Analysis of Polychlorinated Biphenyls in Transformer Oil by Using

Sep 5, 2011 - Polychlorinated biphenyls (PCBs) that are present in transformer oil are a common global problem because of their toxicity and environme...
2 downloads 0 Views 4MB Size
ARTICLE pubs.acs.org/ac

Analysis of Polychlorinated Biphenyls in Transformer Oil by Using LiquidLiquid Partitioning in a Microfluidic Device Arata Aota,* Yasumoto Date, Shingo Terakado, Hideo Sugiyama, and Naoya Ohmura* Biotechnology Sector, Central Research Institute of Electric Power Industry, 1646 Abiko, Abiko City, Chiba, 270-1194 Japan ABSTRACT: Polychlorinated biphenyls (PCBs) that are present in transformer oil are a common global problem because of their toxicity and environmental persistence. The development of a rapid, low-cost method for measurement of PCBs in oil has been a matter of priority because of the large number of PCBcontaminated transformers still in service. Although one of the rapid, low-cost methods involves an immunoassay, which uses multilayer column separation, hexane evaporation, dimethyl sulfoxide (DMSO) partitioning, antigenantibody reaction, and a measurement system, there is a demand for more cost-effective and simpler procedures. In this paper, we report a DMSO partitioning method that utilizes a microfluidic device with microrecesses along the microchannel. In this method, PCBs are extracted and enriched into the DMSO confined in the microrecesses under the oil flow condition. The enrichment factor was estimated to be 2.69, which agreed well with the anticipated value. The half-maximal inhibitory concentration of PCBs in oil was found to be 0.38 mg/kg, which satisfies the much stricter criterion of 0.5 mg/kg in Japan. The developed method can realize the pretreatment of oil without the use of centrifugation for phase separation. Furthermore, the amount of expensive reagents required can be reduced considerably. Therefore, our method can serve as a powerful tool for achieving a simpler, low-cost procedure and an on-site analysis system.

T

he most important application of polychlorinated biphenyls (PCBs), prior to their discontinuation in the late 1970s, was in electrical equipment, including mainly capacitors as well as transformers, owing to their high chemical and thermal stability and low conductivity.1 However, as their toxicity and environmental persistence became apparent, their production and new applications began to be prohibited around the world. After the notorious Yusho incident in 1968,2 Japan became one of the first nations to ban the production and use of PCBs in 1972;3 the United States did the same in 1976.4 Eventually, PCBs were banned worldwide by the Stockholm Convention on Persistent Organic Pollutants (POPs) on May 23, 2001.5 Unfortunately, the high stability of the compounds coupled with their widespread use prior to the ban has resulted in a situation where PCB contamination still remains a global environmental problem. In Japan, the present regulations for transformer oil ordain that it must not have a PCB concentration of more than 0.5 mg/kg. This is a much stricter regulation than that currently imposed around the world; for example, the United States has imposed a maximum PCB concentration criterion of 50 mg/kg.6 In Japan, even though PCBs have deliberately not been used since 1973, PCB-contaminated transformer oil with concentration levels greater than 0.5 mg/kg can still be found in service and in storage.7 In order to find and eliminate these contaminated transformers, the Japanese government has mandated that every transformer in Japan be checked for PCB contamination when it is removed from service; the government has further recommended that the transformers still in service also be tested. The methods that have generally been applied to the detection of PCBs in transformer oil, which primarily include gas r 2011 American Chemical Society

chromatography with either mass spectrometry (GC/MS) or electron capture detection (GC/ECD), are time-consuming and costly, especially given the scale of the program. Many researchers have proposed simple analysis methods.810 Quintana et al.8 developed a microfluidic lab-on-valve for solid-phase microextraction. They determined the presence of PCBs in solid waste leachates by using a lab-on-valve interfaced with GC/MS. This process was 25 times less expensive than the offline and online solid-phase extraction counterparts. Criado et al.9 developed a PCB analysis using concentrated sulfuric acid treatment, dimethyl sulfoxide (DMSO) partitioning, solid-phase extraction, and GC/AED (atomic emission detection). By using the above pretreatment, they obtained good chromatograms. Takada et al.10 developed a PCB analysis using DMSO partitioning, solid-phase extraction, and GC/MS. Their cleanup procedure of oil samples is simple. The correlation between their method and the standard method was good when the PCB concentration in oil was more than 1 mg/kg. Since these methods use gas chromatography, the oil samples of transformers or oil-filled cables must be delivered and analyzed in laboratories or analytical centers. The storage volume is limited because the delivered oil samples must be strictly and safely stored in special rooms. Therefore, developing a rapid and low-cost method and an on-site analysis system for the measurement of PCBs has been a matter of priority. We previously reported the development of an immunoassay for PCBs in oil and a suitable extraction and pretreatment Received: June 21, 2011 Accepted: September 5, 2011 Published: September 05, 2011 7834

