Mechanical and Chemical Protection of a Wired Enzyme Oxygen

When implanted in animals, enzyme-containing battery electrodes, biofuel cell ... The mechanical strengthening of the lyotropic liquid crystal by the ...
0 downloads 0 Views 141KB Size
Download PDF
Anal. Chem. 2007, 79, 1173-1180

Mechanical and Chemical Protection of a Wired Enzyme Oxygen Cathode by a Cubic Phase Lyotropic Liquid Crystal Pawel Rowinski, Chan Kang,† Hyosul Shin,† and Adam Heller*

Department of Chemical Engineering, University of Texas, Austin, Texas 78712

When implanted in animals, enzyme-containing battery electrodes, biofuel cell electrodes, and biosensors are often damaged by components of the biological environment. An O2 cathode, superior to the classical platinum cathode, which would be implanted, as part of a caseless physiological pH miniature Zn-O2 battery or as part of a caseless and membraneless miniature glucose-O2 biofuel cell, is rapidly damaged by serum urate at its operating potential. The cathode is made by electrically connecting, or wiring, reaction centers of bilirubin oxidase to carbon with an electron-conducting redox hydrogel. In the physiological pH 7.3 electrolyte battery or biofuel cell, the O2 cathode should operate at, or positive of, 0.3 V (Ag/AgCl), where the urate anion, a common serum component, is electrooxidized. Because an unidentified urate electrooxidation intermediate, formed in the presence of O2, damages the wired bilirubin oxidase electrocatalyst, urate must be excluded from the cathode. Unlike O2, which permeates through both the lipid and the aqueous interconnected networks of cubic-phase lyotropic liquid crystals, urate permeates only through their continuous threedimensional aqueous channel networks. The aqueous channels have well-defined diameters of ∼5 nm in the monoolein/water cubic-phase liquid crystal. Through tailoring the wall charge of the aqueous channels, the anion/cation permeability ratio can be modulated. Thus, doping the monoolein of the monoolein/water liquid crystal with 1,2-dioleoyl-sn-glycero-3-phosphate makes the aqueous channel walls anionic and reduces the urate permeation in the liquid crystal. As a result, the ratio of the urate electrooxidation current to the O2 electroreduction current is reduced from 1:3 to 1:100 for 5-mm O2 cathodes rotating at 1000 rpm. Doping with 1,2-dioleoylsn-glycero-3-phosphate also increases the shear strength of the cubic-phase monoolein/water lyotropic liquid crystal. While the undoped liquid crystal is promptly damaged at the 0.1 N m-2 average shear stress generated by rotating the 5-mm-diameter disk cathode at 1000 rpm in a physiological aqueous solution, the 10 mol % 1,2† Current address: Department of Chemistry, Research Institute of Physic and Chemistry, Chonbuk National University, Chonju 561-756, The Republic of Korea. * To whom correspondence should be addressed. E-mail: heller@ che.utexas.edu.

10.1021/ac061325m CCC: $37.00 Published on Web 01/05/2007

© 2007 American Chemical Society

dioleoyl-sn-glycero-3-phosphate-doped film remains intact. The mechanical strengthening of the lyotropic liquid crystal by the two-tailed 1,2-dioleoyl-sn-glycero-3-phosphate is attributed to cross-linking hydrophobic bonds (i.e., bonds resulting of the increase in entropy upon the freeing of the translation and rotation of multiple water molecules), which is analogous to the strengthening of polymer-based plastic materials by cross-linking through covalent bonds. The wired bilirubin oxidase and the wired laccase O2 cathodes are considerably superior to the platinum O2 cathode, enabling the electroreduction of O2 to water at potentials very near the halfcell potentials of the reversible O2/H2O half-cell.1-3 If implanted in living tissue, as part of a caseless Zn-O2 battery,4 or as part of a caseless and membraneless glucose-O2 biofuel cell,5-7 the cathode would operate at potentials positive of +0.3 V versus Ag/ AgCl, at which serum urate is electrooxidized.8-10 The electrooxidation of urate, in the presence of dissolved O2, destabilizes the wired bilirubin oxidase electrocatalyst of the O2 cathode. While the cathode operates for well over 1 week in a physiological saline buffer solution at 37 °C, it operates in serum only for a few hours. An intermediate product, formed upon the electrooxidation of urate in the presence of O2, damages the cathode’s bilirubin oxidase-containing electrocatalyst.11,12 The life of the electrocatalyst could be extended by an O2-permeable membrane, if the membrane reduced the flux of urate to the cathode. Such a membrane is of essence in transient operation, where the urate electrooxidation current can be disproportionately large, because the wired (1) Mano, N.; Fernandez, J. L.; Kim, Y.; Shin, W.; Bard, A. J.; Heller, A. J. Am. Chem. Soc. 2003, 125, 15290-15291. (2) Soukharev, V.; Mano, N.; Heller, A. J. Am. Chem. Soc. 2004, 126, 83688369. (3) Mano, N.; Soukharev, V.; Heller, A. J. Phys. Chem. B 2006, 110, 1118011187. (4) Heller, A. Anal. Bioanal. Chem. 2006, 385, 469-473. (5) Mano, N.; Mao, F.; Heller, A. J. Am. Chem. Soc. 2003, 125, 6588-6594. (6) Heller, A. Phys. Chem. Chem. Phys. 2004, 6, 209-216. (7) Mano, N.; Mao, F.; Heller, A. ChemBioChem 2004, 5, 1703-1705. (8) Gilmartin, M. A. T.; Hart, J. P.; Birch, B. Analyst (Cambridge, U. K.) 1992, 117, 1299-1303. (9) Ernst, H.; Knoll, M. Anal. Chim. Acta 2001, 449, 129-134. (10) Dutt, J. S. N.; Cardosi, M. F.; Davis, J. Analyst (Cambridge, U. K.) 2003, 128, 811-813. (11) Kang, C.; Shin, H.; Heller, A. Bioelectrochemistry 2006, 68, 22-26. (12) Kang, C.; Shin, H.; Zhang, Y.; Heller, A. Bioelectrochemistry 2004, 65, 83-88.

