Improved Planar Amperometric Nitric Oxide Sensor Based on

Anal. Chem. 2004, 76, 536-544

Improved Planar Amperometric Nitric Oxide Sensor Based on Platinized Platinum Anode. 1. Experimental Results and Theory When Applied for Monitoring NO Release from Diazeniumdiolate-Doped Polymeric Films Youngmi Lee, Bong Kyun Oh, and Mark E. Meyerhoff*

Department of Chemistry, University of Michigan, 930 North University Avenue, Ann Arbor, Michigan 48109-1055

An improved miniature amperometric nitric oxide sensor design with a planar sensing tip (ranging from 150 µm to 2 mm in diameter) is reported. The sensor is fabricated using a platinized platinum anode and a Ag/AgCl cathode housed behind a microporous poly(tetrafluoroethylene) (PTFE; Gore-tex) gas-permeable membrane. Platinization of the working platinum electrode surface dramatically improves the analytical performance of the sensor by providing ∼10-fold higher sensitivity (0.8-1.3 pA/nM), ∼10-fold lower detection limit (e1 nM), and extended (at least 3-fold) stability (>3 d) compared to sensors prepared with bare Pt electrodes. These improvements in performance arise from increasing the kinetics and lowering the required potential for the 3-electron oxidation of NO to nitrate, relative to that observed using a nonplatinized working electrode. The outer porous PTFE membrane provides complete selectivity for NO over nitrite ions (up to 10 mM nitrite). The new sensor is applied for surface measurements of NO released from diazeniumdiolate-loaded silicone rubber films (SR-DACA-6/N2O2). The effects of sensor size (for sensor dimensions of 0.15-, 1-, and 2-mm o.d.) and the distance of the sensor from the surface of the NO-emitting polymer film are investigated via experiments as well as theoretical calculations. A significant analyte trapping effect is demonstrated, the degree of which depends on the sensor size and its distance from the surface. It is further demonstrated that surface NO concentrations for fresh SR-DACA-6/N2O2 loaded films are also influenced by the polymer film thickness, with thicker films generating higher surface concentrations of NO. Since nitric oxide was first discovered as the elusive endothelial-derived relaxing factor in 1987,1 there has been a significant research effort to examine NO’s many other important physiological functions. Indeed, in addition to regulating blood pressure via its vasodilating activity, NO is also a potent inhibitor of platelet adhesion and activation,2 is a mediator in a wide range of * Corresponding author: (Fax) 734-647-4865. (E-mail) [email protected]. (1) Ignarro, L. J.; Bugga, G. M.; Wood, K. S. Byrns, R. E.; Chaudhuri, G. Proc. Natl. Acad. Sci. U.S.A. 1987, 84, 9265-9269.

536 Analytical Chemistry, Vol. 76, No. 3, February 1, 2004

antimicrobial as well as antitumor activities,3 and also serves as an important neurotransmitter.4 Because of these critical physiological functions, the development of new methods to quantitate NO at biologically relevant concentrations has also been an area of considerable research interest. Further, profiling NO concentrations locally, near its source (generally from the oxidation of L-arginine by nitric oxide synthase within cells), such as the layer of endothelial cells that line all healthy blood vessels, presents a particularly difficult analytical challenge. Quantitative measurement of NO in biological fluids is problematic because of its low concentration and relatively short lifetime,5 due to autoxidation by reactions with endogenous oxygen or hemoglobin to form nitrite (NO2-) or nitrate (NO3-). Thus, most methods of NO measurements are indirect ones based on detection of NO oxidation products with spectroscopic approaches (e.g., Griess assay for nitrite, detection of nitrate and nitrite with reductase enzymes, and detection of methemoglobin after NO reaction with oxyhemoglobin).6 More direct methods include chemiluminescence (after reaction with ozone), Raman spectroscopy, EPR, mass spectrometry, and amperometric/voltammetric techniques.6,7 To monitor local surface concentrations of NO near its source, electrochemical methods have significant advantages over the other techniques. Indeed, miniature and microelectrochemical gas sensors (amperometric/voltammetric) can be positioned in proximity to the NO source and provide a means to estimate the local surface concentration of NO.6 Electrochemical sensors for use in this and other applications must exhibit very high selectivity for NO, good sensitivity, fast response times, and long-term calibration stability and possess a small enough size to enable placement in proximity to the NO source. In addition, if assessing concentration (2) Radomski, M. W.; Palmer, R. M. J.; Moncada, S. Proc. Natl. Acad. Sci. U.S.A. 1990, 87, 5193-5197. (3) (a) Nathan, C. F.; Hibbs, J. B. Curr. Opin. Immunol. 1991, 3, 65-70. (b) Langrehr, J. M.; Hoffman, R. A.; Lancaster, J. R.; Simmons, R. L. Transplantation 1993, 55, 1205-1212. (4) Ohta, A.; Takagi, H.; Matsui, T.; Hamai, J.; Iida, S.; Esumi, H. Neurosci. Lett. 1993, 158, 33-35. (5) Thomas, DD.; Liu, X.; Kantrow, S. P.; Lancaster, J. R. Jr. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 355-360. (6) Methods in Nitric Oxide Research; Feelish, M., Stamler, J., Eds.; John Wiley: Chichester, U.K., 1996. (7) Archer, S. FASEB J. 1993, 7, 349-360. 10.1021/ac035064h CCC: $27.50