dx.doi.org/10.1021/ac2015867 | Anal. Chem. 2011, 83, 7834–7840

Analytical Chemistry process.1116 A multilayer solid-phase extraction for the cleanup of PCBs in oil, a flow-based highly sensitive immunoassay method, and a simple hand-held readout instrument have all been developed. These systems have been approved by the Japanese Ministry of the Environment and are now commercially available.17 The correlation between data from our method and from the standard method using high-resolution GC/highresolution MS was found to be good.15 Furthermore, its applicability for screening real oil samples has been evaluated by use of the receiver operating characteristic plots for 500 oil samples.16 In the case where the maximum permissible concentration of the total PCBs was set as 0.5 mg/kg, the most suitable cutoff concentration was calculated to be 0.4 mg/kg with false negative and false positive rates of 0% and 7%, respectively. However, the application of the present system to an on-site analysis is difficult because oil pretreatment requires large evaporation equipment. A reduction of the evaporation volume will lead to a simpler evaporation process or even make it possible to eliminate this process altogether. For this purpose, a suitable process for the pretreatment of a small volume of oil should be developed. In recent years, there has been great interest in lab-on-a-chip or micro total analysis systems (μ-TAS), which are microfluidic devices developed for miniaturizing chemical systems and integrating various chemical processes.1820 These devices have many advantages, including short analysis time; reduction of sample, reagent, and waste volume; more effective reactions owing to the large specific interfacial area and short diffusion distance; and portability of the analysis systems. If the chemical processes of the immunoassay for PCBs in oil are integrated on a microchip, a portable PCB analysis system may be realized. Microchemical processes can be designed by considering microunit operations, as proposed by Kitamori and co-workers.21,22 The chemical processes related to the immunoassay for PCBs in oil consist of solid-phase extraction, evaporation, liquidliquid extraction, mixing and reaction, surface reaction, and measurement. If the highly efficient and simple DMSO partitioning method is realized for a small volume of oil, the amount of expensive reagent required can be reduced, the instrument can be miniaturized, and the evaporation processes can be simplified or omitted. Here, we focused on liquidliquid extraction in microchips. Liquidliquid extraction methods in microchips have been previously investigated in many studies.21,2328 The methods employed in these studies mostly utilize droplets2325 or parallel multiphase microflows.21,2628 These methods are effective in achieving rapid and highly efficient liquidliquid extraction. The phase separation in these methods depends on the Laplace pressure induced by interfacial tension between the two phases.24,25,29,30 However, the interfacial tension between the insulating oil and the DMSO used for oil pretreatment is low. Therefore, phase separation via the Laplace pressure is not effective in this case. The other liquidliquid extraction method uses a confined extractant, which comes in contact with the sample solution under the flow condition.3133 This method confines the extractant through surface patterning31 or a microchannel structure.32 The former depends on interfacial tension, and the latter employs microrecesses. Microrecesses can be used for DMSO partitioning. Fang and co-workers32,33 demonstrated the preconcentration of butylrhodamine B in the microrecesses by laser-induced fluorescence and chemiluminescence detection. However, this method cannot be applied to immunoassay detection in the microrecesses owing to the instability of the antibody in DMSO. In this paper, we have reported DMSO partitioning, which is proposed for the cleanup of PCBs in insulating oil by use of a

ARTICLE

microfluidic device. Fluid control in a fabricated microfluidic device was examined. Dependence of the efficiency of extraction and elution on the flow rates and dependence of the measurement signals on the PCB concentration were investigated. The enrichment was estimated from the comparison between conventional and microfluidics-based methods. Finally, the applicability of the developed method for practical use has been discussed.

