Separation and Enhanced Detection of Anesthetic Compounds Using Solid Phase Micro-Extraction (SPME)–Raman Spectroscopy Ikechukwu C. Nwaneshiudu,a Chinwe A. Nwaneshiudu,b Daniel T. Schwartza,* a b
University of Washington, Department of Chemical Engineering, Box 351750, Seattle, WA 98195, USA University of Washington, Department of Anesthesiology, Box 356540, Seattle, WA 98195, USA
Polydimethylsiloxane (PDMS)-based solid-phase micro-extraction (SPME) was used along with Raman spectroscopy (RS) to separate and enhance the detection of five anesthetic compounds (halothane, propofol, isoflurane, enflurane, and etomidate) from aqueous and serum phases. Raman signals in the spectral ranges 250–450 cm1 and 950–1050 cm1 allowed the unique characterization of all five compounds when extracted into the PDMS phase. The SPME-RS detection of clinically relevant concentrations of aqueous propofol (6.5 lM) and halothane (200 lM) is shown. We quantify the partition coefficient for aqueous halothane in PDMS as log K ¼ 1.9 6 0.2. Solid-phase micro-extraction of the anesthetics makes their detection possible without the strong autofluorescent interference of serum proteins. Because of low solubility and/or weak Raman scattering, we found it challenging to detect enflurane, isoflurane, and etomidate directly from the aqueous phase, but could we do so with SPME enhancement. These studies show the potential of SPME-RS as a method for the direct detection of anesthetics in blood. Index Headings: Anesthetics; Solid-phase micro-extraction; SPME; Partition; Propofol.
INTRODUCTION Clinical anesthesiology is continually adapting to new and improved anesthetic compounds.1 The transition from volatile compounds like halothane toward such anesthetics as isoflurane and sevoflurane is largely responsible for the reduced number of anesthesiarelated complications.1–3 In addition, the use of fastacting intravenous anesthetics such as propofol, ketamine, and etomidate has also improved clinical practice.1,2 However, a continuing challenge in the field is accounting for the distribution of these anesthetics in the body.4,5 Although overall estimates can be made using the detection of inhaled and exhaled gas concentrations, the actual concentrations in the body remain elusive due to the partitioning of these compounds into lipid-like phases in the blood and tissue.4 This has led to complications during and after surgical procedures.5–7 Ways to directly measure concentrations in relevant biological fluids may be beneficial to researchers and clinicians. The partitioning of volatile and intravenous anesthetics between physiological phases such as blood and brain, alveolar gas and blood, and blood and lipid has Received 7 November 2013; accepted 2 May 2014. * Author to whom correspondence should be sent. E-mail: dts@u. washington.edu. DOI: 10.1366/13-07362
1254
Volume 68, Number 11, 2014
been well documented.8–13 The magnitude of partitioning is directly estimated by the equilibrium partition coefficient (K), the thermodynamic parameter that quantitatively relates the analyte concentration in each phase. Although physiological partitioning can complicate administration and detection of anesthetics,9 the same partitioning process can be used to enhance the detection of analytes via the method solid-phase microextraction (SPME).14 Solid-phase micro-extraction is a well-known analytical technique that uses the hydrophobic traits of polymers to extract and pre-concentrate nonpolar analytes for better detection.14–20 The use of polydimethylsiloxane (PDMS), a staple biocompatible polymer for SPME analytics, has been well established for liquid as well as gas phase detection. Techniques that benefit from SPME pre-concentration are well documented in the literature.16,21–24 In addition to signal enhancement, SPME has the potential to separate macromolecules that may interfere with the detection of smaller target analytes. This sort of separation could prove very useful when dealing with analytes in a complex system such as blood or serum. Solid-phase micro-extraction coupled with Raman spectroscopy (SPME-RS), is emerging as a quantitative analytical method, showing improved detection of one to three orders of magnitude over normal Raman spectroscopy (RS).17,21 In the past, Raman scattering-based devices such as the RASCAL (Raman scattering analyzer) were used in hospitals for monitoring gas-phase anesthetics during surgery.25 New opportunities may exist for combining SPME-RS with anesthesiology in both research and clinical practice. Here we explore using PDMS-based SPME to enhance RS detection directly in relevant aqueous and serum liquid phases. Three volatile anesthetic compounds (isoflurane, enflurane, and halothane) and two intravenous anesthetics (propofol and etomidate) were examined for their enhanced detectability using SPME polymer phases in solution.
MATERIALS AND METHODS Many of the SPME method details are provided in Nwaneshiudu et al. 17 We summarize key features briefly next. Polydimethylsiloxane and Organic/Aqueous Solution Preparation. Polydimethylsiloxane Samples. Slygard 184 PDMS and catalyst were obtained (Dow Corning). Samples of PDMS used in the experiment
0003-7028/14/6811-1254/0 Q 2014 Society for Applied Spectroscopy
APPLIED SPECTROSCOPY
FIG. 1. Raman spectra of the neat anesthetic compounds. (a) Halothane. (b ) Propofol. (c) Isoflurane. (d ) Enflurane. (e) Etomidate. Multipliers next to spectra denote the scale used to make the spectral features visible on the same graph.
were prepared as follows. Uncured PDMS was mixed using the standard 10 :1 polymer-to-catalyst ratio and poured into 1.5 mm deep poly(methyl methacrylate) (PMMA) mold. This was cured at 70 8C overnight to make an 1.5 mm thick sheet of PDMS. Samples were punched out of the PDMS sheet using a 2 mm diameter punch. The resulting PDMS plugs were sealed and kept until ready for use. Organic/Aqueous Solutions. Halothane, propofol, isoflurane, enflurane, and etomidate were all purchased (Sigma Aldrich), and all were used as purchased. Pooled human serum was purchased (Biochemed Services, Winchester, VA). Aqueous solutions were prepared by adding the appropriate amount of reagent to 20 ml deionized water. Serum samples used 2 ml aliquots. The measuring was done volumetrically for liquid anesthetics (halothane, propofol, isoflurane, and enflurane). The solutions were then shaken vigorously, tightly capped, and left to fully dissolve. The spiked solutions were used within 15 min of preparation to prevent evaporation loses. Solutions of etomidate were prepared by introducing solid etomidate into 20 ml deionized water and letting the solution equilibrate for 72 h. These solutions were filtered before use to separate out undissolved solids from the saturated solution. Solid-Phase Micro-Extraction–Raman Spectroscopy Procedure. Water Solid-Phase Micro-Extraction. We used 20 ml scintillation glass vials for the SPME procedure. Solutions of halothane, propofol, isoflurane, enflurane, and etomidate were prepared in separate 20 ml vials. The PDMS plugs were introduced into each vial; the vials were capped and left to equilibrate undisturbed at room temperature for at least 48 h. Serum Solid-Phase Micro-Extraction. We used glass vials (2 ml) for the SPME procedure. The vials were filled with serum and spiked with halothane and propofol to reach the desired concentrations. The PDMS plugs were introduced into each vial; the vials were capped and left to equilibrate undisturbed at room temperature for at least 48 h.
