ARTICLE pubs.acs.org/ac
Optimization of the Coating Procedure for a High-Throughput 96-Blade Solid Phase Microextraction System Coupled with LCMS/MS for Analysis of Complex Samples Fatemeh S. Mirnaghi,† Yong Chen,‡ Leonard M. Sidisky,‡ and Janusz Pawliszyn*,† † ‡
Department of Chemistry, University of Waterloo, 200 University Avenue West, Waterloo, Ontario N2L 3G1, Canada Supelco Inc., 595 North Harrison Road, Bellefonte, Pennsylvania 16823, United States
bS Supporting Information ABSTRACT: Biocompatible C18-polyacrylonitrile (PAN) coating was used as the extraction phase for an automated 96-blade solid phase microextraction (SPME) system with thin-film geometry. Three different methods of coating preparation (dipping, brush painting, and spraying) were evaluated; the spraying method was optimum in terms of its stability and reusability. The highthroughput sample preparation was achieved by using a robotic autosampler that enabled simultaneous preparation of 96 samples in 96-well-plate format. The increased volume of the extraction phase of the C18-PAN thin film coating resulted in significant enhancement in the extraction recovery when compared with that of the C18-PAN rod fibers. Various factors, such as reusability, reproducibility, pH stability, and reliability of the coating were evaluated. The results showed that the C18-PAN 96-blade SPME coating presented good extraction recovery, long-term reusability, good reproducibility, and biocompatibility. The limits of detection and quantitation were in the ranges of 0.10.3 and 0.51 ng/mL for all four analytes.
I
n order to improve the sample preparation strategies for the fast analysis of complex samples, there has been a growing need for automated high-throughput analysis in different areas of clinical, pharmaceutical, food, and environmental sciences.112 The automation facilitates demanding and time-consuming sample preparation steps and improves precision. Sample throughput can be improved by increasing the sampling rate or analyzing several samples simultaneously.13 Automation of solid-phase extraction (SPE) has raised dramatically during the last 2 decades; the introduction of automated 96-well-plate SPE in 1996 provided significant improvements in the parallel sample preparation techniques.1416 Since then, various automatic 96-well-format workstations and different chemistries of solid phase extraction material have been commercialized.1725 However, 384 and 1536 well-plates are also available to further improve sample throughput,13,26,27 the 96-well format are commonly used for parallel sample preparation. The success of the 96-well format SPE has also led to the application of the format to other sample preparation techniques such as liquidliquid extraction (LLE) which had usually been considered to be difficult to automate.2730 Recently, the automation of solid phase microextraction (SPME) in a 96-well-plate format was also successfully attained in order to simultaneously prepare up to 96 samples via full automation prior to liquid chromatography (LC)tandem mass spectrometry (MS/MS) analysis.13,3134 Xie et al. reported on an automated in-tip fiber SPME approach in 96-well format coupled with LCMS/MS for high-throughput clinical application. In this research, SPME fibers were fitted in the tips of pipettes and a r 2011 American Chemical Society
commercially available 96 workstation was used to aspirate and dispense the sample solution and desorption solvent to the pipet tips. The in-tip fiber SPME system has the limitation to be applied for the complex and viscous samples; this is largely due to the narrow opening of the pipet tips and difficulties in expelling all sample residues. As a result, in-tip fibers should mainly be considered disposables in order to prevent carryover/contamination issues.34 Additional studies from the Pawliszyn group reported an in-house multifiber device enabling the simultaneous utilization of 96-coated SPME fiber and application of an automated robotic unit for direct extraction of drugs from biological fluids.32,3537 For example, Vuckovic et al. reported on the development and application of an automated C18-coated 96-fiber system for whole blood analysis, without any need for sample pretreatment.32 A focus in the development of the SPME method is the improvement of the extraction capacity and, consequently, overall method sensitivity. According to the fundamental principles of SPME (eq 1), the amount of analyte extracted by SPME is proportional to the volume of the extraction phase (Vf): n¼
Kfss Vf Vs C0 Kfs Vf þ Vs
ð1Þ
where n represents the amount of analyte extracted at equilibrium, Kfs represents the distribution constant between the extraction Received: April 20, 2011 Accepted: June 28, 2011 Published: June 28, 2011 6018
dx.doi.org/10.1021/ac2010185 | Anal. Chem. 2011, 83, 6018–6025
Analytical Chemistry phase and the sample matrix, Vf represents the volume of extraction phase, Vs represents the volume of the sample, and C0 represents the original concentration of the analyte.38 Increasing the volume of the extraction phase can be accomplished by either increasing the thickness of the stationary phase coating36,39 or increasing the surface area.40 The use of thicker coatings increases the extraction equilibrium time leading to lower sample throughput.41 Increasing the surface area can be achieved by increasing the diameter of the SPME fiber or using thin-film geometry. A large increase in fiber diameter would require more volume in the sample well and cause displacement of the sample solution, thus limiting the maximum sample volume that could be placed in the wells.36 In contrast, using thin-film geometry occupies less space and more sample volume can be used. In addition, thin film-geometry provides much more effective agitation and improved extraction recovery without sacrificing time. Furthermore, the initial rate of SPME extraction is proportional to the surface area of the extraction phase, and it provides another important benefit of thin-film microextraction versus traditional SPME.42 The previous work in the Pawliszyn research group demonstrated the thin-film microextraction theoretical concept by immobilizing a layer of C18 coating on a thin metal surface obtained by flattening one end of a stainless steel rod. The report