Fast, low-pressure gas chromatography triple quadrupole tandem mass spectrometry for analysis of 150 pesticide residues in fruits and vegetables

We developed and evaluated a new method of low-pressure gas chromatography-tandem mass spectrometry (LP-GC/MS-MS) using a triple quadrupole instrument for fast analysis of 150 relevant pesticides in four representative fruits and vegetables. This LP-GC (vacuum outlet) approach entails coupling a 10 m, 0.53 mm i.d., 1 μm film analytical column between the MS transfer line and a 3 m, 0.15 mm i.d. capillary at the inlet. The MS creates a vacuum in the 10 m analytical column, which reduces the viscosity of the He carrier gas and thereby shifts the optimal flow rate to greater velocity. By taking advantage of the H(2)-like properties of He under vacuum, the short analytical column, a rapid oven temperature ramp rate, and the high selectivity and sensitivity of MS/MS, 150 pesticides were separated in <6.5 min. The 2.5 ms dwell time and 1 ms interscan delay of the MS/MS instrument were critical for achieving >8 data points across the 2-3 s wide peaks. To keep dwell and cycle times constant across all peaks, each segment consisted of 30 analytes (60 transitions). For assessment, we injected extracts of spiked broccoli, cantaloupe, lemon, and sweet potato from the updated QuEChERS sample preparation method. Average recoveries (n=72) were 70-120% for 144 of the pesticides, and reproducibilities were <20% RSD for all but 4 analytes. Also, detection limits were <5 ng/g for all but a few pesticides, depending on the matrix. In addition to high quality performance, the method gave excellent reliability and high sample throughput, including easy peak integration to obtain rapid results.     

Evaluation of low-pressure gas chromatography linked to ion-trap tandem mass spectrometry for the fast trace analysis of multiclass pesticide residues

A rapid multiresidue method for the analysis of 72 pesticides has been developed using a single injection with low-pressure gas chromatography/tandem mass spectrometry (LP-GC/MS/MS). The LP-GC/MS/MS method used a short capillary column of 10 m x 0.53 mm i.d. x 0.25 microm film thickness coupled with a 0.6 m x 0.10 mm i.d. restriction at the inlet end. Optimal LP-GC conditions were determined which achieved the fastest separation in MS/MS detection mode. Also MS/MS conditions were optimized in order to increase sensitivity and selectivity. The analytical parameters of the LP-GC/MS/MS method were compared with those obtained by GC/MS/MS using a conventional capillary column (30 m x 0.25 mm i.d. x 0.25 microm film thickness). Better precision and sensitivity values were obtained with the LP-GC/MS/MS approach. The limits of detection (LOD) of the compounds ranged from 0.1 to 14.1 microg L(-1) for LP-GC/MS/MS, lower than those obtained for conventional GC/MS/MS that ranged from 0.1 to 17.5 microg L(-1). The peak widths obtained with the short column in LP-GC are similar to those obtained using conventional capillary GC columns, and the peaks can be successfully identified by MS/MS detection with the conventional scan speed of ion-trap instruments. In addition, the analysis time was significantly reduced with LP-GC/MS/MS (32 min) versus GC/MS/MS (72 min), allowing the number of samples analyzed per day in a routine laboratory to be doubled.     

Fast and sensitive determination of pesticide residues in vegetables using low-pressure gas chromatography with a triple quadrupole mass spectrometer

In this study, the feasibility of low-pressure gas chromatography (LP-GC) in conjunction with a triple quadrupole mass spectrometer, as a route towards fast pesticide residue analysis, was investigated. A Varian GC-MS system equipped with a mass spectrometer model 1200 was used. LP-GC-MS experiments were performed on a HP-5 10 m x 0.32 mm x 0.25 microm analytical column connected to a 2.5 m x 0.15 mm non-coated restriction precolumn at the inlet end. For comparison purposes conventional GC-MS analysis was performed on a RTX-5 30 m x 0.25 mm x 0.5 microm column. Under the optimized conditions the analysis time was reduced to 13.3 min with the LP-GC approach which corresponds to an almost threefold gain in speed versus the conventional GC (37 min). Despite the poorer separation power of the LP-GC column, the experiments conducted with tomato and onion extracts spiked with 78 pesticides proved that LP-GC-MS is of practical value to perform full scan screening analysis. Moreover, the rate of false negative results was higher in the case of conventional GC-MS while the LP-GC-MS enabled correct identification of pesticides at lower levels since the peaks were improved in both size and shape. Validation experiments were performed on a sample of 12 representative pesticides for comparison of performance characteristics of the LP-GC and GC approaches with mass spectrometer operated in scan, SIM and MS/MS mode. The LP-GC column set-up interfaced to the MS detector was found to be superior to the conventional GC with respect to obtained linearity, accuracy and precision parameters. Also, lower limits of detection in real extracts were achieved using the LP-GC approach. Finally, the LP-GC-MS/MS analysis of tomato samples with incurred pesticide residues demonstrated the applicability of the developed method for analysis of real samples.     

