Anal. Chem. 1999, 71, 1257-1264
Enhanced Microdialysis Extraction Efficiency of Ibuprofen in Vitro by Facilitated Transport with β-Cyclodextrin Alexander N. Khramov and Julie A. Stenken*
Department of Chemistry, Rensselaer Polytechnic Institute, 110 8th Street, Troy, New York 12180-3590
A novel approach to increase microdialysis recovery (extraction efficiency, Ed) by facilitated transport through the microdialysis membrane is described. This new approach facilitates mass transport into the microdialysis probe by inclusion of a complexation agent in the microdialysis perfusion fluid. In these studies, β-cyclodextrin (β-CD) (0.25-2.0 w/v%) was included in the microdialysis perfusion fluid consisting of a Ringer’s solution (155 mM NaCl, 4.0 mM KCl, 2.4 mM CaCl2). β-CD forms known inclusion complexes with 2-(4-isobutylphenyl)propionic acid (ibuprofen). Ibuprofen Ed was significantly enhanced (1.5-2.0 times) through different microdialysis membrane materials. The effect of microdialysis membrane material (polycarbonate/polyether, AN-69, cuprophan), pH, β-CD concentration, and ibuprofen concentration on the Ed was examined. Only the polycarbonate/ polyether membrane was able to give an Ed greater than 100%. In general, a maximum increase in Ed was found when 0.5 w/v% β-CD was included in the perfusion fluid. Variations in the ibuprofen concentration external to the microdialysis probe did not significantly change Ed when 0.5 w/v% β-CD was included in the perfusion fluid. In contrast to the ibuprofen data, β-CD inclusion in the microdialysis perfusion fluid did not affect antipyrine Ed. Antipyrine does not form known inclusion complexes with β-CD. The ability of β-CD to increase microdialysis Ed is explained by facilitated transport. Microdialysis is a well-known in vivo sampling method and has found numerous applications in neurochemistry, neurophysiology, and pharmacology.1 Microdialysis sampling techniques for pharmacokinetic2 and neurochemical applications have been reviewed.3 Microdialysis sampling is a dynamic technique, and diffusion of analytes through the semipermeable dialysis membrane occurs under nonequilibrium conditions. The technique involves perfusion of an iso-osmotic solution through the inner fiber lumen of the microdialysis probe, while the outside of the microdialysis membrane is immersed in either a tissue (in vivo) or an aqueous (industrial or in vitro applications) sample. The concentration of the outlet perfusion fluid is related to the true * Corresponding author: (phone) (518) 276-2045; (fax) (518) 276-4887; (e-mail)
[email protected]. (1) Robinson, T., Justice, J. B., Eds. Microdialysis in the Neurosciences; Elsevier: Amsterdam, 1991. (2) Elmquist, W. F.; Sawchuk, R. J. Pharm. Res. 1997, 14, 267-288. (3) Parsons, L. H.; Justice, J. B., Jr. Crit. Rev. Neurobiol. 1994, 8, 189-220. 10.1021/ac9811930 CCC: $18.00 Published on Web 02/20/1999
© 1999 American Chemical Society
sample concentration via the calibration of the microdialysis probe, or its extraction efficiency, Ed. Methods for calculating Ed in microdialysis sampling have been reviewed.4,5 Ed can be calculated using eq 1, where Cinlet is the analyte inlet perfusion concentration,
extraction efficiency ) Ed ) (Cinlet - Coutlet)/(Cinlet - Csample) (1)
Coutlet is the analyte outflowing dialysate concentration, and Csample is the analyte concentration in the sample far away from the probe. Because microdialysis is entirely a diffusion-based separation (diffusion through the sample, membrane, and dialysate), Ed values of greater than 100% cannot be obtained either in vitro or in vivo. However, microdialysis Ed approaching nearly 100% can be obtained by using slow perfusion methods with perfusion flow rates of less than 100 nL/min.6,7 One of the principal difficulties of applying microdialysis sampling in vivo is that of obtaining significant sample concentration in the outflowing dialysate with drugs that