Raman Spectra Acquisition. Raman spectra were collected using an inVia Raman micro-spectrometer (Renishaw) attached to a DM IRBE upright optical microscope (Leica). A 785 nm diode laser operated at full power was used to irradiate samples through a 503 objective lens (numerical aperture (NA) 0.8). The spot area was 50 lm2. Raman scattered light was acquired through the same objective lens and detected using a thermoelectrically cooled (60 8C) charge-coupled detector. All PDMS sample spectral measurements were acquired for 10 s. Acquisition times for the water-sample measurements ranged from 10 to 300 s, depending on signal-to-noise concerns. The spectra of the neat compounds (halothane, isoflurane, enflurane, and propofol) were collected using a 103 objective, which was irradiated in a capped 2 mL glass vial. The spectra of the equilibrated PDMS and water samples were collected using a wet sample holder sealed with a glass coverslip. Laser stability was confirmed using the 520 cm1 peak of silicon. The PDMS peak at 1410 cm1 was acquired with all the solid-phase spectra and served as an internal standard that helped normalize the sample-to-sample variations in the system focus. Raman Scattering Intensity Normalization and Peak Analysis. The spectral peaks were analyzed using WiRE 2.0 software (Windows-based Raman Environment; Renishaw) and fit to standard Voigt distribution profiles. The curve fit parameters were used to calculate the integrated peak areas. The integrated peak areas of the PDMS internal standard were used to normalize the reported analyte peak areas. To account for acquisitiontime differences, peak intensities are reported in counts per second.
RESULTS AND DISCUSSION Figure 1 shows the characteristic Raman spectra of all five of these anesthetic compounds (neat) in the 200– 1090 cm1 range. These compounds represent diverse anesthetics, and over this range, there are unique spectral features for these compounds that make RS effective for fingerprinting. Halothane is the most strongly scattering and etomidate is the weakest scattering of the compounds tested. Halothane is used to illustrate the quantitative aspect of SPME partitioning (methods detailed in prior work17). We examined the C–Cl bending mode of halothane at 310 cm1 in the spectral ranges 285–350 cm1. These peaks do not overlap with the PDMS or the water peaks. Figure 2 shows the Raman spectral peaks for the C–Cl bending mode of halothane at 310 cm1 taken in the equilibrated PDMS (Fig. 2, top) and aqueous (Fig. 2, bottom) phases. The baseline subtracted spectral data points are presented along with the corresponding curve fits and scale bars to show the intensity of the Raman signal in counts per second. Comparing detection in PDMS to that in water, we can see an 50-fold peak-intensity enhancement. Because of the low halothane signal in water, the aqueous spectra at the concentrations 2 and 0.2 mM are not shown in Fig. 2. This figure illustrates that SPME-RS enables the straightforward detection of halothane at concentrations that are clinically relevant (0.25–0.50 mM).12
APPLIED SPECTROSCOPY
1255
FIG. 2. Raman spectra of the 310 cm1 halothane peak. (top) Acquired in the PDMS phase. (bottom) Acquired in the water phase. Lines a–e indicate aqueous concentrations of 20, 15, 10, 2, and 0.2 mM, respectively. Data points are raw data, and lines are the Voigt curve fits.
Quantifying Partitioning. Figure 3 shows a plot of the 310 cm1 halothane peak area as a function of the aqueous halothane concentration. Peaks areas were normalized to the PDMS internal standard peak at 1410 cm1, eliminating small variations in PDMS sample preparation. Data points represent triplicate measurements performed on two separate samples and were fit using linear regression analysis. The linearity of this plot suggests we are in the dilute range where there is negligible polymer swelling and solute–solute interactions are modest. The normalized peak areas in Fig. 3 are proportional to the analyte concentration, Raman cross section of the analyte in each media, and instrumental factors. If we assume the halothane scattering cross section and instrumental factors are identical in water and PDMS (assumptions that have proven reasonable for other SPME–Raman scattering analytes17) then a simple relationship for the equilibrium partition coefficient of halothane (Kh) emerges: Kh ¼
SlopePDMS;h SlopeWater;h
ð1Þ
where SlopePDMS,h is the slope of the best-fit line for halothane measured in the PDMS phase and SlopeWater,h is the slope of the best-fit line measured in the aqueous phase. The partition coefficient estimate from the data in Fig. 3 is log Kh = 1.9 6 0.2. Although no halothane partitioning values for the PDMS or water systems could be found in the literature, others reported values of log Kh between 2.0 and 3.0 for aqueous–lipid partitioning.26,27
1256
Volume 68, Number 11, 2014
FIG. 3. Integrated areas of the 310 cm1 halothane peak versus aqueous halothane concentrations. (top) Acquired in the PDMS phase. (bottom) Acquired in the water phase.
Because PDMS is a hydrophobic polymer known to display lipid-like partitioning, our estimated values appears reasonable.28,29 Qualitative Detection. Next, we examine propofol, a commonly used intravenous anesthetic. Unlike halothane, which has a relatively high water solubility (20 mM), the water solubility of propofol is lower (700 lM). At these concentrations and using reasonable acquisition times ( 300 s), we could not easily detect propofol in the aqueous phase. Concentrations of administered propofol typically range from 1 to 60 lM in blood.4 Using SPME, we could detect propofol readily at clinically relevant concentrations using 10 s acquisition times. Figure 4a shows spectra of the characteristic 1046 cm1 aromatic peak of propofol taken in the equilibrated PDMS. All spectra were acquired with 10 s of exposure, which highlights the improved signal-to noise ratio of the SPME approach. As expected, Fig. 4b shows that the partitioning of propofol behaves linearly at these dilute concentrations. Fingerprinting using RS is challenging for complex multicomponent samples such as blood and urine. The SPME pre-concentration has the potential to selectively separate and concentrate the desired analyte, thereby avoiding interference from other components. As a simple proof of concept, we examined a mixture of 10 mM halothane and 60 lM propofol in human pool serum. Figure 5a shows a Raman spectrum obtained directly in serum, and Fig. 5b shows a typical spectrum in PDMS after the serum mixture was equilibrated with the polymer. Figure 5a displays the broad autofluorescence and overlapping protein spectra that we expect from a complex aqueous protein mixture such as serum. None of
FIG. 4. (panel 1) Spectra a–e acquired in the PDMS phase at aqueous propofol concentrations of 650, 500, 350, 65, and 35 lM, respectively. (panel 2) Representative integrated peak areas of the spectra plotted against aqueous propofol concentration.