Analytical Chemistry, Vol. 79, No. 3, February 1, 2007 1173

Chart 1. Structures of Monoolein and 1,2-Dioleoyl-sn-glycero-3-phosphate

bilirubin oxidase electrocatalyst is an anion exchanger in which the concentration of anions, including urate, is higher than in the solution.13 Cubic-phase lyotropic liquid crystals form spontaneously from mixtures of lipids and aqueous solutions. They are clear, isotropic, highly viscous, jellylike substances and have ordered, continuous, fixed-diameter water channels of ∼5 nm. The monoolein/water system forms at high hydration a bicontinuous Pn3m diamondlike space group phase lyotropic liquid crystal. The phase contains two continuous aqueous channel systems. In each, four aqueous channels are connected at knots. Its water-contacting specific surface is exceptionally large, 500-600 m2 g-1.14 The cubic-phase lyotropic liquid crystal is stable in water.15 It can host enzymes16,17 including laccase,18 nonenzyme catalysts,19 and redox mediators of amperometric biosensors.20 Its doping with a cationic lipid, like cationic dioctadecyldimethylammonium chloride or dioctadecylammonium chloride, slows the release of anionic drugs.21,22 Here we show that doping the monoolein/water cubic-phase lyotropic liquid crystal with 1,2-dioleoyl-sn-glycero-3-phosphate,23-25 an anionic lipid having two hydrophobic chains, (Chart 1) greatly increases its mechanical strength and reduces the urate/O2 permeation ratio. In our earlier attempt to reduce the urate-caused damage to the wired bilirubin oxidase electrocatalyst by coating it with Nafion, the coating caused partial phase separation of the (13) Chen, T.; Friedman, K. A.; Lei, I.; Heller, A. Anal. Chem. 2000, 72, 37573763. (14) Engstrom, S.; Norden, T. P.; Nyquist, H. Eur. J. Pharm. Sci. 1999, 8, 243254. (15) Koynova, R.; Caffrey, M. Chem. Phys. Lipids 2002, 115, 107-219. (16) Razumas, V.; Kanapieniene, J.; Nylander, T.; Engstroem, S.; Larsson, K. Anal. Chim. Acta 1994, 289, 155-162. (17) Ropers, M.-H.; Bilewicz, R.; Stebe, M.-J.; Hamidi, A.; Miclo, A.; Rogalska, E. Phys. Chem. Chem. Phys. 2001, 3, 240-245. (18) Rowinski, P.; Bilewicz, R.; Stebe, M.-J.; Rogalska, E. Anal. Chem. 2004, 76, 283-291. (19) Rowinski, P.; Bilewicz, R.; Stebe, M.-J.; Rogalska, E. Anal. Chem. 2002, 74, 1554-1559. (20) Barauskas, J.; Razumas, V.; Talaikyte, Z.; Bulovas, A.; Nylander, T.; Tauraite, D.; Butkus, E. Chem. Phys. Lipids 2003, 123, 87-97. (21) Shah, J. C.; Sadhale, Y.; Chilukuri, D. M. Adv. Drug Delivery Rev. 2001, 47, 229-250. (22) Lynch, M. L.; Ofori-Boateng, A.; Hippe, A.; Kochvar, K.; Spicer, P. T. J. Colloid Interface Sci. 2003, 260, 404-413. (23) Aota-Nakano, Y.; Li, S. J.; Yamazaki, M. Biochim. Biophys. Acta 1999, 1461, 96-102. (24) Li, S. J.; Masum, S. M.; Yamashita, Y.; Tamba, Y.; Yamazaki, M. J. Biol. Phys. 2002, 28, 253-266. (25) Li, S. J.; Yamashita, Y.; Yamazaki, M. Biophys. J. 2001, 81, 983-993.

