Anal. Chem. 2001, 73, 5328-5333
Reducing the Thrombogenicity of Ion-Selective Electrode Membranes through the Use of a Silicone-Modified Segmented Polyurethane Maria J. Berrocal,§ Ibrahim H. A. Badr,† Dayong Gao,‡ and Leonidas G. Bachas*,§
Department of Chemistry and Center of Membrane Sciences and Department of Mechanical Engineering and Center for Biomedical Engineering, University of Kentucky, Lexington, Kentucky 40506-0055
The susceptibility of segmented polyurethanes (SPUs) to in vivo oxidative cleavage and hydrolysis constitutes a drawback in the use of these materials in the fabrication of implantable devices. The introduction of poly(dimethylsiloxane) (PDMS) groups into the polymer main chain has been previously reported to enhance the stability of SPUs. Herein, we evaluated the use of BioSpan-S, a siliconemodified SPU, in the design of membranes for cationselective electrodes. The resulting electrodes exhibited good potentiometric response with all of the tested ionophores (valinomycin, sodium ionophore X, and nonactin). The obtained selectivity coefficients meet the selectivity requirements for the determination of sodium and potassium in blood. Moreover, as reflected by SEM studies, membranes prepared with BioSpan-S showed less adhesion of platelets than membranes prepared with conventional poly(vinyl chloride) (PVC). These results lead to the conclusion that BioSpan-S would be an appropriate candidate for the fabrication of implantable ion-selective electrodes. Ion-selective electrodes (ISEs) have been used extensively in the determination of several analytes in physiological media.1-3 In most of the early literature, ISEs were employed in ex vivo and bedside applications, but more recent work reflects the increasing interest in in-vivo applications, which are more challenging.4 As with any other device that comes into contact with biological environments, implantable sensors interact with proteins and cells. Therefore, the material of the sensor surface plays a critical role and ultimately determines its biocompatibility.5 If * To whom correspondence should be addressed. Fax: (859) 323-1069. Phone: (859) 257-6350. E-mail:
[email protected]. † On leave from the Department of Chemistry, Faculty of Science, Ain-Shams University, Cairo, Egypt. ‡ Department of Mechanical Engineering and Center for Biomedical Engineering. § Department of Chemistry and Center of Membrane Sciences. (1) Bakker, E.; Bu ¨ hlmann, P.; Pretsch, E. Chem. Rev. 1997, 97, 3083-3132. (2) Bu ¨ hlmann, P.; Pretsch, E.; Bakker, E. Chem. Rev. 1998, 98, 1593-1667. (3) Hahn, C. E. W. Analyst 1998, 6, 57R-86R. (4) Bakker, E.; Diamond, D.; Lewenstam, A.; Pretsch, E. Anal. Chim. Acta 1999, 393, 11-18. (5) Loh, I.-H.; Sheu, M.-S.; Fisher, A. B. In Desk Reference of Functional Polymers: Synthesis and Applications; Arshady, R., Ed.; American Chemical Society: Washington, DC, 1997; pp 657-675.