are highly protein bound. Microdialysis sampling only samples the free fraction of the analyte in the medium in which the probe is immersed, e.g., the tissue extracellular fluid space (ECF) or blood.8 Compounds that are significantly protein-bound have low ECF concentrations.9 With greatly reduced free fractions of the drug, analysis becomes a challenge whether or not microdialysis sampling is applied. Since microdialysis sampling as presently applied only gives an Ed near or below 100%, the analytical challenge becomes even greater with highly protein-bound drugs. An additional analytical challenge manifests itself with respect to the nature of the microdialysis sample. Microdialysis perfusion flow rates typically range between 0.5 and 2.0 µL/min for samples that will be analyzed by HPLC methods. These low perfusion flow rates give sample volumes that range between 5 and 20 µL. These small sample volumes make preconcentration techniques difficult. Approaches to solving this problem in microdialysis have included the use of sample stacking in capillary electrophoresis to achieve preconcentration.10,11 However, sample stacking in capillary elec(4) Justice, J. B., Jr J. Neurosci. Methods 1993, 48, 263-276. (5) Stenken, J. A. Anal. Chim. Acta 1999, 379, 337-358. (6) Menacherry, S.; Hubert, W.; Justice, J. B., Jr. Anal. Chem. 1992, 64, 577583. (7) Lada, M. W.; Vickroy, T. W.; Kennedy, R. T. Anal. Chem. 1997, 69, 45604565. (8) Bungay, P. M.; Morrison, P. F.; Dedrick, R. L. Life Sci. 1990, 46, 105-119. (9) Oravcova´, J.; Bo ¨hs, B.; Lindner, W. J. Chromatogr., B 1996, 677, 1-28.
Analytical Chemistry, Vol. 71, No. 7, April 1, 1999 1257
trophoresis is only effective when analytes possess a charge. Our approach is to develop a method that can increase the analyte concentration in microdialysis samples that is not based on special instrumental techniques. We have focused on developing a method to facilitate analyte mass transport into the microdialysis inner fiber lumen. Facilitated or carrier-mediated transport based on mobile carriers in membranes is a well-known separation process.12 Because facilitated transport combines diffusion with chemical reaction, the flux of the diffusing species will be increased.13 Facilitation of transport across polymeric membranes generally involves incorporating a mobile carrier into a supported liquid membrane.14 The donor and receiver conditions as well as the carrier that is used are such that quite selective transport conditions can be obtained. This selectivity has been applied to categories of chemicals as diverse as catecholamines15 and inorganic anions.16 It would be difficult to prepare a stable liquid membrane for use in microdialysis sampling. Furthermore, the mobile carriers in supported liquid membrane separations are often placed in organic solvents such as o-nitrophenyl n-octyl ether (NPOE), which may be quite harmful to an in vivo tissue sample. Our approach to facilitated transport is similar to liquid membrane transport, but instead of impregnating the membrane with a mobile carrier, we have included the carrier in the microdialysis perfusion medium. Our carrier or complexation reagent in this case is β-cyclodextrin (β-CD). Cyclodextrins are well-known cyclic oligosacharides that have the capability to form inclusion complexes with various organic molecules by the capture of the guest molecule into a hydrophobic central cavity.17,18 The host/guest complexation reactions with cyclodextrins are very important to drug delivery systems and improvements of drug properties, such as solubility.19,20 In our case, the complexation reaction with cyclodextrin on one side of a membrane leads to facilitated diffusion of analyte. This facilitation is due to the reaction of the analyte with the cyclodextrin, which lowers the effective concentration of the analyte at the membrane wall. By decreasing the effective analyte concentration, the concentration