© 2004 American Chemical Society Published on Web 12/30/2003

profiles of NO at different distances from the source, it is essential that the sensing tip of the device have a planar configuration so that NO levels at precise distances away from the source can be determined with good spatial resolution. Electrochemical detection of NO is generally based on either of two strategies. In one case, direct electrochemical oxidation of NO (to nitrate) at a solid electrode operated in an amperometric mode is employed to generate the analytical signal. In these type of sensors, platinum, gold, glassy carbon, or carbon fiber/disk working electrode surfaces are often employed, but these need to be modified with some organized layers to enhance the selectivity for NO (to prevent response from other easily oxidized species, including ascorbate, nitrite, etc.). Indeed, a variety of membranes (e.g., cellulose acetate,8 Nafion,9 chloroprene,10 and polycarbazole11) have been used to modify the surfaces of such working electrodes via electropolymerization of appropriate monomers or classical dip coating procedures. To further improve selectivity and sensitivity, multilayers of different membrane materials12-15 have also been applied. Electrochemical NO sensors can also be based on electrocatalytic oxidation of NO at polymeric metalloporphyrin film-modified electrodes. Malinski and Taha first reported the catalytic action of nickel(II) porphyrins for the oxidation of NO using carbon fiber electrodes modified with an electropolymerized nickel(II) porphyrin film and Nafion.12 The exact interaction between NO and the nickel(II) porphyrin is still not clear, and others have suggested that the nickel(II) porphyrin film does not actually play a role in the oxidation of NO, but serves merely as an organic layer modifying the electrode surface.13 Nevertheless, NO sensors based on the nickel(II) porphyrin/Nafion are widely used. Newer multicharged nickel(II) porphyrins,14 nickel(II) phthalocyanine,15 and nickel(II) Schiff base16 structures have also been examined as analogous alternatives to prepare NO sensors.17 Platinum is well known not only as an inert metal working electrode material but also as a catalyst for many chemical reactions. Platinization of working Pt electrodes has been used previously to devise chemical and biological amperometric sensors with improved response characteristics.18-20 Surprisingly, there have been no reports to date regarding the application of platinized Pt working electrodes for fabrication of amperometric NO sensors. Herein, we report an amperometric NO sensor with a planar sensing tip based on a platinized Pt working electrode behind an outer microporous PTFE gas-permeable membrane. It will be shown that the use of a platinized working electrode greatly (8) (a) Meulemans, A. Neurosci. Lett. 1994, 171, 89-93. (b) Pallini, M.; Curulli, A.; Amine, A.; Palleschi, G. Electroanalysis 1998, 10, 1010-1016. (c) Cserey, A.; Gratzel, M. Anal. Chem. 2001, 73, 3965-3974. (9) Bedioui, F.; Tre´vin, S.; Devynck, J. J. Electroanal. Chem. 1994, 377, 295298. (10) Shibuki, K. Neurosci. Res. 1990, 9, 69-76. (11) Prakash, R.; Srivastava, R. C.; Seth, P. K. Polym. Bull. 2001, 46, 487-490. (12) Malinski, T.; Taha, Z. Nature 1992, 358, 676-678. (13) Tre´vin, S.; Bedioui, F.; Devynck, J. Talanta 1996, 43, 303-311. (14) Tre´vin, S.; Bedioui, F.; Devynck, J. J. Electroanal. Chem. 1996, 408, 261265. (15) Bedioui, F.; Tre´vin, S.; Devynck, J.; Lantoine, F.; Brunet, A.; Devynck, M. A. Biosens. Bioelectron. 1997, 12, 205-212. (16) Tre´vin, S.; Bedioui, F.; Gomez Villegas, M. G.; Bied-Charreton, C. J. Mater. Chem. 1997, 7, 923-928. (17) (a) Bedioui, F.; Tre´vin, S.; Devynck, J. Electroanalysis 1996, 8, 1085-1091. (b) Pointe´, M.; Bedioui, F. Analysis 2000, 28, 465-469. (c) Bedioui, F.; Villeneuve, N. Electroanalysis 2003, 15, 5-18.