5328 Analytical Chemistry, Vol. 73, No. 21, November 1, 2001
inserted into the blood stream, the material should be bloodcompatible (i.e., hemocompatible) as well, in order not to initiate the blood coagulation cascade.6 Hence, biocompatibility is one of the most critical properties of the material to be used in the fabrication of sensors for in vivo biomedical applications. Poly(vinyl chloride), or PVC, is the conventional material used in the preparation of ISE polymeric membranes. However, the polymer itself is not fully biocompatible,7 and it exhibits poor adhesion to the supports used in the fabrication of microelectrodes.8 Different approaches have been proposed to overcome the aforementioned problems. Modification of the ISE membrane surface is a strategy that has been explored to enhance their biocompatibility. Polymeric coatings, such as grafted poly(ethylene oxide), or hydrogel coatings, such as poly(2-hydroxyethyl methacrylate), have been used to improve the biocompatibility of ISE membranes.9,10 Modification of the surface of the membranes with covalently attached biomolecules (e.g., anticoagulants), has also been employed to obtain sensors with better blood compatibility.11 Recently, membranes have been proposed that continuously release active species, such as nitric oxide (NO).12,13 These membranes are doped with different NO-generating diazeniumdiolate compounds. Such compounds, when in contact with water, allow the release of NO over extended periods of time, decreasing the adhesion and activation of platelets onto the membranes, both in vitro and in vivo. New materials, such as substituted PVC,14,15 cellulose triacetate,16,17 polyurethanes,18-22 polymethacrylates,23-26 and silicone rubber,27-30 are among the polymers that have been explored as (6) Anderson, J. M.; Kottake-Merchant, K. CRC Crit. Rev. Biocompat. 1985, 1, 111-203. (7) Simon, M. A.; Kusi, R. P. J. Biomed. Mater. Res. 1996, 30, 313-320. (8) Cha, G. S.; Liu, D.; Meyerhoff, M. E.; Cantor, H. C.; Midgley, A. R.; Goldberg, H. D.; Brown, R. B. Anal. Chem. 1991, 63, 1666-1672. (9) Espadas-Torre, C.; Meyerhoff, M. E. Anal. Chem. 1995, 67, 3108-3114. (10) Cosofret, V. V.; Erdosy, M.; Johnson, T. A.; Bellinger, D. A.; Buck, R. P.; Ash, R. B.; Neuman, M. R. Anal. Chim. Acta 1995, 314, 1-11. (11) Brooks, K. A.; Allen, J. R.; Feldhoff, P. W.; Bachas, L. G. Anal. Chem. 1996, 68, 1439-1443. (12) Mowery, K. A.; Schoenfisch, M. H.; Baliga, N.; Wahr, J. A.; Meyerhoff, M. E. Electroanalysis 1999, 11, 681-686. (13) Schoenfisch, M. H.; Mowery, K. A.; Baliga, N.; Wahr, J. A.; Meyerhoff, M. E. Anal. Chem. 2000, 72, 1119-1126. (14) Cosofret, V. V.; Lindner, E.; Buck, R. P.; Kusy, R. P.; Whitley, J. Q. Electroanalysis 1993, 5, 725-730. (15) Cosofret, V. V.; Buck, R. P.; Erdosy, M. Anal. Chem. 1994, 66, 3592-3599. (16) Cha, G. S.; Meyerhoff, M. E. Talanta 1989, 36, 271-278. 10.1021/ac010375i CCC: $20.00
© 2001 American Chemical Society Published on Web 09/28/2001
Figure 1. Structure of BioSpan-S.
alternatives to PVC in the fabrication of ISE membranes. These polymers present several advantages over PVC. Some of these materials need less, if any, plasticizer25-30 and show better adhesion to the substrates used in microfabrication.8 But perhaps what is most important is that some of them, like polyurethane and silicone rubber, are used in the fabrication of biomedical devices because of their enhanced biocompatibility;5 however, a limitation to the use of polyurethanes is their susceptibility to in vivo oxidative cleavage and hydrolysis.31 Herein, we investigate the design of ISE membranes with a new polymeric material, BioSpan-S (Figure 1), a silicone-modified segmented polyurethane (or SPU).32 This material shows the same mechanical properties as the corresponding unmodified polyurethane, BioSpan, which is used in the fabrication of prostheses and implants because of its good mechanical properties and biocompatibility.32 The poly(dimethylsiloxane) (PDMS) groups of BioSpan-S are introduced in the main chain of the polymer during its synthesis and, therefore, are present in the bulk as well as at the surface. BioSpan-S exhibits higher in vivo stability than does the corresponding unmodified SPU, a