gradient to the membrane wall becomes greater and therefore analyte flux is increased by this facilitated transport as compared to the free diffusion mode as shown in Figure 1. The increase in Ed in the presence of cyclodextrin in perfusion fluid is expected to correlate with the binding constant of the guest/cyclodextrin complex. (10) Hadwiger, M. E.; Torchia, S. R.; Park, S.; Biggin, M. E.; Lunte, C. E. J. Chromatogr., B 1996, 681, 241-249. (11) Zhao, Y.; McLaughlin, K.; Lunte, C. E. Anal. Chem. 1998, 70, 4578-4585. (12) Cussler, E. L. Diffusion: Mass Transfer in Fluid Systems; Cambridge University Press: Canbridge, 1984; pp 395-408. (13) Crank, J. The Mathematics of Diffusion, 2nd ed.; Oxford University Press: Oxford, 1975. (14) Bartsch, R. A., Way, J. D., Eds. Chemical Separations with Liquid Membranes; ACS Symposium Series 642; American Chemical Society: Washington, DC, 1996. (15) Paugam, M.-F.; Bien, J. T.; Smith, B. D.; Chrisstoffels, L. A. J.; de Jong , F.; Reinhoudt, D. N. J. Am. Chem. Soc. 1996, 118, 9820-9825. (16) Visser, H. C.; Rudkevich, D. M.; Verboom, W.; de Jong, F.; Reinhoudt, D. N. J. Am. Chem. Soc. 1994, 116, 11554-11555. (17) Connors, K. A. Chem. Rev. 1997, 97, 1325-1357. (18) Rekharsky, M. V.; Inoue, Y. Chem. Rev. 1998, 98, 1875-1917. (19) Uekama, K.; Hirayama, F.; Irie, T. Chem. Rev. 1998, 98, 2045-2076. (20) Loftsson T.; Brewster M. E. J. Pharm. Sci. 1996, 85, 1017-1025.
1258 Analytical Chemistry, Vol. 71, No. 7, April 1, 1999
Figure 1. Schematic representation of two different microdialysis modes, which demonstrates the approach of enhancing the microdialysis Ed using cyclodextrin additive to perfusion fluid. (A) In routine systems, microdialysis separation is based on the diffusion behavior and affected only by the membrane structure and analyte diffusion properties. (B) A schematic model depicting the facilitated diffusion based on the complexation reaction of analyte with cyclodextrin in perfusate. The Roman numerals I and II refer to the derivation in the Appendix.
Cyclodextrins have been used extensively in different areas of analytical chemistry. These applications have been reviewed by Li and Purdy.21 It is in the area of separations that cyclodextrins have been extensively utilized.22 Cyclodextrins have also been extensively used in chiral separations in gas chromatography, liquid chromatography, and capillary electrophoresis.23-25 Cyclodextrins have been used in liquid membrane separations. 26 To apply cyclodextrins to improve Ed in microdialysis sampling, we studied experimentally the effect of pH, β-CD concentration, and analyte concentration on the microdialysis Ed of 2-(4isobutylphenyl)propionic acid (ibuprofen) as a model system. We chose β-CD because it has been well-studied as a complexing agent. Ibuprofen was chosen as a model substance because of its well-known ability to form the inclusion complexes with β-CD.27,28 Antipyrine was chosen as a control analyte because it does not form known inclusion complexes with β-CD. EXPERIMENTAL SECTION Chemicals. Racemic ibuprofen (4-isobutyl-R-methylphenylacetic acid) was purchased from Aldrich (Milwaukee, WI). Antipyrine was purchased from Sigma (St. Louis, MO). β-CD (pharmaceutical grade) was obtained as a generous gift from Wacker Chemicals (Norwalk, CT). All other chemicals were of analytical-grade purity. (21) Li, S.; Purdy, W. C. Chem. Rev. 1992, 92, 1457-1470. (22) Hinze, W. L. Sep. Purif. Methods 1981, 10, 159-237. (23) Armstrong, D. W. Anal. Chem. 1987, 59, 84A-91A. (24) Armstrong, D. W.; Tang, Y.; Ward, T.; Nichols, M. Anal. Chem. 1993, 65, 1114-1117. (25) Gratz, S. R.; Stalcup, A. M. Anal. Chem. 1998, 70, 5166-5171. (26) Armstrong, D. W.; Jin, H. L. Anal. Chem. 1987, 59, 2237-2241. (27) Rawjee, Y. Y.; Staerk, D. U.; Vigh, G. J. Chromatogr. 1993, 635, 291-306. (28) Mura, P.; Bettinetti, G. P.; Manderioli, A.; Faucci, M. T.; Bramanti, G.; Sorrenti, M. Int. J. Pharm. 1998, 166, 189-203.
