Potentiometric Sensing of Chemical Warfare Agents: Surface

The proliferation and use of chemical and biological warfare agents have ..... of Chemical Warfare Agents, An Approach to the Environmental Monitoring...
0 downloads 0 Views 116KB Size
Anal. Chem. 2004, 76, 2689-2693

Potentiometric Sensing of Chemical Warfare Agents: Surface Imprinted Polymer Integrated with an Indium Tin Oxide Electrode Yanxiu Zhou, Bin Yu, Eric Shiu, and Kalle Levon*

Polymer Research Institute, Polytechnic University, Six Metrotech Center, Brooklyn, New York 11201

Rapid and specific recognition of methylphosphonic acid (MPA), the degradation product of nerve agents sarin, soman, VX, etc., was achieved with potentiometric measurements using a chemical sensor fabricated by a surface imprinting technique coupled with a nanoscale transducer, indium tin oxide (ITO). An octadecylsiloxane thin layer was covalently bound to the ITO-coated glass surface in the presence of MPA. After extraction of MPA, potentiometric measurements showed selective detection of MPA. The selectivity of the sensor has been tested on other alkylphosphonic acids, such as ethylphosphonic acid and propylphosphonic acid, as well as tert-butylphosphonic acid. The viability of the sensor in the presence of other chemical analogues, such as organophosphorus pesticides and herbicides, was investigated. The proliferation and use of chemical and biological warfare agents have become relevant in our society. Lethal compounds such as sarin (isopropyl methylphosphonofluoridate), soman (pinacolyl methylphosphonofluoridate), and VX (O-ethyl-S-2-diisopropylaminoethyl methylphosphonothioate) are highly toxic nerve agents, lethal at low dosages,1,2 and have been used in the recent past.3-8 As the nerve agents are hydrolyzed in the environment, the detection of the degradation product, such as methylphosphonic acid, has been usually performed for proof of the use of nerve agents.9-12 There have been many innovations * Corresponding author. E-mail: [email protected]. Tel: 718 260 3339. Fax: 718 260 3125. (1) Trapp, R. SIPRI Chemical & Biological Warfare Studies. 3. The Detoxification and Natural Degradation of Chemical Warfare Agents; Taylor & Francis: Philadelphia PA, 1985; p 104. (2) Compton, J. A. F. Military Chemical and Biological Agents Chemical and Toxicological Properties; The Telford Press: Caldwell, NJ, 1987; p 458. (3) Tu, A. T. Natural and Selected Synthetic Toxins; American Chemical Society: Washington, DC, 2000; Chapter 20. (4) Black, R. M.; Clarke, R. J.; Read, R. W.; Reid, M. T. J. J. Chromatogr.,, A 1994, 662, 301-321. (5) Report S-16433; United Nations Security Council: New York, 1984. (6) Robinson, J. P. P. SIPRI Yearbook, World Armaments and Disarmament; Taylor & Francis: London, 1985; Chapter 6. (7) Lundin, S. J. SIPRI Yearbook, World Armaments and Disarmament; Talyor & Francis: London, 1989; Chapter 4. (8) Institute of Medicine. Sarin. In Gulf War and Health; National Academy of Sciences Press: Washington, DC, 2000; pp 135-164. (9) Tu, A. T. J. Mass Spectrom. Soc. Jpn. 1996, 44, 293. (10) Trace Analysis of Chemical Warfare Agents, An Approach to the Environmental Monitoring of Nerve Agents; Hirsja¨rvi, P., Miettinen, J. K., Paasivirta, J., Kanolahti, E., Eds.; Ministry of Foreign Affairs of Finland: Helsinki, 1981; pp 27, 28, 37-39, 59-64, 72-79, 90-99. 10.1021/ac035072y CCC: $27.50 Published on Web 04/14/2004