1174

Analytical Chemistry, Vol. 79, No. 3, February 1, 2007

electrocatalyst, which was made of the electrostatic adduct of polycationic redox polymer wire and polyanionic bilirubin oxidase.12 The Nafion competed with the polyanionic bilirubin oxidase for the polycationic, Os2+/3+ complex rich, enzyme-wire. Although the 1,2-dioleoyl-sn-glycero-3-phosphate-doped monoolein/water lyotropic liquid crystal has anionic water channel walls, the walls do not bind strongly the polycationic wire of the enzyme and do not cause phase separation of the wire and the enzyme. When the aqueous channel walls are made anionic, the permeation of urate is significantly reduced. Because oxygen dissolves in both the lipid and the aqueous solution of the liquid crystal, its permeation is not as severely impeded by the doped cubic-phase lyotropic liquid crystal as that of urate, and the O2 electroreduction to urate electrooxidation current ratio increases 30-fold. EXPERIMENTAL SECTION Chemicals and Materials. Bilirubin oxidase from Trachyderma tsunodae (Tt; 1.89 units/mg, Amano, Lombard, IL), poly(ethylene glycol) (400) diglycidyl ether (PEGDGE) (Polysciences, Warrington, PA), and uric acid (Sigma, St. Louis, MO) were used as received. Monoolein (1-oleoyl-rac-glycerol) was purchased from Sigma, and 1,2-dioleoyl-sn-glycero-3-phosphate monosodium salt was from Avanti Polar Lipids, Inc. (Alabaster, AL). The measurements were performed in a pH 7.3 physiological buffer, containing 20 mM phosphate and 0.15 M NaCl. The urate solution was prepared by dissolving uric acid in 20 mM NaOH and then buffered to pH 7.3 with 20 mM KH2PO4, to produce a 10 mM urate stock solution. The wire used, a copolymer of acrylamide and N-vinylimidazole in which all the imidazole functions were complexed with [Os(4,4′-dichloro-2,2′-bipyridine)2Cl]+/2+, (PAAPVI-[Os(4,4′-dichloro-2,2′-bipyridine)2Cl]+/2+), was synthesized as described.26 Preparation of the Electrodes. The rotating carbon cloth electrodes were prepared as reported.26,27 Smooth 0.196-cm2 vitreous carbon electrodes (Pine Instruments, Raleigh, NC) were polished using 0.05-µm alumina, rinsed, and dried, and then a 0.35mm-thick, 5-mm-diameter carbon cloth disk (Toray TGPH-120, E-TEK, Somerset, NJ) was attached using conducting carbon paste (Structure Probe, Inc. West Chester, PA). After drying for 1 h, the cloth was made hydrophilic by exposure to O2 plasma (10 min, 1 Torr O2). The 10-µL aliquots of a mixture of 20 µL of 6 mg/mL wiring redox polymer solution in water, 3.6 µL of 30 mg/ mL bilirubin oxidase in phosphate buffer, 4.1 µL of 4 mg/mL PEGDGE in water, and 13 µL of the physiological phosphate buffer were pipetted onto the carbon cloth. The electrocatalyst composition and loading were as follows: 0.15 mg/cm2 redox polymer; 0.135 mg/cm2 bilirubin oxidase; and 0.021 mg/cm2 PEGDGE. The wired bilirubin oxidase film formed was cured at room temperature overnight. Preparation of the Cubic-Phase Lyotropic Liquid Crystals. The cubic-phase lyotropic liquid crystals were prepared according to published procedures and phase diagrams.15,25 The monoolein/ water lyotropic liquid crystal consisted of 58 wt % monoolein and 42 wt % physiological buffer solution. The monoolein was weighed in an Eppendorf tube, the buffer solution was added, and the tube (26) Mano, N.; Kim, H.-H.; Zhang, Y.; Heller, A. J. Am. Chem. Soc. 2002, 124, 6480-6486. (27) Barton, S. C.; Kim, H.-H.; Binyamin, G.; Zhang, Y.; Heller, A. J. Am. Chem. Soc. 2001, 123, 5802-5803.