Acetonitrile and methanol (HPLC grade, Fisher Scientific, Fair Lawn, NJ) were used as received. Microdialysis. Three different microdialysis probes each with a 4-mm-length membrane were used in the experiments. The membranes and their commercial manufacturer’s are as follows: polycarbonate/polyether (PC), CMA-12 (CMA/Microdialysis, Acton, MA), 20 000 molecular weight cutoff (MWCO), o.d. 0.5 mm, i.d. 0.4 mm; cuprophan (CUP), a regenerated cellulose membrane, CMA-11 (CMA/Microdialysis), 6000 MWCO, o.d. 0.24 mm, i.d. 0.19 mm; AN-69, a copolymer of polyacrylonitrile and methyl sulfonate, and BR-4 (Bioanalytical Systems, Inc., West Lafayette, IN) 29 000 MWCO, o.d. 0.34 mm, i.d. 0.24 mm. Prior to use, the microdialysis probes were prepared according to the manufacture’s instructions. After use, the probes were placed in distilled/deionized water to prevent membrane drying. Probe membranes were used multiple times. There were no statistical differences in ibuprofen Ed between new and used probe membranes. A CMA/102 microsyringe pump (CMA/Microdialysis) was used to pump the perfusion fluid through the microdialysis probes. An in-house prepared Ringer’s solution (155 mM NaCl, 5.6 mM KCl, and 2.4 mM CaCl2) was used as the perfusion fluid. Sample solutions were also prepared in Ringer’s. All experiments were performed at ambient room temperature 25 ( 0.5 °C. Saline (0.9 w/v%) was buffered with either 10 mM sodium acetate or phosphate to control the pH. Ringer’s solution cannot be added to phosphate buffers in the range around or above pKa2 (H3PO4) because of the formation of insoluble CaHPO4 (pKsp ) 6.58).29 The concentration of β-CD was varied between 0.25 and 2.0 w/v% (0, 0.25, 0.5, 1.0, 1.5, and 2.0 w/v%). The maximum solubility of β-CD in our system is 2.0 w/v%. Analyte solutions were stirred using a Thermolyne (Barnstead) magnetic stirrer at constant stirring rate to reduce solution boundary layer effects and to achieve maximum Ed through the membrane.30 Liquid Chromatography. Dialysate samples were analyzed for ibuprofen or antipyrine by HPLC-UV. The chromatographic experiments were performed with Spectra Systems HPLC equipment (Thermo Separation Products, Inc., Riviera Beach, FL) with UV detection at 220 nm using a column (150 × 2 mm) packed with Spherex 3 C-8 (Phenomenex, Torrance, CA) and a mobile phase of either acetonitrile/methanol/water (75/8.3/16.6 v/v/v%) with a final concentration of 0.01 M sodium phosphate buffer (pH 2.6) (ibuprofen) or (45/55 v/v%) acetonitrile/water with a final concentration of 0.1 M sodium phosphate buffer, pH 2.6 (antipyrine) at 0.18 mL/min. The mobile-phase solutions were filtered through a 0.22-µm filter and degassed by ultrasound for 5 min before use. Analyte concentration was determined by comparing peak area to a calibration curve constructed by using a range of 1-200 µM. Calibration curves that were prepared using different concentrations of β-CD gave the same slope and intercept for both peak area and peak height. RESULTS AND DISCUSSION Influence of β-CD Concentration. Ibuprofen Ed experiments were performed with a range of β-CD concentrations (0, 0.25, 0.5, 1.0, 1.5, and 2.0 w/v%) in the microdialysis perfusion fluid. A (29) Harris, D. C. Quantitative Chemical Analysis, 4th ed.; W. H. Freeman and Co.: New York, 1995; p AP-18. (30) Stenken, J. A.; Topp, E. M.; Southard, M. Z.; Lunte, C. E. Anal. Chem. 1993, 65, 2324-2328.