© 2004 American Chemical Society

for the identification of methylphosphonic acid (MPA), including high-performance liquid chromatography (HPLC), ion chromatography, HPLC/mass spectrometry, gas chromatography, and nuclear magnetic resonance spectroscopy.3-8 Although these instruments do offer quantitative analysis, they are not optimal for rapid detection as all are plagued by at least one limitation such as the following: expensive, require sophisticated, often extensive analysis procedures, or nonportability.3,13-16 The methods that meet the needs of real-time analysis, such as fiber-optic sensors,17 surface acoustic wave devices,18 or microbial biosensors,19 often lack selectivity and do not have an optimal alarm ratio. The selectivity can be improved with approaches such as a molecular imprinting technique, which possesses specific affinities for guest molecules.20,21 Some limitations to the method may result from the fact that the transduction mechanism is still separated from the binding in contrast to the natural process where both recognition and transduction are integrated, for instance, in the function of an ion channel in a natural membrane. Therefore, the mimicking of nature by combining recognition and transduction by imprinting the templates on the surface of nanoscale transducer is a target of intensified research. The formation of monolayers of bonded organosilane reagents on SnO2, TiO2, and glassy carbon electrodes under anhydrous reaction conditions was discovered by Murray et al.22 Sagiv23-32 (11) Identification of Degradation Products of Potential Organophosphorus Warfare Agents, An Approach for the Standardization of Techniques and Reference Data; Hirsja¨rvi, P., Miettinen, J. K., Paasivirta, J., Eds.; Ministry of Foreign Affairs of Finland: Helsinki, 1980; pp 3-10, 18-30 and appendices. (12) Ashley, J. A.; Lin, C.-H.; Wirsching, P.; Janda, K. D. Angew. Chem., Int. Ed. Engl. 1999, 38, 1793-1795. (13) Mesilaakso, M. T. Environ. Sci., Technol. 1997, 31, 518-522. (14) Abu-Qare, W.; Abou-Donia, B. Chromatographia 2001, 53, 251-255. (15) Nassar, A.-E. F.; Lucas, S. V.; Thomas. S. A. Anal. Lett. 1999, 32, 10231035. (16) Rohrbaugh, D. K.; Sarver, E. W. J. Chromatogr., A 1998, 809, 141-150. (17) Jenkins, A.; Uy, O. M.; Murray, G. M. Anal. Chem. 1999, 71, 373-378. (18) Nieuwenhuizen, M. S.; Harteveld, J. L. N. Sens. Actuators, B 1997, 40, 167173. (19) Mulchandani, A.; Chen, W.; Mulchandani, P.; Wang, J. Biosens. Bioelectron. 2001, 16, 225-230. (20) Molecularly imprinted polymers, man-made mimics of antibodies and their application in analytical chemistry; Sellergren, B., Ed.; Elsevier: New York, 2001. (21) Molecular and Ionic Recognition with Imprinted Polymers; Bartsch, R. A., Maeda, M., Eds.; American Chemical Society: Washington, DC, 1998. (22) Untereker, D. F.; Lennox, J. C.; Wier, L. M.; Moses, P. R.; Murray, R. W. J. Electroanal. Chem. 1977, 81, 309-318. (23) Polymeropoulos, E. E.; Sagiv, J. J. Chem. Phys. 1978, 69 (5), 1836-1848. (24) Neizer, L.; Sagiv, J. J. Am. Chem. Soc. 1983, 105, 674-676. (25) Gun, J.; Iscovici, R.; Sagiv, J. J. Colloid Interface Sci. 1984, 101, 201-213.