property that was attributed to a possible protective role of the PDMS end groups.31 The (17) Cha, M. J.; Shin, J. H.; Oh, B. K.; Kim, C. Y.; Cha, G. S.; Shin, D. S.; Kim, B. Anal. Chim. Acta 1995, 315, 311-319. (18) Liu, D.; Meyerhoff, M. E.; Goldberg, H. D.; Brown, R. B. Anal. Chim. Acta 1993, 274, 37-46. (19) Lindner, E.; Cosofret, V. V.; Buck, R. P.; Johnson, T. A.; Ash, R. B.; Neuman, M. R.; Kao, W. Y. J.; Anderson, J. M. Electroanalysis 1995, 7, 864-870. (20) Cosofret, V. V.; Erdosy, M.; Raleigh, J. S.; Johnson, T. A.; Neuman, M. R.; Buck, R. P. Talanta 1996, 43, 143-151. (21) Yun, S. Y.; Hong, Y. K.; Oh, B. K.; Cha, G. S.; Nam, H. Anal. Chem. 1997, 69, 868-873. (22) Levitchev, S.; Smirnova, A.; Bratov, A.; Vlasov, Y. Fresenius’ J. Anal. Chem. 1998, 361, 252-254. (23) Bratov, A.; Abramova, N.; Mun ˜oz, J.; Domı´nguez, C.; Alegret, S.; Bartrolı´, J. Anal. Chem. 1995, 67, 3589-3595. (24) Ambrose, T. M.; Meyerhoff, M. E. Electroanalysis 1996, 8, 1095-1100. (25) Heng, L. Y.; Hall, E. A. H. Anal. Chim. Acta 1996, 324, 47-56. (26) Heng, L. Y.; Hall, E. A. H. Anal. Chem. 2000, 72, 42-51. (27) Shin, J. H.; Sakong, D. S.; Ham, H.; Cha, S. Anal. Chem. 1996, 68, 221225. (28) Ho ¨gg, G.; Lutze, O.; Camman, K. Anal. Chim. Acta 1996, 335, 103-109. (29) Tsujimura, Y.; Sunagawa, T.; Yokohama, M.; Kimura, K. Analyst 1996, 121, 1705-1709. (30) Poplawski, M. E.; Brown, R. B.; Rho, K. L.; Yun, S. Y.; Lee, H. J.; Cha, G. S.; Paeng, K. J. Anal. Chim. Acta 1997, 355, 249-257. (31) Mathur, A. B.; Collier, T. O.; Kao, W. J.; Wiggins, M.; Schubert, M. A.; Hiltner, A.; Anderson, J. M. J. Biomed. Mater. Res. 1997, 36, 246-257. (32) Ward, R. S.; White, K. A. Proceedings from the 8th CIMTEC-Forum of New Materials, Topical Symposium VIII, Materials in Clinical Applications; Florence, Italy, 1994.
presence of the PDMS groups on the surface of the polymer also enhances the biocompatibility and nonthrombogenicity of BioSpanS, possibly as a result of the resultant decrease in surface free energy.31,32 This is consistent with data obtained with other silicone-modified polyurethanes, which have been previously proposed as biocompatible materials.33,34 Finally, as demonstrated in this work, membrane electrodes based on BioSpan-S exhibit good potentiometric response and selectivity; therefore, these characteristics make this material a well-suited candidate for fabricating implantable ISEs. EXPERIMENTAL SECTION Reagents. The segmented polyurethane investigated in this work, BioSpan-S, was obtained from the Polymer Technology Group (Berkeley, CA). Polyurethane Tecoflex SG-80A was donated by Thermedics (Woburn, MA). Valinomycin, 4-tert-butylcalix[4]arene-tetraacetic acid tetraethyl ester (Sodium Ionophore X), nonactin, bis(2-ethylhexyl) sebacate (DOS), bis(2-ethylhexyl) phthalate (DOP), poly(vinyl chloride) (PVC), potassium tetrakis[3,5-bis(trifluoromethyl)phenyl]borate (KTFPB), and tetrahydrofuran (THF) were purchased from Fluka (Ronkonkoma, NY). Tris(hydroxymethyl)aminomethane (Tris) was obtained from Research Organics (Cleveland, OH). N,N-Dimethylacetamide and hexamethyldisilazane (HDMS) were purchased from Aldrich (Milwaukee, WI). Absolute ethanol was obtained from AAPER Alcohol and Chemical (Shelbyville, KY). Phosphate buffered saline, Hanks’ balanced salt solution, and dimethylarsinic acid (cacodylic acid, sodium salt) were obtained from Sigma (St. Louis, MO). The other reagents were of the maximum purity available. All of the aqueous solutions were prepared with deionized distilled water with a Milli-Q Water Purification System from Millipore (Bedford, MA). Apparatus. Membrane potentials were monitored using an in-house custom-built six-channel high-impedance amplifier with unity gain coupled to an analog-to-digital converter (G. W. Instruments; Somerville, MA) connected to a MacIntosh computer using SuperScope v. 1.2 (G. W. Instruments) software. The reference electrode was an Orion Ag/AgCl double-junction electrode. (33) Hergenrother, R. W.; Yu, X. H.; Cooper, S. L. Biomaterials 1994, 15, 635640. (34) Ishihara, K. In Biomedical Applications of Polymeric Materials; Tsuruta, T., Hayashi, T., Kataoka, K., Ishihara, K., Kimura, Y., Eds.; CRC Press: Boca Raton, FL, 1993: Chapter 3.