Figure 2. Polycarbonate/polyether membrane. Plot of ibuprofen Ed vs β-CD concentration at different flow rates: 0.5 ([), 1.0 (1), 1.5 (2), 2.5 (b), and 5.0 µL/min (9).
Figure 3. Polyacrylonitrile (AN-69) membrane. Plot of ibuprofen Ed vs β-CD concentration at different flow rates: 0.5 ([), 1.0 (1), 1.5 (2), 2.5 (b), and 5.0 µL/min (9).
maximum of 2.0 w/v% was used in these experiments because of the limited solubility of β-CD in aqueous solutions.31 Figures 2-4 illustrate the influence of β-CD concentration in the perfusion fluid for ibuprofen Ed through PC (Figure 2), CUP (Figure 3), and AN69 (Figure 4) membranes. Ibuprofen microdialysis Ed was increased through each of the membrane materials particularly when low concentrations of β-CD (0.25 and 0.5 w/v%) were included in the microdialysis perfusion fluid. The extent of increase of the Ed was dependent upon the microdialysis perfusion flow rate and the concentration of β-CD in the perfusion fluid. Table 1 shows the percent enhancement over controls obtained with each perfusion flow rate at different β-CD concentrations perfused through different microdialysis membranes. Only PC (31) Saenger W. Angew. Chem., Int. Ed. Engl. 1980, 19, 344-362.
Analytical Chemistry, Vol. 71, No. 7, April 1, 1999
1259
Table 1. Percent Increase in Ibuprofen Ed through Various Microdialysis Membranes % enhancement for β-CD (w/v%) flow rate (µL/min)
membraneb
control
0.25
0.5
1.0
1.5
2.0
0.5 1.0 1.5 2.5 5.0
PC
67.9 ( 1.6 45.9 ( 0.5 35.6 ( 0.9 24.4 ( 0.1 13.7 ( 0.4
72.7 48.6 36.1 30.0 27.3
56.8 29.4 23.1 4.5 -1.5
34.2 18.3 4.5 2.0 8.6
7.3 0.3 -3.4 -11.7 -10.9
-1.1 -17.3 -27.3 -0.4 -26.7
0.5 1.0 1.5 2.5 5.0
AN-69
50.0 ( 1.4 35.4 ( 0.8 26.2 ( 1.3 18.5 ( 1.0 10.1 ( 0.2
29.3 10.0 9.2 -2.0 -4.9
47.5 20.9 13.9 1.7 -8.9
8.9 -3.1 -5.0 -14.3 -20.3
7.5 -10.8 -13.9 -19.0 -27.4
-3.2 -18.4 -22.8 -34.7 -42.4
0.5 1.0 1.5 2.5 5.0
CUP
44.2 ( 0.3 26.3 ( 0.4 18.7 ( 0.4 11.5 ( 0.4 4.7 ( 0.2
27.5 20.5 22.7 25.4 59.2
61.3 55.0 47.9 49.4 92.2
49.6 35.2 44.5 55.4 122.3
39.6 33.0 32.4 35.2 77.8
27.0 24.9 25.9 34.1 40.8
a Controls are mean ( SD for n ) 3 from one probe in a well-stirred medium. b PC, polycarbonate/polyether; AN-69, copolymer of polyacrylonitrile and methyl sulfonate; CUP, cuprophan.