the peaks from halothane or propofol is discernible. In contrast, the spectrum acquired from the equilibrated PDMS phase has none of the protein-related interference. Therefore, we are able to resolve some of the anesthetic
FIG. 5. (a) Spectra for a mixture of 10 mM halothane and 60 lM propofol in serum. (b) Spectra for a mixture of 10 mM halothane and 60 lM propofol in equilibrated PDMS. The inset in (b) is the baseline subtracted region where halothane peaks are present.
peaks (box and inset of Fig. 5b). However, it is interesting to note that no propofol peaks appear in the PDMS that was equilibrated with serum. Based on the earlier aqueous propofol experiments, we would expect to see propofol peaks at the concentrations used here. The absence of propofol peaks at such high concentrations brings up an interesting aspect of propofol biochemistry; namely, that propofol is known to complex with proteins in serum.4,30,31 The protein–propofol complexation competes with PDMS for the accumulation of propofol, reducing the SPME effect. The quantitative traits of this competitive partitioning warrants further study. Last, we illustrate the improved detection possible using the SPME of isoflurane, enflurane, and etomidate. These anesthetic compounds gave the weakest Raman signals, as shown in Fig. 1. Although the inherently weak signals from these compounds make them undetectable in water, when they were equilibrated with a PDMS phase, distinct signature peaks could be seen. The main spectrum in Fig. 6 (gray) shows the characteristic peaks of PDMS, which hide any anesthetic peaks on this scale. The spectral regions where these anesthetic peaks reside are 250–450 cm1 (Fig. 6, inset 1) and 950–1050 cm1 (Fig. 6, inset 2). The Raman spectra labeled a and b in Fig. 6 (inset 1) are of 20 mM isoflurane and enflurane, respectively; spectrum c is from PDMS alone. The plots labeled (ac) and (bc) are difference spectra that show the presence of characteristic anesthetic peaks when the PDMS background is subtracted. Figure 6 (inset 2) shows the spectral range for identification of etomidate signals. Spectrum d is of PDMS equilibrated with aqueous etomidate; spectrum e is of PDMS. A characteristic 1005 cm1etomidate peak neighboring a PDMS background peak at 995 cm1 can be seen in spectrum d. The difference spectrum (de) in Fig. 6 (inset 2) shows the weak, characteristic 1005 cm1 etomidate peak without the PDMS peak. In Fig. 6 (inset 1), four distinct isoflurane peaks (280, 340, 370, and 420 cm1) can be seen in difference spectrum (ac) and three enflurane peaks (265,
APPLIED SPECTROSCOPY
1257
FIG. 6. The PDMS (gray) spectrum with black boxes indicating the spectral ranges used to identify isoflurane, enflurane, and etomidate. (Inset 1) Spectral region 250–450 cm1. Spectrum a is of PDMS equilibrated in 20 mM isoflurane; spectrum b is of PDMS equilibrated in 20 mM enflurane; spectrum c is of PDMS. The difference spectra (ac) and (bc) are presented as bold curves. (Inset 2) Spectral region 950– 1050 cm1. Spectrum d is of PDMS equilibrated in aqueous etomidate; spectrum e is of PDMS. The difference spectrum (de) is shown as a bold curve.
360, and 430 cm1) can be seen in difference spectrum (bc). As we can see, SPME does enhance the Raman signals of these anesthetics.
CONCLUSION AND IMPLICATIONS We have shown that SPME improves the Raman signals of all the anesthetics examined in this study. With halothane and propofol, which displayed the strongest Raman signals, SPME lets us detect clinically relevant concentration (200 and 35 lM, respectively) with modest acquisition times (10 s). We quantified the partitioning of halothane into PDMS and obtained a range of partition coefficients (log K = 1.9 6 0.2) for the system. Finally, signals for isoflurane, enflurane, and etomidate that were undetectable in the water phase could be discerned in a PDMS background under similar conditions using SPME. Showing that SPME-RS can detect anesthetics in both single-component aqueous mixtures and complex biological fluids such as serum displays its versatility in detecting in a variety of systems. In addition, because SPME is a pre-concentration technique that does not rely on a particular detection system, it can complement several different analytical platforms. This could open doors to the direct detection of these compounds in biological fluids using a variety of optically based systems. ACKNOWLEDGMENTS We gratefully acknowledge the support of the Gates Foundation Millennium Scholarship as well as Andrew Keefe for providing human pooled serum for the multicomponent experiments. Partial support was also provided by the Boeing Sutter endowment. 1. D.H. Robinson, A.H. Toledo. ‘‘Historical Development of Modern Anesthesia’’. J. Invest. Surg. 2012. 25(3): 141-149.