Figure 1. (a) Schematic diagram of the new amperometric gas sensor; (b) an inverted (to make distance values positive) blow-up of the sensing tip region in cylindrical coordinates when the device is used to monitor NO levels near the surface of NO release polymers or other materials.

enhances the analytical performance of the device. Further, the effect of the sensor’s overall size (determined by diameter of glass insulating sheath) on NO concentration measurements near the surface of novel polymeric films that continuously release NO with fluxes comparable to normal endothelial cells is also investigated. Experimental results for surface NO levels are compared to theoretical predictions based on computer simulations of NO diffusion and reactivity. EXPERIMENTAL SECTION Materials. Pt wire (76-µm o.d.) and Ag wire (127-µm o.d.) were obtained from Aldrich (Milwaukee, WI) and Medwire (Mt. Vernon, NY), respectively. Borosilicate glass capillaries (with 1and 2-mm o.d.) were products of World Precision Instrumentation Inc. (Sarasota, FL). Nitric oxide, argon, and nitrogen gases were obtained from Cryogenic Gases (Detroit, MI). Hydrochloric acid, sodium chloride, phosphate-buffered saline (PBS), sodium nitrite, and platinizing solution (3% chloroplatinic acid in water) were purchased from Aldrich, Mallinckrodt Laboratory Chemicals (Phillipsburg, NJ), Sigma (St. Louis, MO), Riedel-de Hae¨n (Seelze, Germany), and LabChem Inc. (Pittsburgh, PA), respectively. All solutions were prepared with 18 MΩ cm-1 deionized water with reagent grade compounds as obtained without further purification. Micorporous PTFE membranes (50% porosity; pore size 0.2 µm) (available from W. L. Gore and Associates, Elkton, MD) were obtained as a gift from Dr. Mark Arnold, University of Iowa. Preparation of NO Sensors. Amperometric planar NO sensors were prepared as depicted in Figure 1a. The sensor was composed of a glass-sealed platinized Pt working electrode and a coiled Ag wire (127-µm o.d.), which are housed behind the outer Analytical Chemistry, Vol. 76, No. 3, February 1, 2004