Table 1. Effects of Liquid Crystal Film Coatings on the Voltammetric Characteristics of Smooth Vitreous Carbon Electrodesa [Ru(NH3)6]3+

[Fe(CN)6]3-

coating

Ef

E a - Ec

% of Ibe

Ef

Ea - E c

% of Ibe

none monoolein/water monoolein/water/2% 1,2dioleoyl-sn-glycero-3-phosphate

-159 -158 -192

69 76 68

100.0 8.6 60.0

+199 +197 +196

88 87 102

100.0 5.5 3.0

a Potentials (E) in mV; I refers to the current of the bare electrode; E is the formal potential; E - E is the difference between the anodic and be f a c the cathodic maxima; 10 mV s-1 scan rate.

Figure 1. Cyclic voltammograms of 1 mM [Ru(NH3)6]3+ (top left), 1 mM Fe(CN)63- (top right), and 0.5 mM urate (bottom). (1) Bare electrode; (2) electrode modified with the monoolein/water liquid crystal; (3) electrode modified with monoolein/water/2 mol % 1,2-dioleoyl-sn-glycero-3phosphate liquid crystal. Voltammogram 4 at the bottom is of the bare electrode in buffer without an added redox-active solute. 5-mm diameter, smooth, vitreous carbon electrodes; deoxygenated pH 7.3, 0.15 M NaCl, 0.02 M phosphate buffer; 37.5 °C; 10 mV s-1 scan rate; 150-µm lyotropic liquid crystal film thickness.

was closed and centrifuged several times for 5 min in a 5412 Eppendorf centrifuge at 15000G. The 1,2-dioleoyl-sn-glycero-3phosphate-doped monoolein/water crystals were similarly prepared, but first the 1,2-dioleoyl-sn-glycero-3-phosphate was weighed, and than monoolein was added to produce 0.5-10 mol % 1,2-dioleoyl-sn-glycero-3-phosphate-containing monoolein. To avoid phase separation by fractional crystallization, the lipids were dissolved in a drop of chloroform, which was flash evaporated at ∼60 °C, and then the buffer was added to 30 wt % lipids. The Eppendorf tube was tightly closed, and the mixture was aged for at least 1 day. When placed between crossed polarizers, just before their application, the cubic-phase lyotropic liquid crystals appeared dark and featureless, except at their rim, where water was lost by evaporation and the phase was no longer cubic. The undoped monoolein/water lyotropic liquid crystal was stable for ∼6 months,

Table 2. Urate Electrooxidation on Uncoated and Lyotropic Liquid Crystal-Coated Smooth Vitreous Carbon Electrodesa coating

Ep (mV)

% of Ibe

none monoolein/water monoolein/water (2 mol % 1,2dioleoyl-sn-glycero-3-phosphate)

+341 +362 +388

100.0 16.5 12.5

a I be is the urate electrooxidation current on uncoated smooth vitreous carbon. Ep is the potential of the voltammetric peak of the first scan; 10 mV s-1 scan rate.

but the monoolein/water/10 mol % 1,2-dioleoyl-sn-glycero-3phosphate liquid crystal was stable only for ∼3 weeks, because of slow hydrolysis of the lipid. Analytical Chemistry, Vol. 79, No. 3, February 1, 2007

1175

Figure 2. Dependence of the ratio of the peak urate electrooxidation currents of electrodes coated with increasingly 1,2-dioleoyl-sn-glycero3-phosphate-doped monoolein/water liquid crystal films. Conditions as in Figure 1.

Coating of the Electrodes with Lyotropic Liquid Crystals. The liquid crystal coating was applied by placing a 6-mm-i.d., 0.50mm-high plastic spacer on the 0.35-mm-thick carbon cloth, already cemented to the vitreous carbon rotating electrode. Alternatively, the vitreous carbon rotating electrode itself was coated with the lyotropic liquid crystal, using a 0.15-mm-high spacer. The spacer was filled with the liquid crystal using a pipet tip, and the excess of the liquid crystal was removed with a spatula, before lifting off the spacer. The weight of the liquid crystal on the carbon cloth bound to the carbon disk was 5.5-6.0 mg, and that on the 5-mmdiameter vitreous carbon disk was 4.0-4.5 mg. The liquid crystal film on the carbon cloth did not substantially penetrate the pores, and its thickness equaled the height of the spacer, 0.15 mm. Instrumentation and Cell. The electrodes were rotated with a Pine Instruments (Grove, PA) rotator. The measurements were performed using a model CHI832 potentiostat (CH Instruments, Austin, TX) in a three-electrode cell, maintained at 37.5 °C by an isothermal circulator (Fisher Scientific, Pittsburgh, PA). The cell had a platinum wire counter electrode and an Ag/AgCl (3 M NaCl) reference electrode. The reported potentials are versus this reference. RESULTS AND DISCUSSION Increase of Permeation of the Ru(NH3)63+ Cation and Decrease of the Permeation of the Fe(CN)63- Anion upon 1,2-Dioleoyl-sn-glycero-3-phosphate Doping of the Monoolein/ Water Lyotropic Liquid Crystal. The voltammetric data in Table 1, obtained at a 10 mV s-1 scan rate, summarize the observations for uncoated (bare) smooth vitreous carbon electrodes, for vitreous carbon electrodes coated with undoped monoolein/water, and for vitreous carbon electrodes coated with monoolein/water doped with 2 mol % 1,2-dioleoyl-sn-glycero-3-phosphate. The electrochemically reversible reductions of the cation Ru(NH3)63+ (1 mM) to Ru(NH3)62+ and of the anion Fe(CN)63- (1 mM) to Fe(CN)64- were probed, along with the irreversible oxidation of urate (0.5 mM). (Figure 1) The electrolyte was an aqueous, pH 7.3, 0.15 M NaCl, 20 mM phosphate solution. Application of the monoolein/water lyotropic liquid crystal reduced the voltammetric peak height of the Ru(NH3)63+ wave 11-fold and height of the 1176 Analytical Chemistry, Vol. 79, No. 3, February 1, 2007