1258
Volume 68, Number 11, 2014
2. R.D. Urman, S.P. Desai. ‘‘History of Anesthesia for Ambulatory Surgery’’. Curr. Opin. Anaesthesiol. 2012. 25(6): 641-647. 3. A. Dabbagh, S. Rajaei. ‘‘Halothane: Is There Still Any Place for Using the Gas as an Anesthetic?’’ Hepat. Mon. 2011. 11(7): 511-512. 4. F. Kivlehan, F. Garay, J.D. Guo, E. Chaum, E. Lindner. ‘‘Toward Feedback-Controlled Anesthesia: Voltammetric Measurement of Propofol (2, 6-Diisopropylphenol) in Serum-Like Electrolyte Solutions’’. Anal. Chem. 2012. 84(18): 7670-7676. 5. M. Grossherr, A. Hengstenberg, T. Meier, L. Dibbelt, B.W. Igl, A. Ziegler, P. Schmucker, H. Gehring. ‘‘Propofol Concentration in Exhaled Air and Arterial Plasma in Mechanically Ventilated Patients Undergoing Cardiac Surgery’’. Br. J. Anaesth. 2009. 102(5): 608-613. 6. C.D. Kent, G.A. Mashour, N.A. Metzger, K.L. Posner, K.B. Domino. ‘‘Psychological Impact of Unexpected Explicit Recall of Events Occurring During Surgery Performed Under Sedation, Regional Anaesthesia, and General Anaesthesia: Data from the Anesthesia Awareness Registry’’. Br. J. Anaesth. 2013. 110(3): 381-387. 7. C.K.E. Moll, A. Sharott, W. Hamel, A. Mu¨nchau, C. Buhmann, U. Hidding, S. Zittel, M. Westphal, D. Mu¨ller, A.K. Engel. ‘‘Waking Up the Brain: A Case Study of Stimulation-Induced Wakeful Unawareness During Anaesthesia’’. Prog. Brain Res. 2009. 177: 125-145. 8. U. Norinder, P. Sjoberg, T. Osterberg. ‘‘Theoretical Calculation and Prediction of Brain-Blood Partitioning of Organic Solutes Using Molsurf Parametrization and PLS Statistics’’. J. Pharm. Sci. 1998. 87(8): 952-959. 9. M.S. Burch, R.K. McAllister, T.A. Meyer. ‘‘Treatment of LocalAnesthetic Toxicity with Lipid Emulsion Therapy’’. Am. J. Health Syst. Pharm. 2011. 68(2): 125-129. 10. Q.C. Meng, H. Zou, J.S. Johansson, R.G. Eckenhoff. ‘‘Determination of the Hydrophobicity of Local Anesthetic Agents’’. Anal. Biochem. 2001. 292(1): 102-106. 11. A. Pohorille, M.A. Wilson, M.H. New, C. Chipot. ‘‘Concentrations of Anesthetics Across the Water-Membrane Interface; the MeyerOverton Hypothesis Revisited’’. Toxicol. Lett. 1998. 101: 421-430. 12. M. Weinrich, T.K. Rostovtseva, S.M. Bezrukov. ‘‘Lipid-Dependent Effects of Halothane on Gramicidin Channel Kinetics: A New Role for Lipid Packing Stress’’. Biochemistry. 2009. 48(24): 5501-5503. 13. C.C. Peng, M.T. Burke, A. Chauhan. ‘‘Transport of Topical Anesthetics in Vitamin E Loaded Silicone Hydrogel Contact Lenses’’. Langmuir. 2012. 28(2): 1478-1487. 14. H. Lord, J. Pawliszyn. ‘‘Introduction to SPME for In Vivo Analysis’’. LC GC Eur. 2012. 25(4): 180. 15. E.L. Difilippo, R.P. Eganhouse. ‘‘Assessment of PDMS-Water Partition Coefficients: Implications for Passive Environmental Sampling of Hydrophobic Organic Compounds’’. Environ. Sci. Technol. 2010. 44(18): 6917-6925. 16. D.L. Heglund, D.C. Tilotta. ‘‘Determination of Volatile Organic Compounds in Water by Solid Phase Microextraction and Infrared Spectroscopy’’. Environ. Sci. Technol. 1996. 30(4): 1212-1219. 17. I.C. Nwaneshiudu, Q.M. Yu, D.T. Schwartz. ‘‘Quantitative SolidPhase Microextraction (SPME)-Raman Spectroscopy for the Detection of Trace Organics in Water’’. Appl. Spectrosc. 2012. 66(12): 1487-1491. 18. D.C. Stahl, D.C. Tilotta. ‘‘Screening Method for Nitroaromatic Compounds in Water Based on Solid-Phase Microextraction and Infrared Spectroscopy’’. Environ. Sci. Technol. 2001. 35(17): 35073512. 19. D. Vuckovic, X. Zhang, E. Cudjoe, J. Pawliszyn. ‘‘Solid-Phase Microextraction in Bioanalysis: New Devices and Directions’’. J. Chromatogr. A. 2010. 1217(25): 4041-4060. 20. S. Risticevic, H. Lord, T. Gorecki, C.L. Arthur, J. Pawliszyn. ‘‘Protocol for Solid-Phase Microextraction Method Development’’. Nat. Protoc. 2010. 5(1): 122-139. 21. M.J. Jager, D.P. McClintic, D.C. Tilotta. ‘‘Measurement of Petroleum Fuel Contamination in Water by Solid-Phase Microextraction with Direct Raman Spectroscopic Detection’’. Appl. Spectrosc. 2000. 54(11): 1617-1623. 22. H. Kataoka. ‘‘Current Developments and Future Trends in SolidPhase Microextraction Techniques for Pharmaceutical and Biomedical Analyses’’. Anal. Sci. 2011. 27(9): 893-905. 23. J.J. Langenfeld, S.B. Hawthorne, D.J. Miller. ‘‘Quantitative Analysis of Fuel-Related Hydrocarbons in Surface Water and Wastewater Samples by Solid-Phase Microextraction’’. Anal. Chem. 1996. 68(1): 144-155.
24. W.T. Stringfellow, K.C. Oh. ‘‘Comparison of SPME Headspace Analysis to US EPA Method 5030/8260B for MTBE Monitoring’’. Ground Water Monit. Rem. 2005. 25(2): 52-58. 25. D. Lawson, S. Samanta, P.T. Magee, D.E. Gregonis. ‘‘Stability and Long-Term Durability of Raman-Spectroscopy’’. J. Clin. Monitor. 1993. 9(4): 241-251. 26. H. Kamaya, S. Kaneshina, I. Ueda. ‘‘Partition Equilibrium of Inhalation Anesthetics and Alcohols Between Water and Membranes of Phospholipids with Varying Acyl Chain Lengths’’. Biochim. Biophys. Acta. 1981. 646(1): 135-142. 27. S.A. Simon, T.J. McIntosh, P.B. Bennett, B.B. Shrivastav. ‘‘Interaction of Halothane with Lipid Bilayers’’. Mol. Pharmacol. 1979. 16(1): 163-170.
28. P. Mayer, J. Tolls, L. Hermens, D. Mackay. ‘‘Equilibrium Sampling Devices’’. Environ. Sci. Technol. 2003. 37(9): 184A-191A. 29. X.Y. Cui, P. Mayer, J. Gan. ‘‘Methods to Assess Bioavailability of Hydrophobic Organic Contaminants: Principles, Operations, and Limitations’’. Environ. Pollut. 2013. 172: 223-234. 30. J. Ohmori, S. Maeda, H. Higuchi, M. Ishii, Y. Arai, Y. Tomoyasu, A. Kohjitani, M. Shimada, T. Miyawaki. ‘‘Propofol Increases the Rate of Albumin-Unbound Free Midazolam in Serum Albumin Solution’’. J. Anesth. 2011. 25(4): 618-620. 31. R.L. Zhou, J.M. Perez-Aguilar, Q.C. Meng, J.G. Saven, R.Y. Liu. ‘‘Opioid Binding Sites in Human Serum Albumin’’. Anesth. Analg. 2012. 114(1): 122-128.