’ EXPERIMENTAL SECTION Chemicals. Commercial mixtures of PCBs (Kanechlor [KC] 300, 400, 500, and 600, 1:1:1:1 mixture) were purchased from GL Sciences Inc. (Tokyo, Japan). DMSO (catalog no. 346-03615) was purchased from Dojindo Laboratories (Kumamoto, Japan). Bovine serum albumin (BSA) and fluorescein sodium salt were purchased from Sigma Aldrich (Tokyo, Japan). Hexane was purchased from Wako Pure Chemical Industries Ltd. (Osaka, Japan). Phosphate-buffered saline (PBS, consisting of 137 mM sodium chloride, 3 mM potassium chloride, 20 mM disodium hydrogen phosphate, 1.5 mM potassium dihydrogen phosphate, and 1.5 mM sodium azide, pH 7.0) was prepared in-house. PBSB consisted of PBS supplemented with 0.1% (w/w) BSA. PBSBD consisted of PBSB supplemented with 2% (v/v) DMSO. PCB-free insulating oil for the electric transformers was purchased from Matsumura Oil Co. Ltd. (Osaka, Japan). The pretreatment column kit was purchased from Sumika Chemical Analysis Service Ltd. (Osaka, Japan). The positive photoresist (S-1818) was purchased from Nippon Kayaku Co. Ltd. (Tokyo, Japan). Fluorescent polystyrene microparticles (F8823) were purchased from Life Technologies Japan Ltd. (Tokyo, Japan). Monoclonal antiPCB antibody was purchased from Kyoto Electrics Manufacturing Co. Ltd. (Kyoto, Japan). Cy-5-labeled F(ab0 )2 fragment goat antimouse IgG (heavy and light chains) was purchased from Jackson ImmunoResearch Laboratories Inc. (West Grove, PA). Microchip. The microchip was fabricated by a photolithographic wet etching method. Tempax glass plates that were 0.7 mm thick were used (top and bottom plates). The glass plates were washed with piranha solution (3:1 v/v mixture of 95% sulfuric acid and 30% hydrogen peroxide), distilled water, and 2-propanol. A Cr layer was sputtered on the glass plates. A positive photoresist was spin-coated on the Cr layer and baked at 95 °C. UV light was exposed through a photomask by using a mask aligner to transfer the microchannel pattern onto the photoresist. The photoresist was developed, and a pattern was obtained. The photoresist was baked at 180 °C for 30 min. The Cr layer was etched with ammonium cerium nitrate solution. The bare glass surface with the microchannel pattern was etched with a 47% hydrofluoric acid solution. After the etching, the remaining photoresist was removed by the piranha solution and Cr was removed by the ammonium cerium nitrate solution. The microchip was washed with ethanol and hexane after each experimental use, and it was repeatedly used in the experiments. Operating Procedures. The experimental setup for the flow control was similar to that described in the previous paper.34 Briefly, the solutions were introduced through microsyringes and flow rates were controlled via microsyringe pumps. A poly(ether ether ketone) (PEEK) tube was used to connect the syringes to steel blocks with Teflon ferrules, which were pressed onto an inlet of the microchip. PCB Immunoassay. PCB immunoassays were performed on the KinExA 3000 supplied by Sapidyne Instruments Inc. (Boise, ID). The theory and operation of this instrument have been previously described by us11,35,36 and by others.3739 Briefly, 7835

dx.doi.org/10.1021/ac2015867 |Anal. Chem. 2011, 83, 7834–7840

Analytical Chemistry

ARTICLE

Figure 1. Schematic of multilayer silica gel column used for cleanup of PCBs in oil.