[Fe(CN)6]3- wave 18-fold. Doping of the monoolein/water lyotropic liquid crystal with 1,2-dioleoyl-sn-glycero-3-phosphate increased the Ru(NH3)63+ peak height 6-fold but decreased that of [Fe(CN)6]3-1.8-fold. Because the solution-phase concentrations of [Ru(NH3)6]3+ and [Fe(CN)6]3- were the same, and because their diffusion coefficients at 25 °C differ only by 20% (7.17 × 10-6 cm2 s-1 at pH 3.0 in 0.5 M KCl for [Fe(CN)6]3-; 5.48 × 10-6 cm2 s-1 in pH 7.0 phosphate buffer, for [Ru(NH3)6]3+),28 the observed 40% difference in peak currents after application of the undoped lyotropic liquid crystal is attributed in part to the larger diameter of the [Fe(CN)6]3- anion, which reduces its diffusion in the narrow water channels of the lyotropic liquid crystal. The 6-fold increase in the peak height of [Ru(NH3)6]3+upon doping the liquid crystal with 2 mol % 1,2-dioleoyl-sn-glycero-3-phosphate is attributed to the anionic phosphate in the aqueous channel walls. Because the anionic charge is balanced by the charge of solution-phase cations, including [Ru(NH3)6]3+, the concentration of [Ru(NH3)6]3+ in the liquid crystal is higher than in the solution, and the permeability, which is the product of the diffusion coefficient and the concentration, increases for the cation, but not for the anion, correspondingly. The 1.8-fold decrease in the peak height for [Fe(CN)6]3upon 1,2-dioleoyl-sn-glycero-3-phosphate doping is attributed to its electrostatic repulsion from the anionic walls of the aqueous channels, which narrows the effective channel diameter for anion diffusion. Coating the smooth vitreous carbon electrode with monoolein/ water/2 mol % 1,2-dioleoyl-sn-glycero-3-phosphate shifted the formal potential of [Ru(NH3)6]2+/3+ by +33 mV but did not change the peak separation. The upshift is attributed to greater ion pairing of [Ru(NH3)6]3+ than of [Ru(NH3)6]2+ with wall anions. While the peak height of [Ru(NH3)6]3+ is as much as 60% of that of the bare electrode for the electrode is coated with the anionic lipid-doped liquid crystal, the peak current of [Fe(CN)6]3- is only 3.0% of that of the bare electrode after it is coated. Unlike the formal potential of [Ru(NH3)6]2+/3+, which is upshifted by the preferred electrostatic interaction of the oxidized member of the redox couple, the formal potential of [Fe(CN)6]3-/4- is not changed by 2 mol % 1,2dioleoyl-sn-glycero-3-phosphate doping. In summary, making the walls of the aqueous channels of the liquid crystals anionic increases the permeation rate of cations and slightly reduces the permeation rate of anions, increasing 12-fold, from 1.6 to 20, the [Ru(NH3)6]3+/[Fe(CN)6]3- electroreduction current ratio. Reduction of the Urate Electrooxidation Current by Undoped and by 1,2-Dioleoyl-sn-glycero-3-phosphate-Doped Monoolein/Water Lyotropic Liquid Crystal Coatings in Stagnant Solutions. Figure 1 (bottom) and Table 2 show the effect of coating the smooth vitreous carbon electrodes with monoolein/ water and with monoolein/water/2 mol % 1,2-dioleoyl-sn-glycero3-phosphate on the electrooxidation of 0.5 mM urate. In the absence of the liquid crystal coatings, the urate electrooxidation wave peaks in the first voltammogram at +0.341 V and then shifts positive as the electrode is increasingly fouled by the electrooxidative polymerization product of urate.29 The monoolein/water coating shifted the peak potential by + 21 mV and the monoolein/ water/2 mol % 1,2-dioleoyl-sn-glycero-3-phosphate coating by (28) Baur, J. E.; Wightman, R. M. J. Electroanal. Chem. Interfacial Electrochem. 1991, 305, 73-81. (29) Binyamin, G.; Chen, T.; Heller, A. J. Electroanal. Chem. 2001, 500, 604611.