APPLIED SPECTROSCOPY
1259
University of Washington, Department of Chemical Engineering, Box 351750, Seattle, WA 98195, USA University of Washington, Department of Anesthesiology, Box 356540, Seattle, WA 98195, USA
Polydimethylsiloxane (PDMS)-based solid-phase micro-extraction (SPME) was used along with Raman spectroscopy (RS) to separate and enhance the detection of five anesthetic compounds (halothane, propofol, isoflurane, enflurane, and etomidate) from aqueous and serum phases. Raman signals in the spectral ranges 250–450 cm1 and 950–1050 cm1 allowed the unique characterization of all five compounds when extracted into the PDMS phase. The SPME-RS detection of clinically relevant concentrations of aqueous propofol (6.5 lM) and halothane (200 lM) is shown. We quantify the partition coefficient for aqueous halothane in PDMS as log K ¼ 1.9 6 0.2. Solid-phase micro-extraction of the anesthetics makes their detection possible without the strong autofluorescent interference of serum proteins. Because of low solubility and/or weak Raman scattering, we found it challenging to detect enflurane, isoflurane, and etomidate directly from the aqueous phase, but could we do so with SPME enhancement. These studies show the potential of SPME-RS as a method for the direct detection of anesthetics in blood. Index Headings: Anesthetics; Solid-phase micro-extraction; SPME; Partition; Propofol.
INTRODUCTION Clinical anesthesiology is continually adapting to new and improved anesthetic compounds.1 The transition from volatile compounds like halothane toward such anesthetics as isoflurane and sevoflurane is largely responsible for the reduced number of anesthesiarelated complications.1–3 In addition, the use of fastacting intravenous anesthetics such as propofol, ketamine, and etomidate has also improved clinical practice.1,2 However, a continuing challenge in the field is accounting for the distribution of these anesthetics in the body.4,5 Although overall estimates can be made using the detection of inhaled and exhaled gas concentrations, the actual concentrations in the body remain elusive due to the partitioning of these compounds into lipid-like phases in the blood and tissue.4 This has led to complications during and after surgical procedures.5–7 Ways to directly measure concentrations in relevant biological fluids may be beneficial to researchers and clinicians. The partitioning of volatile and intravenous anesthetics between physiological phases such as blood and brain, alveolar gas and blood, and blood and lipid has Received 7 November 2013; accepted 2 May 2014. * Author to whom correspondence should be sent. E-mail: dts@u. washington.edu. DOI: 10.1366/13-07362
1254
Volume 68, Number 11, 2014
been well documented.8–13 The magnitude of partitioning is directly estimated by the equilibrium partition coefficient (K), the thermodynamic parameter that quantitatively relates the analyte concentration in each phase. Although physiological partitioning can complicate administration and detection of anesthetics,9 the same partitioning process can be used to enhance the detection of analytes via the method solid-phase microextraction (SPME).14 Solid-phase micro-extraction is a well-known analytical technique that uses the hydrophobic traits of polymers to extract and pre-concentrate nonpolar analytes for better detection.14–20 The use of polydimethylsiloxane (PDMS), a staple biocompatible polymer for SPME analytics, has been well established for liquid as well as gas phase detection. Techniques that benefit from SPME pre-concentration are well documented in the literature.16,21–24 In addition to signal enhancement, SPME has the potential to separate macromolecules that may interfere with the detection of smaller target analytes. This sort of separation could prove very useful when dealing with analytes in a complex system such as blood or serum. Solid-phase micro-extraction coupled with Raman spectroscopy (SPME-RS), is emerging as a quantitative analytical method, showing improved detection of one to three orders of magnitude over normal Raman spectroscopy (RS).17,21 In the past, Raman scattering-based devices such as the RASCAL (Raman scattering analyzer) were used in hospitals for monitoring gas-phase anesthetics during surgery.25 New opportunities may exist for combining SPME-RS with anesthesiology in both research and clinical practice. Here we explore using PDMS-based SPME to enhance RS detection directly in relevant aqueous and serum liquid phases. Three volatile anesthetic compounds (isoflurane, enflurane, and halothane) and two intravenous anesthetics (propofol and etomidate) were examined for their enhanced detectability using SPME polymer phases in solution.
MATERIALS AND METHODS Many of the SPME method details are provided in Nwaneshiudu et al. 17 We summarize key features briefly next. Polydimethylsiloxane and Organic/Aqueous Solution Preparation. Polydimethylsiloxane Samples. Slygard 184 PDMS and catalyst were obtained (Dow Corning). Samples of PDMS used in the experiment
0003-7028/14/6811-1254/0 Q 2014 Society for Applied Spectroscopy
APPLIED SPECTROSCOPY
FIG. 1. Raman spectra of the neat anesthetic compounds. (a) Halothane. (b ) Propofol. (c) Isoflurane. (d ) Enflurane. (e) Etomidate. Multipliers next to spectra denote the scale used to make the spectral features visible on the same graph.
were prepared as follows. Uncured PDMS was mixed using the standard 10 :1 polymer-to-catalyst ratio and poured into 1.5 mm deep poly(methyl methacrylate) (PMMA) mold. This was cured at 70 8C overnight to make an 1.5 mm thick sheet of PDMS. Samples were punched out of the PDMS sheet using a 2 mm diameter punch. The resulting PDMS plugs were sealed and kept until ready for use. Organic/Aqueous Solutions. Halothane, propofol, isoflurane, enflurane, and etomidate were all purchased (Sigma Aldrich), and all were used as purchased. Pooled human serum was purchased (Biochemed Services, Winchester, VA). Aqueous solutions were prepared by adding the appropriate amount of reagent to 20 ml deionized water. Serum samples used 2 ml aliquots. The measuring was done volumetrically for liquid anesthetics (halothane, propofol, isoflurane, and enflurane). The solutions were then shaken vigorously, tightly capped, and left to fully dissolve. The spiked solutions were used within 15 min of preparation to prevent evaporation loses. Solutions of etomidate were prepared by introducing solid etomidate into 20 ml deionized water and letting the solution equilibrate for 72 h. These solutions were filtered before use to separate out undissolved solids from the saturated solution. Solid-Phase Micro-Extraction–Raman Spectroscopy Procedure. Water Solid-Phase Micro-Extraction. We used 20 ml scintillation glass vials for the SPME procedure. Solutions of halothane, propofol, isoflurane, enflurane, and etomidate were prepared in separate 20 ml vials. The PDMS plugs were introduced into each vial; the vials were capped and left to equilibrate undisturbed at room temperature for at least 48 h. Serum Solid-Phase Micro-Extraction. We used glass vials (2 ml) for the SPME procedure. The vials were filled with serum and spiked with halothane and propofol to reach the desired concentrations. The PDMS plugs were introduced into each vial; the vials were capped and left to equilibrate undisturbed at room temperature for at least 48 h.