537

PTFE gas-permeable membrane. The Ag wire spontaneously forms a Ag/AgCl reference/counter electrode in the 30 mM NaCl and 0.3 mM HCl internal solution of the sensor. Working electrodes were prepared as follows: The Pt disk electrodes (with 76-µm o.d.) shielded in glass (with 1- and 2-mm o.d.) were prepared by a method previously reported.21 For a smaller electrode diameter, the glass insulator sheath surrounding the Pt disk was conically diminished down to 150-µm o.d. by polishing. The detailed procedure has been described elsewhere.22 The bare Pt disk electrodes were then platinized in a platinizing solution (3% chloroplatinic acid in water) by cyclic voltammetry (from +0.6 to -0.35 V vs Ag/AgCl) at a scan rate of 20 mV/s using an electrochemical impedance analyzer, model 6310, from EG&G Instruments (Princeton, NJ). The outer sleeve of the NO sensors was made with a micropipet Redi-tip (volume 1-200 µL) (Fisherbrand, Pittsburgh, PA), which was cut at ∼1.5 cm from the distal tip to make a large orifice. This enlarged opening was covered with the PTFE membrane and fixed in place with an O-ring. Finally, the platinized Pt disk electrode with a coiled silver wire as the reference electrode around its glass wall was inserted into the outer sleeve filled with an internal solution (aqueous 30 mM NaCl and 0.3 mM HCl solution as recommended by Shibuki to optimize kinetics of NO oxidation at a platinum working electrode of a NO gas sensor10) and then pushed gently toward the PTFE membrane. Pressure was exerted to cause significant stretching of the membrane, as observed by a protrusion of the inner electrode from the end plane of the outer pipet sleeve. Therefore, the outer diameter of the sensor tips was determined by the working electrode’s insulating glass o.d., regardless of the size of the orifice of the outer container. The length of the protruding inner electrode (∼1 mm) and the thickness of the stretched PTFE membrane (∼30 µm) were controlled under an optical microscope, and reproducibility to within (20% was achieved (N > 30). Surface NO Measurements. Nitric oxide sensors were calibrated before and after NO measurements using standard NO solutions prepared by bubbling NO gas into PBS solutions deaerated with Ar gas purging (assuming a concentration of NO in these saturated solutions of 1.9 mM25). For surface NO measurements, the NO-releasing polymeric films were adhered on the bottom of a glass beaker containing PBS buffer using inert vacuum grease. The NO sensors were positioned over NOreleasing polymer films immersed in the buffer at varying distances (10-70 µm; distance from outer surface of membrane to NO-releasing polymer) using micromanipulators (World Preci(18) Maclay, G. J.; Keyvani, D.; Lee, S. B. In Proceedings of the Second International Symposium on Microstructures and Microfabricated Systems; Chicago, October 1995; Vol. 95-27 (Electrochemical Society Proceedings). (19) (a) Vidal, J. C.; Garcia-Ruiz, E.; Castillo, J. R. J. Pharm. Biomed. Anal. 2000, 24, 51-63. (b) Vidal, J. C.; Garcia-Ruiz, E.; Castillo, J. R. Anal. Sci. 2002, 18, 537-542. (20) Li, Q. S.; Ye, B. C.; Liu, B. X.; Zhong, J. J. Biosens. Bioelectron. 1999, 14, 327-334. (21) Wightman, R. M.; Wipt, D. O. In Electroanalytical Chemistry; Bard, A. J., Ed.; Marcel Dekker: New York, 1988; Vol. 15. (22) Bard, A. J.; Fan, F.-R. F.; Kwak, J.; Lev. O. Anal. Chem. 1989, 61, 132138. (23) Zhang, H.; Annich, G. M.; Miskulin, J.; Osterholzer, K.; Merz, S. I.; Bartlett, R. H.; Meyerhoff, M. E. Biomaterials 2002, 23, 1485-1494. (24) (a) Wink, D. A.; Darbyshire, J. F.; Nims, R. W. Chem. Res. Toxicol. 1993, 6, 23-27. (b) Kharitonov, V. G.; Sundquist, A. R.; Sharma, V. S. J. Biol. Chem. 1994, 269, 5881-5883. (25) Handbook of Chemistry and Physics; Lide, D. R., Ed.; CRC Press, Inc.: Boca Raton, FL, 1990-1991.

538

Analytical Chemistry, Vol. 76, No. 3, February 1, 2004

sion Instrumentation Inc., Sarasota, FL). To measure the surface concentrations of NO released from the polymers, a potential adequate to oxidize NO to nitrate (0.75 V for the platinized working electrode; 0.9 V for the nonplatinized platinum vs Ag/AgCl electrode) (see Results and Discussion) was applied to the inner working electode and the resulting current was measured as a function of time using a high-sensitivity microsensor module (Diamond Electro-Tech Inc., Ann Arbor, MI). The measured current was converted to the concentration of NO using prior calibration curves recorded immediately before surface measurements were started. The NO-releasing polymer film used in these experiments was a diazeniumdiolated silicone rubber material (SRDACA-6/N2O2), prepared as described elsewhere.23 Theory of NO Measurements Near NO Source as a Function of Sensor Diameter and Distance. The time-dependent current as a function of sensor size as well as distance between the sensor and a NO-releasing surface was examined using a numerical solution of the appropriate diffusion equations. A diagram of the distal region of the planar NO sensor in cylindrical coordinates is shown in Figure 1b. Calculations were based on Fick’s second law. Indeed, only diffusion was considered as the means of mass transport of NO to the anode of the electrode and from the surface of the emitting polymer film. For the long time periods required to achieve steady-state currents using the model (>50 s), not only diffusion but also unintentional convection could contribute to mass transport of NO. However, a closer look at this specific system would suggest that contributions from convection should be minimal even at long time periods, given the very confined (i.e., sandwiched) solution phase that exists between the sensor and the surface of the NOemitting polymer film, along with the specific series of parallel interfaces that are in close proximity to each other in the final gas sensor experiment being modeled (see Figure 1b). For example, the sum of the Nernst layer thicknesses (even if a low level of external convection were present in each phase), associated with each of the four parallel interfaces (i.e., platinum working electrode surface, inner surface of PTFE membrane, outer surface of PTFE membrane, and surface of the NO-releasing polymer film) would be a very large fraction (if not total) of the distance between the surface of the working Pt electrode and the surface of the NO-emitting polymer film. Add to this, any solution phase that fills the innermost and outermost layers (likely 5-10 µm on each side) of the 0.2-µm-diameter pores of the stretched PTFE gaspermeable membrane (30 µm thick), where clearly only diffusion can occur, and one can conclude that when the outer surface of the sensor is within 10-20 µm of the NO-emitting polymer film, the vast majority of liquid phases between that surface and the Pt electrode surface are within the Nernst layer thicknesses and diffusion should dominate even at extended time periods. Unfortunately, for the purposes of modeling, it is not easy to describe the diffusion pattern of NO within the PTFE gaspermeable membrane. We consider this barrier as a three-layer structure (one centered air pocket layer sandwiched with two porous aqueous layers). Nitric oxide diffusion within the porous inner and outer aqueous layers is likely slower than the in free solution phase owing to the tortuous path of the porous polymeric structure. On the other hand, NO certainly diffuses through the air pocket region formed in the center layer of the polymer film