Analytical Chemistry, Vol. 76, No. 10, May 15, 2004 2689

Scheme 1. Formation of Selective MPA Cavities on ITO-Coated Glass Plates

was able to form similar oleophobic monomolecular films but containing mixed monolayers of more than one component on similar polar surfaces from anhydrous organic solutions. The covalent polysiloxane monolayer-surface binding with intralayer cross-linking was shown to result in an unusual mechanical, chemical, and electrical stability.24 The covalently bonded silane monolayers were shown to be perfectly stable under conditions that cause a major deterioration of the structure of fatty acid films.25 Sagiv et al. also studied the thermal behavior of the monolayers prepared by covalent, ionic, and physical bonds and found that, with the exception of the covalently bound octadecylsiloxane (ODS), all other monolayers underwent a large melting transition around 110 °C.26 The heating caused only slight disorientation of the chains (by FT-IR investigation) in ODS, and no sharp melting point was observed. The significant structural reversibility upon recooling was related to the stability of the ODS monolayers as the immobilization of the headgroups by the covalent intralayer can be expected to prevent melting of such films. The fact that the mixed monolayers contained both physisorbed and chemisorbed components made the removal of the physisorbed components possible. Sagiv30-32 proposed that the resulting skeleton monolayers have holes of molecular dimensions that may be used for free adsorption sites. To investigate this opportunity, the mixed monolayers were formed in the presence of surfactant dyes, which were only physically adsorbed on the surface in contrast to the monolayers that were covalently linked.31 The dye molecules were washed away, leaving holes in the polymerized silane network. This experiment can be considered as the first successful surface molecular imprinting experiment. Tabushi et al. applied this concept to imprint alkanes in the above-mentioned mixed monolayers.33,34 In their studies, they used n-hexadecane as the template molecules and coimplanted the (26) Cohen, S. R.; Naaman, R.; Sagiv, J. J. Phys. Chem. 1986, 90, 3054-3056. (27) Maoz, R.; Sagiv, J. J. Colloid Interface Sci. 1984, 100, 465-496. (28) Iscovici, R.; Sagiv, J. Thin Solid Films 1983, 99, 235-241. (29) Iscovici, R.; Sagiv, J. Thin Solid Films 1983, 100, 67-76. (30) Sagiv, J. J. Am. Chem. Soc. 1980, 102, 92-98. (31) Sagiv, J. Isr. J. Chem. 1979, 18, 339-345. (32) Sagiv, J. Isr. J. Chem. 1979, 18, 346-353. (33) Tabushi, I.; Kurihara, K.; Naka, K.; Yamamura, K.; Hatakeyama, H. Tetrahedron Lett. 1987, 28, 4299-4302.

2690

Analytical Chemistry, Vol. 76, No. 10, May 15, 2004

templates with the ODS molecules on a SnO2 electrode surface. The formed monolayer was able to sense molecules with alkyl tails, indicating the presence of the space induced by the template molecules. This application confirmed the recognition components but did not take advantage of the active surface properties. The detection of the affinity binding by guest recognition was exemplified by Mosbach et al. using ellipsometry for the binding measurements.35 Previously, we have developed highly selective, sensitive, stable, and fast response chemical sensors for chiral amino acids 36,37 by integration of surface imprinting30-37 with the transduction properties of an indium tin oxide (ITO) electrode. Here, we report chemical sensors for MPA based on the surface imprinting method30-37 coupled with potentiometric detection. EXPERIMENTAL SECTION Materials. Chloroform and carbon tetrachloride were distilled over CaH2. Other chemicals were used without further purification. Thionazin, malathion, and dibutyl chlorendate were obtained as a neat liquid standard from Radian (Radian International, Austin, TX). Neat liquid standards of phosdrin and dichlorvos and solid standards of methyl parathion and dimethoate were purchased from Supelco (Supelco Chromatography Products, Bellefonte, PA). Unless otherwise specified, all solutions and suspensions were prepared from water purified using a Millipore system (resistivity 18.2 MΩ‚cm). The ITO-coated glass plates used as indicator electrodes were obtained from Delta Technologies, Ltd. (Stillwater, MN). Sensor Construction. ITO-coated glass plates were pretreated by a method described by Sagiv.30 MPA (template) and octadecyltrichlorosilane (OTS, C18H37SiCl3; silylating agent) were coadsorbed on the polar solid surface of the ITO glass plate (effective surface area ∼1 × 4 cm2) from a CHCl3/CCl4 solution (2:3 v/v) at 0 °C for a period of time. Then, the electrode was rinsed with CHCl3 for 30 times using 1-mL volumes of chloroform. An ODS-modified electrode without the MPA cavity was also prepared as a control. (34) Yamamura, K.; Hatakeyama, H.; Naka, K.; Tabushi, I.; Kurihara, K. Chem. Commun. 1988, 79-81. (35) Andersson, L. I.; Mandenius, C. F.; Mosbach, K. Tetrahedron Lett. 1988, 29, 5437-5440. (36) Zhou, Y.; Yu, B.; Levon, K. Chem. Mater. 2003, 15, 2774-2779. (37) Zhou, Y.; Nagaoka, T. Chem. Sens. 1998, 14 (Suppl. B), 101-104.