Raman Spectra Acquisition. Raman spectra were collected using an inVia Raman micro-spectrometer (Renishaw) attached to a DM IRBE upright optical microscope (Leica). A 785 nm diode laser operated at full power was used to irradiate samples through a 503 objective lens (numerical aperture (NA) 0.8). The spot area was 50 lm2. Raman scattered light was acquired through the same objective lens and detected using a thermoelectrically cooled (60 8C) charge-coupled detector. All PDMS sample spectral measurements were acquired for 10 s. Acquisition times for the water-sample measurements ranged from 10 to 300 s, depending on signal-to-noise concerns. The spectra of the neat compounds (halothane, isoflurane, enflurane, and propofol) were collected using a 103 objective, which was irradiated in a capped 2 mL glass vial. The spectra of the equilibrated PDMS and water samples were collected using a wet sample holder sealed with a glass coverslip. Laser stability was confirmed using the 520 cm1 peak of silicon. The PDMS peak at 1410 cm1 was acquired with all the solid-phase spectra and served as an internal standard that helped normalize the sample-to-sample variations in the system focus. Raman Scattering Intensity Normalization and Peak Analysis. The spectral peaks were analyzed using WiRE 2.0 software (Windows-based Raman Environment; Renishaw) and fit to standard Voigt distribution profiles. The curve fit parameters were used to calculate the integrated peak areas. The integrated peak areas of the PDMS internal standard were used to normalize the reported analyte peak areas. To account for acquisitiontime differences, peak intensities are reported in counts per second.
RESULTS AND DISCUSSION Figure 1 shows the characteristic Raman spectra of all five of these anesthetic compounds (neat) in the 200– 1090 cm1 range. These compounds represent diverse anesthetics, and over this range, there are unique spectral features for these compounds that make RS effective for fingerprinting. Halothane is the most strongly scattering and etomidate is the weakest scattering of the compounds tested. Halothane is used to illustrate the quantitative aspect of SPME partitioning (methods detailed in prior work17). We examined the C–Cl bending mode of halothane at 310 cm1 in the spectral ranges 285–350 cm1. These peaks do not overlap with the PDMS or the water peaks. Figure 2 shows the Raman spectral peaks for the C–Cl bending mode of halothane at 310 cm1 taken in the equilibrated PDMS (Fig. 2, top) and aqueous (Fig. 2, bottom) phases. The baseline subtracted spectral data points are presented along with the corresponding curve fits and scale bars to show the intensity of the Raman signal in counts per second. Comparing detection in PDMS to that in water, we can see an 50-fold peak-intensity enhancement. Because of the low halothane signal in water, the aqueous spectra at the concentrations 2 and 0.2 mM are not shown in Fig. 2. This figure illustrates that SPME-RS enables the straightforward detection of halothane at concentrations that are clinically relevant (0.25–0.50 mM).12
APPLIED SPECTROSCOPY
1255
FIG. 2. Raman spectra of the 310 cm1 halothane peak. (top) Acquired in the PDMS phase. (bottom) Acquired in the water phase. Lines a–e indicate aqueous concentrations of 20, 15, 10, 2, and 0.2 mM, respectively. Data points are raw data, and lines are the Voigt curve fits.
Quantifying Partitioning. Figure 3 shows a plot of the 310 cm1 halothane peak area as a function of the aqueous halothane concentration. Peaks areas were normalized to the PDMS internal standard peak at 1410 cm1, eliminating small variations in PDMS sample preparation. Data points represent triplicate measurements performed on two separate samples and were fit using linear regression analysis. The linearity of this plot suggests we are in the dilute range where there is negligible polymer swelling and solute–solute interactions are modest. The normalized peak areas in Fig. 3 are proportional to the analyte concentration, Raman cross section of the analyte in each media, and instrumental factors. If we assume the halothane scattering cross section and instrumental factors are identical in water and PDMS (assumptions that have proven reasonable for other SPME–Raman scattering analytes17) then a simple relationship for the equilibrium partition coefficient of halothane (Kh) emerges: Kh ¼
SlopePDMS;h SlopeWater;h
ð1Þ
where SlopePDMS,h is the slope of the best-fit line for halothane measured in the PDMS phase and SlopeWater,h is the slope of the best-fit line measured in the aqueous phase. The partition coefficient estimate from the data in Fig. 3 is log Kh = 1.9 6 0.2. Although no halothane partitioning values for the PDMS or water systems could be found in the literature, others reported values of log Kh between 2.0 and 3.0 for aqueous–lipid partitioning.26,27
1256
Volume 68, Number 11, 2014
FIG. 3. Integrated areas of the 310 cm1 halothane peak versus aqueous halothane concentrations. (top) Acquired in the PDMS phase. (bottom) Acquired in the water phase.
Because PDMS is a hydrophobic polymer known to display lipid-like partitioning, our estimated values appears reasonable.28,29 Qualitative Detection. Next, we examine propofol, a commonly used intravenous anesthetic. Unlike halothane, which has a relatively high water solubility (20 mM), the water solubility of propofol is lower (700 lM). At these concentrations and using reasonable acquisition times ( 300 s), we could not easily detect propofol in the aqueous phase. Concentrations of administered propofol typically range from 1 to 60 lM in blood.4 Using SPME, we could detect propofol readily at clinically relevant concentrations using 10 s acquisition times. Figure 4a shows spectra of the characteristic 1046 cm1 aromatic peak of propofol taken in the equilibrated PDMS. All spectra were acquired with 10 s of exposure, which highlights the improved signal-to noise ratio of the SPME approach. As expected, Fig. 4b shows that the partitioning of propofol behaves linearly at these dilute concentrations. Fingerprinting using RS is challenging for complex multicomponent samples such as blood and urine. The SPME pre-concentration has the potential to selectively separate and concentrate the desired analyte, thereby avoiding interference from other components. As a simple proof of concept, we examined a mixture of 10 mM halothane and 60 lM propofol in human pool serum. Figure 5a shows a Raman spectrum obtained directly in serum, and Fig. 5b shows a typical spectrum in PDMS after the serum mixture was equilibrated with the polymer. Figure 5a displays the broad autofluorescence and overlapping protein spectra that we expect from a complex aqueous protein mixture such as serum. None of
FIG. 4. (panel 1) Spectra a–e acquired in the PDMS phase at aqueous propofol concentrations of 650, 500, 350, 65, and 35 lM, respectively. (panel 2) Representative integrated peak areas of the spectra plotted against aqueous propofol concentration.