Analytical Chemistry, Vol. 73, No. 21, November 1, 2001
5329
Preparation of Membranes and Potentiometric Setup. Membrane compositions used in this study are given in the Results and Discussion section. In general, 2 mg of the ionophore (corresponding to 1 wt %), different amounts of the lipophilic salt KTFPB, plasticizer, and BioSpan-S were dissolved in 2 mL of N,Ndimethylacetamide. This cocktail was poured into a 22-mmdiameter glass ring on a poly(tetrafluoroethylene) (PTFE) plate, and the membrane was formed after evaporation of part of the solvent overnight at room temperature, followed by placement in an oven at 35 °C for 2 days to complete the evaporation process. The use of a PTFE casting plate was necessary, because the BioSpan-S membranes adhered strongly to glass. This sequence of evaporation steps was required in order to avoid the formation of air bubbles within the membrane. Smaller disks were cut from this membrane and were placed at the tip of a Philips IS-561 electrode body (Glasblaserei Mo¨ller, Zurich, Switzerland). The electrodes were then conditioned overnight in a 1.0 × 10-2 M solution of the primary ion. Potentiometric measurements were obtained by using the following cell assembly: Ag/AgCl|KCl (saturated)||buffer||sample| membrane|0.0100 M of the primary ion in buffer|Ag/AgCl. The change in the potential of the cell, ∆E, was recorded for every addition of aliquots of standard solutions to 50.0 mL of buffer. The buffer that was used throughout the experiments was 0.0100 M Tris-HCl, pH 7.2. The potential changes that are reported reflect averages of at least three electrodes. RESULTS AND DISCUSSION Biocompatibility of the outer surface of a sensor is of high importance if the sensor is to be used in vivo for the determination of clinically relevant analytes. In the case of membrane-based ISEs, the use of a biocompatible polymer in the formulation of the sensing membrane may fulfill this requirement. In this work, a silicone-modified polyurethane, BioSpan-S, was evaluated as matrix for ISE membranes. BioSpan-S has been reported to show reduced thrombogenicity.32 Moreover, this material is more resistant to in vivo oxidative cleavage and hydrolysis than unmodified polyurethanes.31 To function properly, cation-selective electrodes based on neutral ionophores require the presence of anionic sites within the polymer matrix.1 In contrast to PVC, which contains endogenous negative sites,35 the presence of cationic impurities has been reported in the commercial polyurethane Tecoflex.36 Thus, it was of critical importance to determine whether membranes prepared with BioSpan-S also contain ionic sites. Several methods have been proposed in the past to determine the number of sites within PVC membranes, including the use of impedance measurements,35 radiotracers,37 and atomic absorption experiments.38 Nonaqueous titrations and spectral analysis of chromoionophore-containing membranes were used to determine the acidic groups in decylmethacrylate membranes.24 The change in selectivity as a function of site concentration within an ISE membrane has also been (35) Horvai, G.; Gra´f, E.; To´th, K.; Buck, R. P. Anal. Chem. 1986, 58, 27352740. (36) Na¨gele, M.; Pretsch, E. Mikrochim. Acta 1995, 121, 269-279. (37) Thoma, A. P.; Viviani-Nauer, A.; Arvanitis, S.; Morf, W. E.; Simon, W. Anal. Chem. 1977, 49, 1567-1572. (38) Lindner, E.; Gra´f, E.; Niegreisz, Z.; To´th, K.; Pungor, E.; Buck, R. P. Anal. Chem. 1988, 60, 295-301.