antigens were immobilized on N-hydroxysuccinimide (NHS) activated Sepharose as described in our previous paper.40 The antigen-coupled solid phase is captured in a small flow cell held at the focus of a fluorometer. Samples, which are extracted oil in the present case, are mixed with a detection antibody and flowed through the solid phase. The detection antibody that binds the analyte in the sample solution cannot bind to the antigen in the solid phase and passes through the flow cell. A small fraction of the antibody that is not bound in the solution is captured on the solid phase. The captured detection antibody is labeled with a fluorescently labeled anti-species antibody, and the binding signals are measured as fluorescent accumulation on the solid phase. A zero PCB sample gives the highest response, and very high concentrations of PCB lead to total inhibition of the specific binding signal. In our experiments, the concentration of the antibody prepared was set as 250 pM and the measurement solutions were incubated for a minimum of 30 min at room temperature. The DMSO concentration of the measurement solution prepared was set as 2% (v/v). Oil Pretreatment. The cleanup of PCBs was performed on a multilayer column supplied with the commercial kit developed by us. The procedure was based on our method described in a manual published by the Japanese Ministry of the Environment.17 In brief, the upper layer was 0.75 g of anhydrous sodium sulfate, the second layer was 2 g of oleum-impregnated silica gel, the third layer was 0.75 g of anhydrous sodium sulfate, and the lower layer was 1.5 g of aminopropyl silica gel, as shown in Figure 1. A quantity of 295 μL of the oil to be tested was added to the column, and 3 min was allowed for reaction between the oleum and the hydrocarbon chains in the oil. Next, 75 μL of hexane was added twice to the column to promote infiltration of the oil in the upper layer or around the column wall into the oleum-impregnated layer. One minute was allowed for the reaction. Then 10 mL of hexane was added to the column and all of the liquid that passed through the column was collected. The hexane was evaporated on a rotary evaporator with a hot water bath at 40 °C. The remaining liquid was used as experimental oil samples for the microchip-based DMSO partitioning.

’ RESULTS AND DISCUSSION A schematic of the concept of the microchip-based DMSO partitioning is shown in Figure 2. The microrecesses aligned

Figure 2. DMSO partitioning procedures performed by using the microchip. (A) The microchip has a number of microrecesses aligned along both sides of the main microchannel. (B) First, the microchannels were filled with DMSO. (C) Second, the PCB-contaminated oil was introduced into the main microchannel. The oil flowed only in the main microchannel. (D) Third, the oil was removed by introducing air flow. DMSO remained in the microrecesses under the air flow. (E) Finally, the PCBs flowed out by introducing the buffer solution and eluting the DMSO.

along the main microchannel are used for confinement of the DMSO used as extractant of the PCBs. After the microchannel is filled with DMSO, the PCBs in the oil are extracted into the DMSO by passing the PCB-contaminated oil through the microchannel. The DMSO remains in the microrecesses under the oil flow condition. The oil is removed by the air flow used for preventing oil contamination of the samples. Finally, the PCBs in the DMSO are flowed out by buffer elution. The obtained samples are measured by immunoassays. We have considered the enrichment factor in our concept. When conventional batch DMSO partitioning for the same volume of DMSO and oil is performed, the PCB concentration after DMSO partitioning can be expressed as K ¼

CO COi  CD ¼ CD CD

ð1Þ

where K is the apparent partition coefficient of the total PCB congeners and CO, CD, and COi are the PCB concentrations in oil and DMSO after batch extraction and the initial PCB concentration in oil, respectively. In our concept, the DMSO in the microrecesses continuously comes in contact with high PCB concentration oil because fresh oil continuously flows through the main microchannel. Here, the PCB concentration in DMSO is considered when the batch extraction is repeated. Therefore, the PCB concentration in DMSO in our concept can be 7836

dx.doi.org/10.1021/ac2015867 |Anal. Chem. 2011, 83, 7834–7840

Analytical Chemistry

ARTICLE

Figure 3. Photograph and SEM image of the microchip. The main microchannel had a width of 250 μm, depth of 50 μm, and length of 610 mm. Each of the 1472 microrecesses had a circle-like structure, width of 250 μm, and depth of 50 μm. Since the volume of each of the microrecesses was estimated to be 2.3 nL, the total volume of the microrecesses was estimated to be 3.5 μL.

expressed as CDmicro ¼

COi K

ð2Þ

where CDmicro is the PCB concentration in DMSO after microfluidic extraction. By use of eqs 1 and 2, the enrichment factor between the batch extraction and microflow extraction can be expressed as CDmicro K þ 1 ¼ K CD

