to be 0.055 mg O2/L from the
FIGURE 7. DO and nitrate microprofiles of the microbial granule. Point 0 on the X-axis indicates the granular surface. Negative depth values (left of 0) are in the bulk solution, and positive depth values (right of 0) are in the granule. three times of the standard deviation of the control. Similarly, the nitrate reduction peak current also showed a good linear response in a range from 0.001 to 0.1 M NaNO3 at a scan rate of 30 mV/s (Figure 6B). The correlation coefficient of the regression curve was 0.9979, and the largest relative standard deviation (N ) 5) was below 5% at 0.001 M NaNO3. The detection limit of nitrate was calculated to be 1.5 × 10-4 M. The repeated utilization times could be used to indicate the stability and repeatability of the functionalized microelectrode. The responses of DO and nitrate in 0.01 M NaNO3 solution for consecutive 100 measurements were obtained. The DO reduction peak current response was almost at a constant level, while nitrate response became weaker at a slow rate. The nitrate response became 94% of the initial value after 100 measurements. These results demonstrate that the functionalized microelectrode was stable as a DO and nitrate sensor. Interference. The anti-interference ability of the functionalized microelectrode as a DO and nitrate sensor was also evaluated. Various amounts of foreign ions, including Ca2+, K+, NH4+, Cl-, F-, H2PO4-, HPO42-, PO43-, Fe3+, and NO2- were added into 0.01 M NaNO3 solution and their interferences were explored. All the ions, except Fe3+, at a maximum dose of 0.2 M were found to have no interference (response difference