the peaks from halothane or propofol is discernible. In contrast, the spectrum acquired from the equilibrated PDMS phase has none of the protein-related interference. Therefore, we are able to resolve some of the anesthetic
FIG. 5. (a) Spectra for a mixture of 10 mM halothane and 60 lM propofol in serum. (b) Spectra for a mixture of 10 mM halothane and 60 lM propofol in equilibrated PDMS. The inset in (b) is the baseline subtracted region where halothane peaks are present.
peaks (box and inset of Fig. 5b). However, it is interesting to note that no propofol peaks appear in the PDMS that was equilibrated with serum. Based on the earlier aqueous propofol experiments, we would expect to see propofol peaks at the concentrations used here. The absence of propofol peaks at such high concentrations brings up an interesting aspect of propofol biochemistry; namely, that propofol is known to complex with proteins in serum.4,30,31 The protein–propofol complexation competes with PDMS for the accumulation of propofol, reducing the SPME effect. The quantitative traits of this competitive partitioning warrants further study. Last, we illustrate the improved detection possible using the SPME of isoflurane, enflurane, and etomidate. These anesthetic compounds gave the weakest Raman signals, as shown in Fig. 1. Although the inherently weak signals from these compounds make them undetectable in water, when they were equilibrated with a PDMS phase, distinct signature peaks could be seen. The main spectrum in Fig. 6 (gray) shows the characteristic peaks of PDMS, which hide any anesthetic peaks on this scale. The spectral regions where these anesthetic peaks reside are 250–450 cm1 (Fig. 6, inset 1) and 950–1050 cm1 (Fig. 6, inset 2). The Raman spectra labeled a and b in Fig. 6 (inset 1) are of 20 mM isoflurane and enflurane, respectively; spectrum c is from PDMS alone. The plots labeled (ac) and (bc) are difference spectra that show the presence of characteristic anesthetic peaks when the PDMS background is subtracted. Figure 6 (inset 2) shows the spectral range for identification of etomidate signals. Spectrum d is of PDMS equilibrated with aqueous etomidate; spectrum e is of PDMS. A characteristic 1005 cm1etomidate peak neighboring a PDMS background peak at 995 cm1 can be seen in spectrum d. The difference spectrum (de) in Fig. 6 (inset 2) shows the weak, characteristic 1005 cm1 etomidate peak without the PDMS peak. In Fig. 6 (inset 1), four distinct isoflurane peaks (280, 340, 370, and 420 cm1) can be seen in difference spectrum (ac) and three enflurane peaks (265,
APPLIED SPECTROSCOPY
1257
FIG. 6. The PDMS (gray) spectrum with black boxes indicating the spectral ranges used to identify isoflurane, enflurane, and etomidate. (Inset 1) Spectral region 250–450 cm1. Spectrum a is of PDMS equilibrated in 20 mM isoflurane; spectrum b is of PDMS equilibrated in 20 mM enflurane; spectrum c is of PDMS. The difference spectra (ac) and (bc) are presented as bold curves. (Inset 2) Spectral region 950– 1050 cm1. Spectrum d is of PDMS equilibrated in aqueous etomidate; spectrum e is of PDMS. The difference spectrum (de) is shown as a bold curve.
360, and 430 cm1) can be seen in difference spectrum (bc). As we can see, SPME does enhance the Raman signals of these anesthetics.
CONCLUSION AND IMPLICATIONS We have shown that SPME improves the Raman signals of all the anesthetics examined in this study. With halothane and propofol, which displayed the strongest Raman signals, SPME lets us detect clinically relevant concentration (200 and 35 lM, respectively) with modest acquisition times (10 s). We quantified the partitioning of halothane into PDMS and obtained a range of partition coefficients (log K = 1.9 6 0.2) for the system. Finally, signals for isoflurane, enflurane, and etomidate that were undetectable in the water phase could be discerned in a PDMS background under similar conditions using SPME. Showing that SPME-RS can detect anesthetics in both single-component aqueous mixtures and complex biological fluids such as serum displays its versatility in detecting in a variety of systems. In addition, because SPME is a pre-concentration technique that does not rely on a particular detection system, it can complement several different analytical platforms. This could open doors to the direct detection of these compounds in biological fluids using a variety of optically based systems. ACKNOWLEDGMENTS We gratefully acknowledge the support of the Gates Foundation Millennium Scholarship as well as Andrew Keefe for providing human pooled serum for the multicomponent experiments. Partial support was also provided by the Boeing Sutter endowment. 1. D.H. Robinson, A.H. Toledo. ‘‘Historical Development of Modern Anesthesia’’. J. Invest. Surg. 2012. 25(3): 141-149.