ð3Þ

In order to estimate the enrichment factor, K was estimated by measuring the PCBs. PCB-contaminated oil was prepared by adding the commercial mixture of PCBs to pure insulating oil pretreated by the multilayer column and evaporation. A quantity of 300 μL of 2.22 ( 0.17 mg/L PCB oil was extracted to 300 μL of pure DMSO by vortex mixing for 1 min. The mixture was centrifuged at 10 000 rpm for 1 min. A quantity of 170 μL of the lower DMSO phase was collected without the oil. The PCBs in the collected DMSO were measured by use of the KinExA 3000. The PCB concentration in the collected DMSO was estimated to be 1.37 ( 0.13 mg/L; measurement errors corresponded to the standard deviation of triplicate measurements. Therefore, K and the enrichment factor were calculated to be 0.62 ( 0.12 and 2.61 ( 0.51, respectively. Figure 3 shows the photograph and scanning electron microscopy (SEM) image of the microchip fabricated for this study. The main microchannel had a width of 250 μm, depth of 50 μm, and length of 610 mm. Each of the 1472 microrecesses had a circle-like structure, a width of 250 μm, and a depth of 50 μm. Since the volume of each of the microrecesses was estimated to be 2.3 nL, the total volume of the microrecesses was estimated to be 3.5 μL. Flow control in the microchip was examined by dissolving fluorescein sodium salt in DMSO. First, the fluorescent DMSO was introduced in the microchip. Since fluorescence was observed in all the microrecesses, as shown in Figure 4A, it is evident that the microrecesses would be filled only with DMSO without any air bubbles. Second, oil was flowed in the main microchannel. Although small oil droplets intruded into a few microrecesses, the oil flowed only in the main microchannel as shown in Figure 4B. Third, the oil in the main microchannel could be purged by introducing air, as shown in Figure 4C. Finally, PBSB was introduced in the main microchannel. The fluorescent DMSO flowed out toward the outlet of the microchannel by dissolving in PBSB, as shown in Figure 4D. The oil droplets remained in the

Figure 4. Fluid motions in the microchip with microrecesses. Fluorescent DMSO is represented as gray colored areas. (A) Microrecesses were successfully filled with DMSO without any air bubbles. (B) Oil flowed in the main microchannel. Oil droplets rarely intruded into the microrecesses. (C) Oil in the main microchannel was removed and the DMSO remained in the microrecesses after the air flow. (D) DMSO shown in gray was eluted by the buffer solution. Intruded oil droplets remained in the microrecesses after the elution process.

Figure 5. Streamlines in the microrecesses. White lines correspond to the motion of the fluorescent microparticles. A vortex was generated when the oil flowed in the main microchannel.

microrecesses after the elution process. Therefore, the measurement sample solution did not contain oil, which interferes with the immunoassay. The applicability of the fabricated microchip for DMSO partitioning was verified by the experiments. The streamlines in the microrecesses were observed by use of fluorescent microparticles (Figure 5). Fluorescent microparticles 7837

dx.doi.org/10.1021/ac2015867 |Anal. Chem. 2011, 83, 7834–7840

Analytical Chemistry

Figure 6. Dependence of DMSO partitioning efficiency on oil flow rates. Error bars are (1 SD based on triplicate measurements.