Figure 4. Cellulose (cuprophan) membrane. Plot of ibuprofen Ed vs β-CD concentration at different flow rates: 0.5 ([), 1.0 (1), 1.5 (2), 2.5 (b), and 5.0 µL/min (9).
membranes exhibited increases of microdialysis Ed such that the analyte outlet concentration (Coutlet) was greater than that in solution (Csample); i.e., Ed was greater than 100%. In contrast to the ibuprofen data, antipyrine Ed was not affected by the inclusion of β-CD in the microdialysis perfusion fluid. Figure 5 shows the difference between antipyrine Ed at different flow rates for control and 0.5 w/v% β-CD perfusion fluids. Table 2 shows that, for all concentrations of β-CD in the perfusion fluid, antipyrine Ed is not significantly affected. These data indicate that β-CD at high concentration does not hinder antipyrine diffusion through the microdialysis porous membrane material. The increase in ibuprofen microdialysis Ed by the inclusion of β-CD in the perfusion fluid is explained by the coupling of diffusion with chemical reaction in the microdialysis sampling regime. It is well-known that diffusion coupled with chemical reaction will 1260 Analytical Chemistry, Vol. 71, No. 7, April 1, 1999
Figure 5. Plot of antipyrine Ed with 0.5 wt % (b) and without (9) β-CD with a PC probe at pH 7.4 with 100 µM antipyrine external to the probe. Error bars are mean ( SD (n ) 3 for each perfusion flow rate).
increase mass transport flux in membrane systems.12,32,33 Generally, in these systems, reactions occur at the membrane/sample interface such as that of a thin film. In our experiments, β-CD is included as a freely diffusing complexation agent in the microdialysis perfusion fluid and is not bound to the membrane material. In this case, β-CD behaves similarly to a capillary electrophoresis experiment in which cyclodextrins are added to complex analyte and affect retention behavior generally for the purpose of separating enantiomers.24 However, in our case, the role of the β-CD is not to separate enantiomers of ibuprofen (it is racemic) but simply to complex with ibuprofen in solution as it passes through the microdialysis membrane. Although capillary electrophoresis experiments indicate that racemic ibuprofen forms a complex with (32) Goddard, J. D.; Schultz, J. S.; Bassett, R. J. Chem. Eng. Sci. 1970, 25, 665683. (33) Ward, W. J., III AIChE J. 1974, 16, 405-410.
Figure 6. Plot of ibuprofen Ed with 0.5 w/v% β-CD (b) and 0.0 w/v% β-CD (×) vs microdialysis perfusion flow rate for PC probe at pH 7.4 with 100 µM ibuprofen. Error bars are mean ( SD (n ) 3 at each point). Table 2. Effect of β-CD Concentration on Antipyrine Ed through a PC Membranea flow rate (µL/min)
0
0.5 1.0 5.0
89.3 ( 1.2 60.2 ( 1.2 15.9 ( 0.5
a
β-CD (w/v%) 0.5 1 94.0 ( 0.4 62.2 ( 2.0 17.2 ( 0.3
92.9 ( 0.5 64.3 ( 1.0 18.1 ( 0.5
2 93.2 ( 2.8 64.1 ( 2.3 18.1 ( 0.3
All samples are mean ( SD (n ) 3).
β-CD with different binding energies, it is not expected to see a large difference in the microdialysis extraction efficiency between the enantiomers because CE experiments can achieve separation due to the large number of theoretical plates in that experiment. Increasing the concentration of β-CD in the perfusion fluid leads to the decrease of ibuprofen Ed for all membrane materials and flow rates. This type of behavior is unexpected in equilibrium systems. For most equilibrium systems in which binding occurs, a maximum in the dependent variable (Ed in our case) is achieved via a rectangular hyperbolic equation (eq 2) as the independent
y ) dx/(f + ex)
(2)
variable (concentration of some species) is increased as discussed recently by Armstrong for such cases as fractional composition of a weak acid, enzyme kinetics, and changes in NMR signals with different ligand concentrations. 34 Equation 2 is the expected binding isotherm for 1:1 associations, where y is the dependent variable, x is the ligand concentration, and d, e, and f are constants that are associated with the particular binding process under study.34 A theoretical reason for the observed behavior with ibuprofen and increased β-CD concentration is unclear at the moment. This (34) Armstrong, D. W. In Advances in Chromatography; Brown, P. R., Grushka, E., Eds.; Marcel Dekker: New York, 1998; Vol. 39, pp 239-262.