showed that an increase in the surface area of the SPME coating resulted in significant improvement of extraction recovery compared to the rod fiber configuration; however, long-term stability and reusability of the coating were not reported.36 The current study is an improvement upon previous work due to the application of a commercial 96-blade device with thin-film geometry. This device is designed to overcome the limitations of the in-house multifiber rod configuration and the difficulties in manually flattening fibers to make a thin-film. Since the stationary phase of the 96-blade SPME system is coated on the outer surface of the blades, it provides direct open-bed extraction for complex matrixes (including dense fluids and colloidal suspensions) without any need for sample pretreatment or limitation of clogging or contamination that is common in the conventional packed bed systems. The robotic 96-autosampler is designed to directly place the 96-blade device into the 96-wellplates prefilled with sample/solvents, enabling the performance of preconditioning, extraction, washing, and desorption steps. The main focus of the current work is the development and evaluation of a biocompatible and long-lasting SPME coating for the 96-blade SPME system which can be used in complex matrixes with high reproducibility and efficiency for several times. Most of the current commercial SPME coatings present drawbacks, including instability and stripping of the coating after long-term handling, limited extraction recovery, lack of compatibility with complex matrixes, and high costs.43 The adhesion of macromolecules, such as particulates and proteins, to the coating surface can significantly influence the kinetics of extraction and the amount of analyte extracted by the coating.44 One of the biocompatible polymers widely used in the biomedical area such as dialysis and ultrafiltration is polyacrylonitirile (PAN).45,46 In addition to its biocompatible characteristic, PAN provides high chemical and mechanical stability, making it ideal for use as a binder for stationary phase immobilization. A previous study demonstrated the utilization of PAN as a binder for the preparation of different biocompatible SPME coatings using the dipping method for the coating preparation. They proved the biocompatibility of the coatings by X-ray photoelectron spectroscopy.47
ARTICLE
Since the current study sought strategies for preparation of the biocompatible coating with high stability for the automated 96blade SPME system, PAN was chosen as the binder for immobilizing the stationary phase on the stainless steel surface. Also octadecyl-silica based stationary phase (C18) was chosen for the extraction phase because of its good extraction efficiency and applicability for a large variety of compounds. The current work also focused on evaluation of different procedures for coating preparation in order to get the optimum coating for the C18PAN 96-blade SPME system. Dipping, brush painting, and spraying methods were evaluated, and the spraying method showed the highest stability and reproducibility when compared to the dipping and brush painting methods. The performance of the sprayed C18-PAN 96-blade SPME system, coupled with the LCMS/MS method, was evaluated for high-throughput analysis of benzodiazepines from phosphate-buffered saline solution (PBS) and human plasma.
’ EXPERIMENTAL SECTION Chemicals and Materials. Sodium chloride, potassium chloride, potassium phosphate monobasic, sodium phosphate dibasic, polyacrylonitrile, and a flask-type sprayer were purchased from Sigma-Aldrich (MO). Diazepam, lorazepam, oxazepam, nordiazepam, and diazepam-d5 (internal standard) were purchased from Cerilliant (TX) as a 1 mg/mL standard in methanol with the exception of lorazepam which was in acetonitrile. Working standards were prepared from these stock solutions using acetonitrile/water (50:50 v/v) as the diluents and stored at 4 °C in a refrigerator. Acetonitrile (HPLC grade), methanol (HPLC grade), and N,N-dimethylformamide (DMF) were purchased from Caledon Laboratories (ON, Canada). Human plasma was purchase from Lampire Biological Laboratories (PA). Discovery silica-based-C18 particles (5 μm) were obtained from Supelco (PA). Polypropylene Nunc U96 Deep Well plates were purchased from VWR International (ON, Canada). The preparation of the phosphate-buffered saline solution is described in the Supporting Information. Preparation of C18-PAN 96-Blade SPME Coating. Prior to the coating process, the blades were sonicated with concentrated hydrochloric acid for about 60 min in order to condition the stainless steel surface for effective immobilization of the coating. The blades were then washed thoroughly and rinsed with nanopure water. Next, they were dried in an oven for 30 min at 150 °C and then cooled to room temperature. The blades were coated using three different methods of spraying, dipping, and brush painting, and in all three cases the C18 particles were immobilized on the surface of stainless steel blades using biocompatible PAN glue. The previous study showed that combining 10% w/w PAN particles with DMF solvent resulted in the optimum properties of the required glue.47 Since PAN does not dissolve at room temperature, the mixture was heated in the oven at 90 °C for about 1 h, until a yellowish clear solution was obtained. The mixture was then cooled to room temperature and 5 μm C18 particles (20% of the total volume) were added. All three coating were made using the same slurry of the C18PAN mixture. In the case of preparation of the coating with the spraying method, a flask type sprayer (250 mL Erlenmeyer flask with a sprayer head) was used for spraying the slurry on the blade’s surface. The mixture was transferred into the flask-type sprayer, and the source of nitrogen gas was connected to the sprayer head to provide the required pressure for spraying. 6019
dx.doi.org/10.1021/ac2010185 |Anal. Chem. 2011, 83, 6018–6025
Analytical Chemistry
ARTICLE
Figure 1. Concept 96-blade SPME device coated with C18-PAN coating: (a) Concept 96-autosampler including the following labels (b) (A, C, and D) orbital agitators for extraction, desorption, and conditioning, respectively, (B) wash station, (E) 96-blade device, (F) nitrogen evaporation device, and (G) syringe arm.