5330 Analytical Chemistry, Vol. 73, No. 21, November 1, 2001
Figure 2. Potassium chloride response of membranes having 50: 50 mass ratio of DOP:BioSpan-S with 0 (circles), 1 (triangles), 1.5 (diamonds), and 2 wt % KTFPB (squares). No ionophore was added to the membranes.
employed previously to estimate the concentration of ionic sites within the membrane.36 Electrodes with blank BioSpan-S membranes, that is, containing no additives or ionophore, showed anionic response when in contact with KCl solutions, implying the presence of cationic impurities. To determine the amount of these impurities, we adopted a method previously reported by Ambrose and Meyerhoff to evaluate the sites present in methacrylate-based membranes.24 Membranes with 1, 1.5, and 2 wt % of the lipophilic additive KTFPB and a 50:50 mass ratio of DOS:BioSpan-S (but no ionophore) were prepared. KTFPB provides anionic sites that would counterbalance any cationic impurities in BioSpan-S. The potentiometric response of the membranes was evaluated using KCl standard solutions. As seen in Figure 2, there is a progressive conversion of the electrode response from anionic to cationic, with increasing amounts of KTFPB. This reinforces the hypothesis of the presence of cationic impurities in the membrane (probably left from the synthesis process). The concentration of KTFPB within the membranes where the response of the electrode turns from anionic to cationic can be used to estimate the amount of cationic impurities present in the membranes. This amount was estimated to be between 20 and 30 mmol/kg SPU. To purify the material, a sample of BioSpan-S, which was dissolved in N,N-dimethylacetamide, was precipitated and rinsed several times with fresh portions of 2-propanol. The polymer was then dried by evaporating the solvent at 50 °C and kept in a desiccator until further use. Blank membranes prepared with the purified polymer without any additives showed a small cationic response when in contact with potassium chloride solutions. The magnitude of the response increased even when a small amount of KTFPB (0.5 wt %) was added to the membrane. These observations suggest that the cationic impurities had been effectively removed from the polymer. The remaining studies were performed with this purified polymer. Valinomycin-based ISEs are routinely used in clinical analyzers for the determination of potassium in blood serum.1 Membranes containing 1 wt % valinomycin and 60 mol % KTFPB (with respect to the ionophore) were prepared with mass ratios of 10:90, 25:75, and 50:50 DOS:BioSpan-S. The slopes of the calibration plots for the corresponding electrodes were sub-Nernstian. The best results
Figure 3. Potentiometric responses of a BioSpan-S membrane containing 1 wt % valinomycin, 60 mol % KTFPB (relative to the ionophore) and 50 wt % DOP in 0.0100M Tris-HCl, pH 7.2: (1) potassium, (2) ammonium, (3) lithium, (4) sodium, (5) magnesium, and (6) calcium.
Figure 4. Potentiometric responses of a BioSpan-S membrane containing 1 wt % sodium ionophore X, 60 mol % KTFPB (relative to the ionophore), and 50 wt % DOP in 0.0100M Tris-HCl, pH 7.2: (1) sodium, (2) potassium, (3) lithium, (4) ammonium, (5) magnesium, and (6) calcium.
Table 1. Summary of Calculated and Required Selectivity Coefficients log Kpot i,j matrixa PVCb BioSpan-S/DOS BioSpan-S/DOP requiredc
K+ 0 0 0 0
PVCd BioSpan-S/DOP requiredc
-2.0 -2.3