behavior may or may not be related to the equilibrium chemistry of the analyte/carrier complex. Because of the multiple interactions that occur during the microdialysis experiment, i.e., the reaction with β-CD in the flowing perfusion fluid, the possibility of insoluble complexes and/or aggregation, and the possibility of adsorption of complexes onto the dialysis polymeric material, it is difficult to make a proper mathematical explanation of these data without further investigation of multiple analyte/β-CD facilitated transport complexes in microdialysis. This work is currently under investigation in our laboratory.35 In fact, whether or not equilibrium occurs at the membrane interface is difficult to predict. Even though steady-state conditions occur if kinetic processes are faster than diffusion, the analyte that is moving through the membrane is in a continual concentration gradient that extends into the inner center of the membrane inner fiber lumen. Thus, concentrations of the analyte at the wall are not equal to those away from the wall. The inner dead volumes of the active membrane area for the PC, AN-69, and CUP probes are 0.31, 0.11, and 0.04 µL, respectively. For the PC probe (highest internal dead volume) at the lowest flow rates used in these studies, 0.5 µL/ min, this volume would be swept within 37 s. All the others would be swept through much faster at the 0.5 µL/min volumetric flow rate. Whether a true equilibrium could be established under these conditions is unknown at this time. The case with the microdialysis facilitated transport is greatly different from either a chromatographic or electrophoretic separation. The greatest difference that must be clearly pointed out is that a “partitioning” process does not really occur within the membrane itself. These dialysis membranes are porous and have a significant water content within their pores. Mass transport into the inner fiber lumen of the microdialysis device normally occurs via diffusion through the water-filled pores, not partitioning into the membrane followed by diffusion. Therefore, there are generally not multiple equilibrium processes occurring at the membrane/ perfusion fluid interface. Influence of Microdialysis Perfusion Flow Rate. In these experiments, ibuprofen diffuses across the microdialysis membrane and then reacts with β-CD in the microdialysis perfusion solution. This makes the effective concentration of ibuprofen at the inner membrane wall with β-CD in solution lower than in the free diffusion case. This increase in the concentration gradient to the microdialysis membrane causes an increased flux, thus increasing the effective Ed for ibuprofen. Since the β-CD/ibuprofen complex is at a chemical potential different from ibuprofen, it is possible for more ibuprofen to enter the inner fiber lumen, thus achieving total concentrations that are greater than the concentration of ibuprofen external to the probe. The microdialysis perfusion flow rate affects the relative increase in ibuprofen Ed with the inclusion of β-CD. The reason for this is that faster perfusion flow rates make the effective concentration of ibuprofen at the inner membrane wall near zero with or without β-CD. In other words, the analyte does not have enough time to react with β-CD because the perfusion fluid is sweeping across the membrane too fast to allow appreciable concentrations of the analyte to form inclusion complexes with β-CD in the perfusion medium. This observed data is better (35) Khramov, A. N.; Stenken, J. A. unpublished observations.