The coating preparation was performed by spraying very thin layers of C18-PAN slurry on the first 2 cm length of the blades followed by instant thermal curing in the oven at 180 °C for 2 min. For the preparation of coatings with the method of dipping, the first 2 cm portions of the blades were dipped into the coating slurry and then slowly removed out. Next, the coatings were dried immediately at 180 °C for 2 min. In the case of brush paint coating, a thin layer of the coating slurry were spread onto the first 2 cm length of the blades with a fine painting brush whose width was approximately the same as that of the blades and the coating on the blades was dried with the same condition as the other two coatings. For all three types, the coating and curing steps were repeated 10 times in order to ensure uniform coverage and proper thickness of the C18-PAN coating on the surface of the blades. High-Pressure Liquid Chromatography and Mass Spectrometry Conditions. A Shimadzu (LC-10 AD) high-pressure liquid chromatography (HPLC) and an Applied Biosystems API 3000 triple quadrupole mass spectrometer (equipped with TurboIonSpray source) were used for separation and quantitative analyses of compounds. A Waters Symmetry Shield RP18 with dimensions of 2.1 mm 50 mm and 5 μm particles was used as the chromatographic column. Chromatographic conditions used for the separation of the benzodiazepines are discussed elsewhere.31 A sample volume of 20 μL of both the standard and extracted analytes was injected into an LCMS/MS system using a CTC PAL autoinjector from Leap Technologies (CTC Analytics, NC). In the first minute of the chromatographic run time, a bypass pump and a Waters switching valve were used to divert the flow of column effluent into the waste. The MS/MS analysis was performed in the positive mode under multiple reaction monitoring (MRM) conditions. The summary of MS/MS parameters is given in the Supporting Information, Supporting Table S-1. Automated Concept 96-Blade SPME System. The Concept 96-blade SPME device (Professional Analytical System (PAS) Technology, Magdala, Germany) is composed of eight rows of blade sets made from 1.4310 grade stainless steel; each blade set consists of 12 thin-film pins (length, 50 mm; width, 2.5 mm; depth, 0.7 mm). The eight rows of the blade sets are held together using nine interblade holders, creating a 96-blade SPME system that fits in a 96-well-plate and works as a part of an autosampler (Figure 1a). The current study uses the robotic Concept 96-autosampler (PAS Technology, Magdala, Germany) that contains three integrated arms and three separate orbital agitators, all of which are fully controlled by the Concept software. One arm is designed to automatically hold, move, and place the
96-blade SPME device into the 96-well-plates, enabling the performance of the steps of preconditioning, extraction, washing, and desorption. The orbital agitators are allocated to shake the 96-well-plates at a specific speed controlled with Concept software. In cases when enhanced sensitivity is important and reconstitutions and/or preconcentration are required, the second arm is equipped with a nitrogen blow-down device in order to perform solvent evaporation and analyte preconcentration. Finally, the third arm is equipped with a syringe that dispenses a specific volume of the solvent into the individual wells of the 96well-plate. This arm can also be used for sample injection into the HPLC port for chromatographic separation and analysis. Figure 1a,b shows the C18-PAN coated Concept 96-blade device and the Concept 96-autosampler from PAS Technology. Automated SPME Procedure for High-Throughput Analysis. The SPME experimental steps of this work included method development and optimization of the SPME procedure for benzodiazepines (diazepam, lorazepam, oxazepam, nordiazepam) analysis in spiked PBS (adjusted to pH = 7.4) and human plasma. After spiking the analytes for plasma analysis, samples were stored in a refrigerator overnight to ensure proper incubation and to reach drug-plasma protein binding equilibrium. In order to prevent disturbing the equilibrium between the analytes and biofluid matrix, the organic solvents in the final spiked samples were always kept