Figure 3. Polarizations of O2 cathodes without the wired bilirubin oxidase electrocatalyst and their Levich plots measured at -0.9 V (Ag/ AgCl). Bare and smooth vitreous carbon (top); coated with monoolein/water/10 mol % 1,2-dioleoyl-sn-glycero-3-phosphate liquid crystal (center); Koutecky-Levich plots of the bare (open rhombs) and modified (solid rhombs) electrodes poised at -0.9 V (Ag/AgCl). (1) 100, (2) 500, and (3) 1000 rpm. Scan rate, 5 mV s-1; 1 atm O2; other conditions as in Figure 1.

+ 47 mV. The shifts are attributed to reduced urate surface concentration on the carbon electrodes. Because the charge of urate is only 1/3 of that of Fe(CN6)3-, it is less electrostatically repelled from the anionic channel walls, allowing the urate to diffuse through a larger fraction of the cross-sectional channel area. Thus 2 mol % 1,2-dioleoyl-sn-glycero-3-phosphate doping decreased the permeation of urate only by ∼25%, while it decreased the permeation of Fe(CN6)3- by 40%. Higher 1,2dioleoyl-sn-glycero-3-phosphate doping resulted in greater wall charge, and the increased electrostatic repulsion confined the urate to a lesser cross-sectional area, reducing its permeation (Figure 2). Greater doping (>0.5 mol % 1,2-dioleoyl-sn-glycero-3phosphate) changes the Pn3m diamondlike space group cubic-

phase liquid crystal to the also cubic, but more hydrated, Im3m phase.25 This is not the case, however, in 0.15 M NaCl, where the space group of the monoolein/water/10 mol % 1,2-dioleoylsn-glycero-3-phosphate remains cubic Pn3m. At 0.15 M NaCl concentration, the urate electrooxidation current is only 9.3% of that on the bare electrode after the electrode is coated with monoolein/water/10 mol % 1,2-dioleoyl-sn-glycero-3-phosphate. The voltammetric peak for monoolein/water/10 mol % 1,2-dioleoylsn-glycero-3-phosphate-coated electrode is +0.35 V, downshifted by -0.04 V from that of monoolein/water/2 mol % 1,2-dioleoylsn-glycero-3-phosphate, suggesting that the wall charge reduces the dissociation of uric acid, i.e., increasing its effective pKa, and facilitating its oxidation. Analytical Chemistry, Vol. 79, No. 3, February 1, 2007

1177

Figure 4. Polarizations (left) of the 0.5 mM urate-electrooxidizing rotating electrodes and their Levich plots (right) measured with the electrodes poised at + 0.7 V (Ag/AgCl). Bare electrode (top); monoolein/water/10 mol % 1,2-dioleoyl-sn-glycero-3-phosphate liquid crystal-coated electrode (bottom). 1000 rpm; scan rate, 5 mV s-1; other conditions as in Figure 1.

Change in the Oxygen Flux Upon Applying the Monoolein/Water 10 mol % 1,2-Dioleoyl-sn-glycero-3-phosphate Film to Rotating Smooth Vitreous Carbon Electrodes. Figure 3 (top) shows the potential dependence of the O2 electroreduction current as a function of the angular velocity for the bare vitreous carbon electrode. In the Levich plot (eq 1) (Figure 3, top right), the mass transport limited current increases, as expected, linearly with the square root of the angular velocity

i ) 0.620nFAD2/3ω1/2ν-1/6Cs

(1)

between 100 and 1200 rpm (n is the number of electrons exchanged, F is Faraday’s constant, D is the diffusion coefficient of the electroactive species, ν is the kinematic viscosity, Cs is the concentration of O2, and ω is the angular velocity). Figure 3 (center) shows the potential dependence of the O2 electroreduction current as a function of the angular velocity for the electrode coated with monoolein/water/10 mol % 1,2-dioleoyl-sn-glycero-3phosphate. Although the current still increased with ω1/2, the increase was no longer linear, especially at the higher rotation rates (Figure 3, center, right). Such behavior was observed by Gough and Leypoldt30,31 in rotating polymeric membrane-covered (30) Gough, D. A.; Leypoldt, J. K. Anal. Chem. 1980, 52, 1126-1130. (31) Gough, D. A.; Leypoldt, J. K. Anal. Chem. 1979, 51, 439-444.