Analytical Chemistry, Vol. 71, No. 7, April 1, 1999
1261
Table 3. Effect of Internal and External Fluid pH the Ed of Ibuprofen between Controls and 0.5 w/v% β-CDa pH 5.5
a
pH 6.9
pH 8.0
flow rate (µL/min)
0.0 β-CD
0.5 β-CD
0.0 β-CD
0.5 β-CD
0.0 β-CD
0.5 β-CD
5.0 3.5 2.5 1.5 1.0 0.5
14.0 ( 0.5 18.0 ( 0.2 24.9 ( 0.1 35.5 ( 0.8 46.5 ( 0.8 58.4 ( 2.1
26.0 ( 0.4 31.6 ( 1.5 40.9 ( 1.7 62.6 ( 1.8 83.7 ( 0.4 132.3 ( 3.4
17.1 ( 1.3 20.5 ( 0.5 27.5 ( 0.4 39.2 ( 0.7 51.6 ( 1.5 76.4 ( 0.2
21.8 (1.4 27.8 ( 1.8 35.8 ( 2.2 54.2 ( 2.6 71.8 ( 0.7 125.3 ( 4.1
13.8 ( 0.9 17.1 ( 0.9 23.3 ( 0.8 35.0 ( 1.0 48.3 ( 3.1 73.6 ( 0.9
18.2 ( 1.3 25.3 ( 1.3 31.8 ( 0.5 51.7 ( 1.6 69.7 ( 0.2 129.9 ( 3.1
Ed is given as mean ( SD. n ) 3 from each probe at each perfusion flow rate.
Table 4. Effect of Varying Ibuprofen Concentration on Ed with 0.5 w/v% β-CD
a
flow (µL/min)
0 w/v% β-CD 100 mM ibuprofen
100 µM ibuprofen
0.5 w/v% β-CD 50 µM ibuprofen
25 µM ibuprofen
1.0 0.5
57.5 ( 0.4 82.7 ( 1.8a
66.9 ( 117.4 ( 5.3a
63.1 ( 1.9 120.3 ( 1.1a
68.8 ( 1.8a 117.5 ( 0.9a
2.6a
n ) 2; otherwise n ) 3 samples from the same probe in one stirred beaker.
illustrated by eq 3 (eq A7 in the Appendix). In this equation, the
( )
dCL 2πDmem(Co - CL) πRi2kCL ) dx Q Q ln(Ro/Ri)
(3)
change in the free, unbound concentration of the ibuprofen, CL, vs the position of the perfusion fluid in the dialysis probe, x, is related to different experimental conditions, which include the following: Dmem (cm2/s), the analyte diffusion coefficient through the porous, polymeric dialysis membrane; Co (M), the external analyte concentration; Q (cm3/s), the volumetric flow rate through the microdialysis probe; Ro and Ri (cm), the inner and outer radii of the probe membrane; and k (s-1) a first-order kinetic rate constant. The ratio of the kinetic rate constant to the volumetric flow rate in the second term clearly illustrates that as the perfusion flow rate is increased, the ability of the kinetic reactions to increase flux through the membrane diminishes. Influence of pH. Ibuprofen has an ionizable carboxylic group in its molecular structure; therefore, the solubility and ionization of ibuprofen are dependent on pH. In general, the ionized and the nonionized forms of ibuprofen are supposed to form inclusion complexes with β-CD, although the literature data on the complexation of the form are in contradiction to each other. In one work,36 the complexation of ibuprofen was found as pH-independent. However, other authors21 reported that the complex formation constant of protonated ibuprofen is ∼2 times larger that the complexation constant of its anion. To examine the influence of pH on the Ed of ibuprofen, we performed a series of experiments with pH of the internal and external media that varied between 5.0 and 8.0. Table 3 shows the influence of pH on ibuprofen Ed for the PC membrane with 0.5 w/v% β-CD. Figure 5 shows the enhancement for the Ed of ibuprofen between controls and 0.5 w/v% β-CD at pH 7.4. Note that these values are quite similar to those obtained in unbuffered Ringer’s solution. At all perfusion flow rates greater than 0.5 µL/ (36) Menard, F. A. Ph.D. Dissertation, University of Rhode Island, 1988.
1262 Analytical Chemistry, Vol. 71, No. 7, April 1, 1999
min, increases in pH increased the amount of Ed enhancement for ibuprofen. However, the data at 0.5 µL/min suggest that a maximum is reached for all pH values. In the absence of β-CD, low pH (