Figure 1. Potentiometric responses of MPA on the MPA ODS/ITO electrodes with (s) and without template (- - -). [MPA](CHCl3/CCl4) ) 2.5 ×10-2 M; [OTS] ) 8.0 × 10-4 M. Adsorption time 3 min.

Figure 2. Dependence of the potential response of the MPA sensor on the concentration of MPA in CHCl3/CCl4. The other conditions are as in Figure 1.

Experimental Setup and Measurement. All measurements were made in 50 mL of 0.1 M phosphate-buffered saline (PBS; pH 7.2) unless specifically stated, in a 100-mL working volume electrochemical cell, equipped with a magnetic stirrer. The twoelectrode system consisted of an Ag/AgCl (saturated KCl) reference electrode and the MPA ODS/ITO sensor or the control as the working electrode. The potentials of the MPA ODS/ITO indicator electrodes were measured against the Ag/AgCl reference electrode with an Orion 920A potentiometer. Thereafter, the potential response of the sensor was defined as the difference between the electrode potential with and without MPA in solution, i.e., ∆E ) E1 - E0, where E0 and E1 are the electrode potentials before and after MPA addition, respectively. The potential responses of the sensor as a function of pH were determined using the same two-electrode system described above but with one more pH electrode to record the change of pH of the bulk solution. RESULTS AND DISCUSSION The surface imprinting method30-37 was applied by modifying the surface of an ITO-coated glass plate with OTS to fabricate a sensor for MPA (Scheme 1). The ITO plate was soaked in a MPA suspension dispersed in CHCl3/CCl4 containing OTS. Thus, OTS and MPA were coabsorbed on the ITO surface simultaneously and the polysiloxane thin film was further generated in the presence of MPA. Afterward, the MPA was washed out from the polymeric film by solvent extraction,30,36 with the goal of obtaining a skeleton membrane, which would contain a space of the molecular dimensions of MPA as the free adsorption sites for subsequent recognition. The resulted sensor demonstrated a higher complementary steric and functional adsorption ability to the template molecule than that of the untemplated surface. As shown in Figure 1, MPA selectively bound to the MPA-imprinted film (Figure 1, curve -) compared to that of unimprinted one (Figure 1, curve - - -), which indicated that the MPA molecules were originally embodied into the ODS film, removed by solvent washing, and subsequently created recognition sites. It appeared that the conditions chosen for the surface polymerization were extremely important to ensure a selective sensor.

Figure 3. Influence of the coadsorption time of ITO in the deposition solution on the potentiometric response of the MPA sensor. The other conditions are as in Figure 1.

First, the optimal concentration of MPA that was most effectively incorporated into the ODS film was between 1.95 × 10-2 and 3.46 × 10-2 M (Figure 2). The OTS monomer used in the formation of the imprinted polymer is typically reached after a few minutes, as shown in Figure 3. Interestingly, when the coadsorption time was longer than 5 min, the sensor produced smaller potential responses for the MPA. This behavior could be explained in terms of two competitive adsorption processes, the (chemical) adsorption (polymerization) of OTS on the ITO surface and the insertion of MPA into the ODS layer. During the initial phase of polymerization (