1178

Analytical Chemistry, Vol. 79, No. 3, February 1, 2007

Figure 5. Cyclic voltammograms of the wired bilirubin oxidasecoated carbon cloth O2 cathode. (1) without coating; (2) coated with a 0.15-mm thick monoolein/water/10 mol % 1,2-dioleoyl-sn-glycero3-phosphate film. 1 atm O2; scan rate, 5 mV s-1; other conditions as in Figure 1.

disk electrodes, for which the current was id (eq 2), where i is

id ) i

[

]

1 1 + Ps/Pm

(2)

the current in the absence of the membrane, i.e., the current from

Figure 6. Stabilization of the wired bilirubin oxidase-coated carbon cloth O2 cathodes by lyotropic liquid crystal coating. No urate added (left); urate added to 0.5 mM concentration at 2500 s (right). (1) Without coating; (2) coated with a 0.15-mm-thick monoolein/water/10 mol % 1,2dioleoyl-sn-glycero-3-phosphate lyotropic film. 0.15 V (Ag/AgCl); 1 atm O2; other conditions as in Figure 1.

the Levich eq 1; Pm and Ps are the respective permeabilities of the membrane and the solution; Dm is the diffusion coefficient of

Pm ) RDm/δm

(3)

Ps ) D/δd

(4)

the electroactive probe in the membrane; R is a partition coefficient, i.e., the concentration in its membrane divided by the concentration in the solution; δm is the thickness of the membrane; D is the diffusion coefficient of the probe in the solution, and δd is the thickness of the diffusion layer. For mixed membrane and solution transport control, eq 2 predicts linear dependence of the current on the rotation rate at slow rotation rates and high membrane permeability. For low membrane permeability, the current is membrane transport controlled and is independent of rotation rate. Equation 2 can be rewritten32 as eq 5, where ilim is a steady-state limiting current of the membrane

δm 1 1 ) + ilim i nFARDmCs

(5)

covered rotating disk electrode at a sufficiently large overpotential for the reaction not to be limited by the electron-transfer rate on the electrode. According to eq 5, the plots of the reciprocal of the currents versus the reciprocal of the angular velocity, i.e., the Levich-Koutecky plots, have identical slopes for the modified and the bare electrodes, but the intercepts for the membrane-coated electrodes are now defined by the permeabilities of the membranes. Indeed, as seen in Figure 3 (bottom), both plots, for the bare and for the modified electrodes, have slopes differing only by 10%. While application of the lyotropic liquid crystal reduces the effective electrode area on which O2 is electroreduced, the crystal is permeable to O2 and the current still increases, though no longer linearly, with the square root of the angular velocity. (32) Leddy, J.; Bard, A. J. J. Electroanal. Chem. Interfacial Electrochem. 1983, 153, 223-242.

Table 3. Stabilities of the Current Densities (j) of Bare and Monoolein/Water (10 mol % 1,2-Dioleoyl-sn-glycero-3-phosphate)-coated Carbon Cloth Wired Bilirubin Oxidase O2 Cathodes

PBS, O2 PBS, urate, O2

jcoated/jbare at 2450 s

j7200s/j2500s, bare cathode

j7200s /j2500s, liquid crystalcoated cathode

0.44 ( 0.11 0.43 ( 0.04

0.79 ( 0.12 0.72 ( 0.06

0.96 ( 0.03 0.96 ( 0.04

Increasing the O2 Electroreduction: Urate Electrooxidation Current Ratio by Rotating the Monoolein/Water/10 mol % 1,2-Dioleoyl-sn-glycero-3-phosphate-Coated Electrode. The urate electrooxidation current on the bare electrode poised at +0.7 V (Figure 4 top, left) is solution transport controlled, increasing linearly with ω1/2 (Figure 4, top, right). In contrast, the urate oxidation current at the monoolein/water/10 mol % 1,2dioleoyl-sn-glycero-3-phosphate overcoated electrode does not change when the rotation rate is increased from 100 to 1200 rpm (Figure 4, bottom left). The urate transport is controlled now by the monoolein/water/10 mol % 1,2-dioleoyl-sn-glycero-3-phosphate film. Because the O2 transport increases with the rotation rate, but the urate transport does not, the ratio of the O2 electroreduction/urate electrooxidation currents increases with the rate of rotation, reducing the fraction of the current lost to electrooxidation of urate and to O2 scavenging by the intermediate urate electrooxidation product. For the bare carbon electrode in the stagnant solution, the urate electrooxidation current was ∼30% of the O2 electroreduction current. It dropped to 2% upon its coating with monoolein/water/10 mol % 1,2-dioleoyl-sn-glycero3-phosphate, and then dropped further, to 1%, when the electrode was rotated at 1000 rpm. The permeation of urate was drastically reduced, yet the O2 electroreduction current for the rotating electrode coated with the 150-µm-thick monoolein/water/10 mol % 1,2-dioleoyl-sn-glycero-3-phosphate film was 20% of that for the bare electrode. Unlike the urate, the O2 permeated through both the lipid substructure and the aqueous channels of the cubic-phase lyotropic liquid crystal. Analytical Chemistry, Vol. 79, No. 3, February 1, 2007