were seeded into the DMSO at a concentration of 0.02% by volume, and they dispersed uniformly. After introduction of DMSO containing the microparticles in the microrecesses, the oil flowed at a flow rate of 10 μL/min. A vortex was observed in the microrecesses. It was caused by shear force at the interface between DMSO and oil. This vortex flow enhances molecular transport process. The extraction efficiency depends on the contact time between DMSO and oil. The enrichment condition indicated in eq 3 cannot be realized at a high flow rate because the contact time is insufficient for transport of the PCBs to the DMSO. The dependence of extraction efficiency on contact time was examined by changing the oil flow rates. A quantity of 50 μL of pretreated oil with 0.5 mg/kg PCBs was flowed in the main microchannel at flow rates of 10, 20, 40, 100, and 300 μL/min. After the air flow, the DMSO was flowed out by flowing 1 mL of PBSB at 100 μL/min. A quantity of 900 μL of elution solution was prepared to obtain a total volume of 1.5 mL after addition of PBSB, DMSO, and antibody. Since the volume of DMSO in the microrecesses was 3.5 μL, the dilution rate was 476. Relative responses with error bars, resulting from three replicates of the experiments, are shown in Figure 6. Lower relative responses indicate higher concentration of the PCBs in DMSO. The relative responses decreased as the flow rates decreased. There was no difference between the relative responses obtained at flow rates of 10 and 20 μL/min. Therefore, the 20 μL/min flow rate gave sufficient contact time under our experimental conditions. Henceforth, we have used a flow rate of 20 μL/min for DMSO partitioning in our experiments. The elution of DMSO also depends on the flow rate of PBSB, owing to the transport of the PCBs. Dependence of elution on the flow rate of PBSB was examined. A quantity of 1 mL of PBSB was flowed in the main microchannel at flow rates of 100, 300, 500, 700, and 900 μL/min. A quantity of 900 μL of elution solution was prepared to obtain a total volume of 1.5 mL after addition of PBSB, DMSO, and antibody. Relative responses with error bars, resulting from three replicates of the experiments, are shown in Figure 7. Lower relative responses indicate higher elution efficiency. No difference was observed between the relative responses at flow rates of 700 and 500 μL/min. Therefore, a flow rate of 700 μL/min provided sufficient contact time under our experimental conditions. Henceforth, we have used a flow rate of 700 μL/min for the elution process in our experiments. Finally, the concentration dependence of DMSO partitioning in the microchip was examined. Figure 8 shows the standard curve

ARTICLE

Figure 7. Dependence of PCB elution efficiency on flow rates of buffer solution. Error bars are (1 SD based on triplicate measurements.

Figure 8. Dependence of relative responses on PCB concentrations in the oil. Experimental results and fitting curves for (b, —) the microfluidic device and (9, ---) the conventional batch method are shown.

with error bars resulting from three replicates of the experiments. Solid circles correspond to results of the microchip-based DMSO partitioning. Solid squares correspond to the batch DMSO partitioning used as a control experiment. The dilution rate in the batch DMSO partitioning was set as 303. The results were fitted with the following four-parameter logistic equation: ðA1  A2Þ  þ A2 y¼ 1 þ ðx=x0 Þp

ð4Þ

where A1 is the maximum system response, A2 is the minimum response (nonspecific binding), x0 is the concentration of 50% maximum response, and p is the slope factor.40 When the range of measurement is defined as the relative response between 10% and 90%, the ranges of measurements in the batch and microfluidic method were estimated to be 0.055.66 and 0.045.00 mg/kg, respectively. The Japanese regulation criterion of 0.5 mg/kg could be realized. Half-maximal inhibitory concentrations in the batch and microfluidic methods were estimated to be 0.38 and 0.65 mg/kg, respectively. When each dilution rate was considered, the enrichment factor for the microfluidic method was estimated to be 2.69, which agreed well with the theoretical value of 2.61. Therefore, 7838

dx.doi.org/10.1021/ac2015867 |Anal. Chem. 2011, 83, 7834–7840

Analytical Chemistry our concept of DMSO partitioning using a microfluidic device was demonstrated. The developed method realized the DMSO partitioning by use of 50 μL of the oil sample, phase separation only by solution introduction, and enrichment of PCBs in the DMSO. Although 295 μL of oil sample is required in the conventional method, only 50 μL of oil sample is needed in the developed method. Therefore, the amount of expensive reagents required for the multilayer column can be reduced to 1/6 the previous amount. Furthermore, centrifugation for phase separation can be omitted. We have already developed hand-held measurement instruments for PCB immunoassay.15 If a simple evaporation process for on-site operation or omission of the evaporation process is developed, an on-site analysis system can be realized. A quantity of 1.7 mL of hexane is used for elution during column pretreatment in our method. We tried to recover 50 μL of oil residue by collecting the eluate in a microtube and evaporating the hexane by use of a drier. Hexane was quickly evaporated because of its low volume. Our method poses some problems in terms of quantitative performance; however, the procedure is the simplest possible. Since the PCB congeners have different partition coefficients for DMSO and oil and different reactivities for the antibody, the results depend on the composition ratio of the PCB congeners. However, the immunoassay using a multilayer column and DMSO partitioning is applicable to the screening of PCBs in real oil samples.15 If the on-site analysis system is realized, only the oil samples that are judged positive by the screening test will have to be delivered and measured by the standard method in a laboratory or an analytical center.