1258
Volume 68, Number 11, 2014
2. R.D. Urman, S.P. Desai. ‘‘History of Anesthesia for Ambulatory Surgery’’. Curr. Opin. Anaesthesiol. 2012. 25(6): 641-647. 3. A. Dabbagh, S. Rajaei. ‘‘Halothane: Is There Still Any Place for Using the Gas as an Anesthetic?’’ Hepat. Mon. 2011. 11(7): 511-512. 4. F. Kivlehan, F. Garay, J.D. Guo, E. Chaum, E. Lindner. ‘‘Toward Feedback-Controlled Anesthesia: Voltammetric Measurement of Propofol (2, 6-Diisopropylphenol) in Serum-Like Electrolyte Solutions’’. Anal. Chem. 2012. 84(18): 7670-7676. 5. M. Grossherr, A. Hengstenberg, T. Meier, L. Dibbelt, B.W. Igl, A. Ziegler, P. Schmucker, H. Gehring. ‘‘Propofol Concentration in Exhaled Air and Arterial Plasma in Mechanically Ventilated Patients Undergoing Cardiac Surgery’’. Br. J. Anaesth. 2009. 102(5): 608-613. 6. C.D. Kent, G.A. Mashour, N.A. Metzger, K.L. Posner, K.B. Domino. ‘‘Psychological Impact of Unexpected Explicit Recall of Events Occurring During Surgery Performed Under Sedation, Regional Anaesthesia, and General Anaesthesia: Data from the Anesthesia Awareness Registry’’. Br. J. Anaesth. 2013. 110(3): 381-387. 7. C.K.E. Moll, A. Sharott, W. Hamel, A. Mu¨nchau, C. Buhmann, U. Hidding, S. Zittel, M. Westphal, D. Mu¨ller, A.K. Engel. ‘‘Waking Up the Brain: A Case Study of Stimulation-Induced Wakeful Unawareness During Anaesthesia’’. Prog. Brain Res. 2009. 177: 125-145. 8. U. Norinder, P. Sjoberg, T. Osterberg. ‘‘Theoretical Calculation and Prediction of Brain-Blood Partitioning of Organic Solutes Using Molsurf Parametrization and PLS Statistics’’. J. Pharm. Sci. 1998. 87(8): 952-959. 9. M.S. Burch, R.K. McAllister, T.A. Meyer. ‘‘Treatment of LocalAnesthetic Toxicity with Lipid Emulsion Therapy’’. Am. J. Health Syst. Pharm. 2011. 68(2): 125-129. 10. Q.C. Meng, H. Zou, J.S. Johansson, R.G. Eckenhoff. ‘‘Determination of the Hydrophobicity of Local Anesthetic Agents’’. Anal. Biochem. 2001. 292(1): 102-106. 11. A. Pohorille, M.A. Wilson, M.H. New, C. Chipot. ‘‘Concentrations of Anesthetics Across the Water-Membrane Interface; the MeyerOverton Hypothesis Revisited’’. Toxicol. Lett. 1998. 101: 421-430. 12. M. Weinrich, T.K. Rostovtseva, S.M. Bezrukov. ‘‘Lipid-Dependent Effects of Halothane on Gramicidin Channel Kinetics: A New Role for Lipid Packing Stress’’. Biochemistry. 2009. 48(24): 5501-5503. 13. C.C. Peng, M.T. Burke, A. Chauhan. ‘‘Transport of Topical Anesthetics in Vitamin E Loaded Silicone Hydrogel Contact Lenses’’. Langmuir. 2012. 28(2): 1478-1487. 14. H. Lord, J. Pawliszyn. ‘‘Introduction to SPME for In Vivo Analysis’’. LC GC Eur. 2012. 25(4): 180. 15. E.L. Difilippo, R.P. Eganhouse. ‘‘Assessment of PDMS-Water Partition Coefficients: Implications for Passive Environmental Sampling of Hydrophobic Organic Compounds’’. Environ. Sci. Technol. 2010. 44(18): 6917-6925. 16. D.L. Heglund, D.C. Tilotta. ‘‘Determination of Volatile Organic Compounds in Water by Solid Phase Microextraction and Infrared Spectroscopy’’. Environ. Sci. Technol. 1996. 30(4): 1212-1219. 17. I.C. Nwaneshiudu, Q.M. Yu, D.T. Schwartz. ‘‘Quantitative SolidPhase Microextraction (SPME)-Raman Spectroscopy for the Detection of Trace Organics in Water’’. Appl. Spectrosc. 2012. 66(12): 1487-1491. 18. D.C. Stahl, D.C. Tilotta. ‘‘Screening Method for Nitroaromatic Compounds in Water Based on Solid-Phase Microextraction and Infrared Spectroscopy’’. Environ. Sci. Technol. 2001. 35(17): 35073512. 19. D. Vuckovic, X. Zhang, E. Cudjoe, J. Pawliszyn. ‘‘Solid-Phase Microextraction in Bioanalysis: New Devices and Directions’’. J. Chromatogr. A. 2010. 1217(25): 4041-4060. 20. S. Risticevic, H. Lord, T. Gorecki, C.L. Arthur, J. Pawliszyn. ‘‘Protocol for Solid-Phase Microextraction Method Development’’. Nat. Protoc. 2010. 5(1): 122-139. 21. M.J. Jager, D.P. McClintic, D.C. Tilotta. ‘‘Measurement of Petroleum Fuel Contamination in Water by Solid-Phase Microextraction with Direct Raman Spectroscopic Detection’’. Appl. Spectrosc. 2000. 54(11): 1617-1623. 22. H. Kataoka. ‘‘Current Developments and Future Trends in SolidPhase Microextraction Techniques for Pharmaceutical and Biomedical Analyses’’. Anal. Sci. 2011. 27(9): 893-905. 23. J.J. Langenfeld, S.B. Hawthorne, D.J. Miller. ‘‘Quantitative Analysis of Fuel-Related Hydrocarbons in Surface Water and Wastewater Samples by Solid-Phase Microextraction’’. Anal. Chem. 1996. 68(1): 144-155.
24. W.T. Stringfellow, K.C. Oh. ‘‘Comparison of SPME Headspace Analysis to US EPA Method 5030/8260B for MTBE Monitoring’’. Ground Water Monit. Rem. 2005. 25(2): 52-58. 25. D. Lawson, S. Samanta, P.T. Magee, D.E. Gregonis. ‘‘Stability and Long-Term Durability of Raman-Spectroscopy’’. J. Clin. Monitor. 1993. 9(4): 241-251. 26. H. Kamaya, S. Kaneshina, I. Ueda. ‘‘Partition Equilibrium of Inhalation Anesthetics and Alcohols Between Water and Membranes of Phospholipids with Varying Acyl Chain Lengths’’. Biochim. Biophys. Acta. 1981. 646(1): 135-142. 27. S.A. Simon, T.J. McIntosh, P.B. Bennett, B.B. Shrivastav. ‘‘Interaction of Halothane with Lipid Bilayers’’. Mol. Pharmacol. 1979. 16(1): 163-170.
28. P. Mayer, J. Tolls, L. Hermens, D. Mackay. ‘‘Equilibrium Sampling Devices’’. Environ. Sci. Technol. 2003. 37(9): 184A-191A. 29. X.Y. Cui, P. Mayer, J. Gan. ‘‘Methods to Assess Bioavailability of Hydrophobic Organic Contaminants: Principles, Operations, and Limitations’’. Environ. Pollut. 2013. 172: 223-234. 30. J. Ohmori, S. Maeda, H. Higuchi, M. Ishii, Y. Arai, Y. Tomoyasu, A. Kohjitani, M. Shimada, T. Miyawaki. ‘‘Propofol Increases the Rate of Albumin-Unbound Free Midazolam in Serum Albumin Solution’’. J. Anesth. 2011. 25(4): 618-620. 31. R.L. Zhou, J.M. Perez-Aguilar, Q.C. Meng, J.G. Saven, R.Y. Liu. ‘‘Opioid Binding Sites in Human Serum Albumin’’. Anesth. Analg. 2012. 114(1): 122-128.
APPLIED SPECTROSCOPY
1259