much faster than in a free solution phase but slower than in a free air phase due to the impedance of the solid porous polymer membrane network. Therefore, as a first approximation, the combined mass transport barrier for the PTFE membrane is considered as an average of these processes and is taken as the barrier that would occur in a pure stagnant solution phase of the same thickness. For the calculations, the following reaction at a sensor surface was considered: kf

NO + 2H2O - 3e- {\ } 4H+ + NO3k b

(1)

where kf and kb are the heterogeneous electron-transfer rate constants. The following autoxidation reaction of NO in the surrounding solution phase was also considered: kaq

4NO + O2 + 2H2O 98 4HNO2

(2a)

where kaq is the reaction constant, 6 × 106 M-2 s-1.24 The NO reaction rate of the above reaction is second order in NO concentration and first order in O2 concentration and can be expressed by the following rate equation:24

-d[NO]/dt ) kaq[NO]2[O2]

(2b)

The concentration of NO is denoted as c(r,z,t) and the diffusion equation in cylindrical coordinates is expressed as

(

)

∂ 2c ∂2c ∂c ∂c )D 2+ + 2 ∂t r∂r ∂z ∂r

(3)

where r and z are the coordinates in directions parallel and normal to the electrode surface, respectively, D is the diffusion coefficient of NO, 2.7 × 10-5 cm2 s-1 at 23 °C,25 and c is the concentration of NO. Mixed boundary conditions were applied via the following equations:

∂c(0,z)/∂r ) 0 for 0 e z < d

(4)

∂c(b,z)/∂r ) 0 for dd e z < 0

(5)

∂c(r,0)/∂z ) 0 for a e r < b

(6)

∂c(rs,z)/∂r ) 0 for dd e z < d

(7)

∂c(r,dd)/∂z ) 0 for b e r < rs

(8)

c(r,0) ) 0 for 0 e r < a

(9)

∂c(r,d)/∂z ∂t ) (JNO/a2) for 0 e r < rs

(10)

where JNO is the NO flux from NO-releasing polymer films. The current for the entire surface of a sensing electrode can be obtained as a function of sensor-sample surface distance:

i(d,t) ) nFD

∫

a

0

2πr(∂c/∂z)z ) 0 dr

(11)

Figure 2. Linear sweep voltammograms obtained using (a) bare Pt disk electrodes and (b) platinized Pt disk electrodes. Solution contained 90 µM NO or 180 µM NO2- with 30 mM NaCl and 0.3 mM HCl. Diameter of working electrode, 0.5 mm (note: large diameter working electrode was used because the EG&G electrochemical analyzer employed for the linear sweep voltammetry was not sensitive enough to detect pA levels without noise); scan rate, 20 mV/s.

For the simulations, [O2] . [NO] and a constant flux of NO (JNO) from NO-releasing polymer film was assumed. The typical NO flux pattern of SR-DACA-6/N2O2 was a function of time; however, the film generates a relatively constant flux within 1.53-h time frame after soaking in PBS with