’ CONCLUSIONS We have developed and demonstrated a DMSO partitioning method that utilizes a microfluidic device containing microrecesses. DMSO confined in the microrecesses was used for extraction of PCBs in oil. Liquidliquid extraction and phase separation were achieved under flow conditions. The enrichment of PCBs in the DMSO extracted from the insulating oil was estimated to be 2.69. The developed microfluidic device can serve as a powerful tool for pretreatment during on-site analysis of PCBs. Since the microfluidic device can reduce the required volume of expensive reagents, eliminate centrifugation processes, and serve as a small analysis device, it can be useful for realizing a low-cost, rapid, on-site analysis system for screening PCBs in oil. ’ AUTHOR INFORMATION Corresponding Author

*(N.O.) Phone: +81-471-82-1181. Fax: +81-471-83-9947. E-mail: [email protected]. (A.A.) Phone: +81-471-82-1181. Fax: +81-471-83-9947. E-mail: [email protected].

’ ACKNOWLEDGMENT We are grateful to Professor Tomokazu Matsue of Tohoku University for help in fabrication of microchips and to Professor Akihide Hibara of the University of Tokyo for help in measuring contact angles and interfacial tensions. ’ REFERENCES (1) Erickson, M. D. Analytical Chemistry of PCBs, 2nd ed.; CRC Press: Boca Raton, FL, 1997.