1179

Improving the Shear Strengths of the Lyotropic Liquid Crystals by Their Doping. The monoolein/water/10 mol % 1,2dioleoyl-sn-glycero-3-phosphate film on the 5-mm carbon disk was mechanically stable for at least 4 h under the high, 0.1 N m-2 average shear stress,33 produced by rotation at 1000 rpm in the physiological buffer solution, while the undoped and the less doped monoolein/water/2 mol % 1,2-dioleoyl-sn-glycero-3-phosphate films were unstable. Both the undoped monoolein/water and the monoolein/water/2 mol % 1,2-dioleoyl-sn-glycero-3phosphate films rapidly lost their cubic structure, the currents dropping to nil because of discontinuity of the water channels in the altered phases. The massive improvement in the mechanical strength is attributed to the hydrophobic bonds between the long and aligned alkyl chains of the lipids in the lyotropic liquid crystals. Lipids with two alkyl chains (Chart 1) form two sets of hydrophobic bonds, acting in a manner analogous to cross-linkers of polymers, well known to mechanically strengthen plastics. Increasing the 1,2-dioleoyl-sn-glycero-3-phosphate doping is analogous to increasing the density of cross-links in polymer-based plastics. Stabilization of the Rotating Wired Bilirubin Oxidase O2 Carbon Cloth Cathode by Coating with Monoolein/Water/ 10 mol % 1,2-Dioleoyl-sn-glycero-3-phosphate. Oxygenelectroreducing wired copper enzyme (laccase)-coated carbon cloth electrodes were earlier made and characterized by Barton et al.27 Figure 5 shows voltammograms of wired bilirubin oxidase carbon cloth O2 cathodes without and with the monoolein/water/ 10 mol % 1,2-dioleoyl-sn-glycero-3-phosphate film. The stability of their currents is seen in Figure 6, left, in the absence of urate and Figure 6, right, with urate added at t ) 2500 s. The results (mean values ( standard deviation from three experiments) are summarized in Table 3. In the 4700-s period following t ) 2500, 79 ( 12% of the current of the uncoated electrode rotating at 1000 rpm persisted in the absence of urate, but 96 ( 3% persisted when the electrode was coated with monoolein/water/10 mol % 1,2dioleoyl-sn-glycero-3-phosphate. The uncoated, shear-stressed,

rotating cathode lost its wired bilirubin oxidase electrocatalyst, while the mechanically strong liquid crystal coated electrode did not. When urate was added, and the protective lyotropic liquid crystal was not applied, the remaining current was 72 ( 6%. However, even in the presence of urate, as much as 96 ( 4% of the O2 electroreduction current persisted when the wired bilirubin oxidase cathode was protected by the monoolein/water/10 mol % 1,2-dioleoyl-sn-glycero-3-phosphate film, which greatly decreased the permeation of urate but decreased much less the permeation of O2.

(33) Binyamin, G.; Heller, A. J. Electrochem. Soc. 1999, 146, 2965-2967.

AC061325M

1180

Analytical Chemistry, Vol. 79, No. 3, February 1, 2007

CONCLUSION The 1,2-dioleoyl-sn-glycero-3-phosphate dopant, with two hydrophobic carbon chains, mechanically strengthens the one-chain, monoolein-water cubic-phase lyotropic liquid crystal, just as crosslinking with a bifunctional cross-linker strengthens a polymer. O2 permeates rapidly in the monoolein/water (10 mol % dopant) lyotropic liquid crystal, and half of the O2 flux to the electrode is retained when the crystal is 0.15 mm thick (Figure 6). Urate does not permeate as rapidly in the undoped liquid crystal, and when the crystal’s water channel walls are made anionic, the urate flux is further reduced. The doped liquid crystal stabilizes the wired bilirubin oxidase O2 cathode, preventing mechanical stripping of its electrocatalyst when sheared, and reducing its urate electrooxidation-associated chemical damage. The observed mechanical strengthening by the two-tailed dopant explains why all of the surfactants in lipid bilayer cell membranes are two-tailed: The cell membranes, as thin as they are, must withstand the huge tensile/compressive stresses in the breathing lung, or in the pumping heart, where they are also subjected to the shear stress of the rapidly flowing blood. ACKNOWLEDGMENT Support of the Office of Naval Research (award N000140210144) and the Welch Foundation is gratefully acknowledged. Received for review July 20, 2006. Accepted November 13, 2006.