ARTICLE

(2) Aoki, Y. Environ. Res. 2001, 86, 2–11. (3) Japan, The Ministry of International Trade and Industry. Act on the Evaluation of Chemical Substances and Regulation of Their Manufacture, etc., 1973. http://www.meti.go.jp/english/policy/mono_info_ service/kagaku/chemical_substances/index.html. (4) U.S. EPA. PCB Containing Waste; Disposal Procedures, 41 FR 14133, 1976. http://www.epa.gov/osw/hazard/tsd/pcbs/pubs/laws. htm. (5) Stockholm Convention on Persistent Organic Pollutants, Stockholm, Sweden, 2001. http://chm.pops.int/default.aspx. (6) U.S. EPA. PCB Question and Answer Manual; U.S. Government Printing Office: Washington, DC, 2001. (7) Japan, The Ministry of the Environmental Government. State of PCB Waste in Storage under the PCB Special Measures Law, 2006. http://www.env.go.jp/en/press/2003/1024a.html. (8) Quintana, J. B.; Boonjob, W.; Miro, M.; Cerda, V. Anal. Chem. 2009, 81, 4822–4830. (9) Criado, M. R.; Pereiro, I. R.; Torrijos, R. C. J. Chromatogr. A 2004, 1056, 263–2666. (10) Takada, M.; Toda, H.; Uchida, R. Chemosphere 2001, 43, 455–459. (11) Ohmura, N.; Lackie, S. J.; Saiki, H. Anal. Chem. 2001, 73, 3392–3399. (12) Glass, T. R.; Saiki, H.; Joh, T.; Taemi, Y.; Ohmura, N.; Lackie, S. J. Biosens. Bioelectron. 2004, 20, 397–403. (13) Glass, T. R.; Ohmura, N.; Morita, K.; Sasaki, K.; Saiki, H.; Takagi, Y.; Kataoka, C.; Ando, A. Anal. Chem. 2006, 78, 7240–7247. (14) Ohmura, N.; Glass, T. R.; Sasaki, K.; Joh, T.; Taemi, Y.; Yokobori, N. Bunseki Kagaku 2006, 55, 1–7. (15) Ohmura, N.; Glass, T. R.; Sasaki, K.; Joh, T.; Taemi, Y.; Imanishi, K.; Terakado, S. Bunseki Kagaku 2006, 55, 519–523. (16) Ohmura, N.; Glass, T. R.; Sasaki, K.; Joh, T.; Taemi, Y.; Yokobori, N.; Terakado, S. Bunseki Kagaku 2006, 55, 341–348. (17) Japan, The Ministry of the Environmental Government, 2011. http://www.env.go.jp/press/file_view.php?serial=17471&hou_id=13760 (in Japanese). (18) Reyes, D. R.; Iossifids, D.; Auroux, P.-A.; Manz, A. Anal. Chem. 2002, 74, 2623–2636. (19) Auroux, P.-A.; Iossifids, D.; Reyes, D. R.; Manz, A. Anal. Chem. 2002, 74, 2637–2652. (20) Arora, A.; Simone, G.; Salieb-Beugelaar, G. B.; Kim, J. T.; Manz, A. Anal. Chem. 2010, 82, 4830–4847. (21) Tokeshi, M.; Minagawa, T.; Uchiyama, K.; Hibara, A.; Sato, K.; Hisamoto, H.; Kitamori, T. Anal. Chem. 2002, 74, 1565–1571. (22) Kitamori, T.; Tokeshi, M.; Hibara, A.; Sato, K. Anal. Chem. 2004, 76, 52A–60A. (23) Song, H.; Tice, J. D.; Ismagilov, R. F. Angew. Chem., Int. Ed. 2003, 42, 768–772. (24) Kralj, J. G.; Sahoo, H. R.; Jensen, K. F. Lab Chip 2007, 7, 256–263. (25) Hibara, A.; Kasai, K.; Miyaguchi, H.; Kitamori, T. In Proceedings of Micro Total Analysis Systems 2008; Locascio, L. E., Gaitan, M., Paegel, B. M., Ross, D. J., Vreeland, W. N., Eds.; Chemical and Biological Microsystems Society: San Diego, CA, 2008; pp 13261328. (26) Aota, A.; Nonaka, M.; Hibara, A.; Kitamori, T. Angew. Chem., Int. Ed. 2007, 46, 878–880. (27) Maruyama, T.; Matsushita, H.; Uchida, J.; Kubota, F.; Kamiya, N; Goto, M. Anal. Chem. 2004, 76, 4495–4500.  nidarsic-Plazl, P.; Plazl, I. Lab Chip 2007, 7, 883–889. (28) Z (29) Aota, A.; Hibara, A.; Kitamori, T. Anal. Chem. 2007, 79, 3919– 3924. (30) Aota, A.; Mawatari, K.; Takahashi, S.; Matsumoto, T.; Kanda, K.; Anraku, R.; Hibara, A.; Tokeshi, M.; Kitamori, T. Microchim. Acta 2009, 164, 249–255. (31) Kikutani, Y.; Mawatari, K.; Hibara, A.; Kitamori, T. Microchim. Acta 2009, 164, 241–247. (32) Shen, H.; Fang, Q.; Fang, Z.-L. Lab Chip 2006, 6, 1387– 1389. (33) Chen, H.; Fang, Q.; Yin, X.-F.; Fang, Z.-L. Lab Chip 2005, 5, 719–725. 7839

dx.doi.org/10.1021/ac2015867 |Anal. Chem. 2011, 83, 7834–7840

Analytical Chemistry

ARTICLE

(34) Tokeshi, M.; Minagawa, T.; Kitamori, T. Anal. Chem. 2000, 74, 1711–1714. (35) Glass, T. R.; Saiki, H.; Blake, D. A.; Blake, R. C., II; Lackie, S. J.; Ohmura, N. Anal. Chem. 2004, 76, 767–772. (36) Ohmura, N.; Tsukidate, Y.; Shinozaki, H.; Lackie, S. J.; Saiki, H. Anal. Chem. 2003, 75, 104–110. (37) Blake, R. C., II; Pavlov, A. R.; Blake, D. A. Anal. Biochem. 1996, 272, 123–134. (38) Blake, D. A.; Chakrabarti, P.; Khosraviani, M.; Hatcher, F. M.; Westhoff, C. M.; Goebel, P.; Wylie, D. E.; Blake, R. C., II. J. Biol. Chem. 1996, 271, 27677–27685. (39) Carter, R. M.; Blake, R. C.; Nguyen, T. D.; Bostanian, L. A. Biosens. Bioelectron. 2003, 18, 69–72. (40) Glass, T. R.; Ohmura, N.; Taemi, Y.; Joh, T. Environ. Sci. Technol. 2005, 39, 5005–5009.

7840

dx.doi.org/10.1021/ac2015867 |Anal. Chem. 2011, 83, 7834–7840