Sequential injection sample introduction microfluidic-chip based capillary electrophoresis system

A sequential injection micro-sample introduction system was coupled to a microfluidic-chip based capillary electrophoresis system through a split–flow sampling interface integrated on the micro-chip. The microfluidic system measured 20×70×3 mm in dimension, and was produced using a non-lithographic approach with components readily available in the analytical laboratory. In the H-configuration channel design the horizontal separation channel was a 75 μm I.D.×60 mm quartz capillary, with two vertical side arms produced from plastic tubing. The conduits were embedded in silicon elastomer with a planar glass base. Sequential introduction of a series of samples with about 2.5% carryover was achieved at 48 h⁻¹ throughput with samples containing a mixture of fluorescein isothiocyanate (FITC)-labeled amino acids using SI sample volumes of 3.3 μl and carrier flow-rate of 2.0 ml min⁻¹. Baseline separation was achieved for FITC-labeled arginine, phenylalanine, glycine and FITC (laser induced fluorescence detection) in sodium tetraborate buffer (pH 9.2) within 8–80 s, at separation lengths of 25–35 mm and electrical field strengths of 250–1500 V cm⁻¹, with plate heights in the 0.7–3 μm range.     

Fast Temperature Programming in Gas Chromatography using Resistive Heating

The features of a resistive-heated capillary column for fast temperature-programmed gas chromatography (GC) have been evaluated. Experiments were carried out using a commercial available EZ Flash GC, an assembly which can be used to upgrade existing gas chromatographs. The capillary column is placed inside a metal tube which can be heated, and cooled, much more rapidly than any conventional GC oven. The EZ Flash assembly can generate temperature ramps up to 1200°/min and can be cooled down from 300 to 50°C in 30 s. Samples were injected via a conventional split/splitless injector and transferred to the GC column. The combination of a short column (5 m×0.25 mm i. d.), a high gas flow rate (up to 10 mL/min), and fast temperature programmes typically decreased analysis times from 30 min to about 2.5 min. Both the split and splitless injection mode could be used. With n-alkanes as test analytes, the standard deviations of the retention times with respect to the peak width were less than 15% (n = 7). First results on RSDs of peak areas of less than 3% for all but one n-alkane indicate that the technique can also be used for quantification. The combined use of a short GC column and fast temperature gradients does cause some loss of separation efficiency, but the approach is ideally suited for fast screening as illustrated for polycyclic aromatic hydrocarbons, organophosphorus pesticides, and triazine herbicides as test compounds. Total analysis times – which included injection, separation, and equilibration to initial conditions – were typically less than 3 min.     

Measurement of PCDDs, PCDFs, and non-ortho-PCBs by comprehensive two-dimensional gas chromatography-isotope dilution time-of-flight mass spectrometry (GC x GC-IDTOFMS)

Comprehensive two-dimensional gas chromatography with isotope-dilution time-of-flight mass spectrometry (GC × GC-IDTOFMS) was used to measure polychlorinated dibenzo-p-dioxin (PCDD), polychlorinated dibenzofuran (PCDF), and coplanar polychlorinated biphenyl (cPCB) concentrations in ash, sediment, vegetation, and fish samples. The GC × GC capability was achieved by using a quad jet, dual stage, thermal modulator. Zone compression of the GC peaks from modulation resulted in a significant increase of the signal intensity over classical GC-IDTOFMS. The GC × GC column set used an Rtx-Dioxin 2 phase as the first dimension (¹D) and an Rtx-500 as the second dimension (²D). The chromatographic separation of the 17 PCDD/Fs and the 4 cPCBs was attained in ¹D except for 2,3,7,8-TCDD and CB126 for which deconvoluted ion currents (DIC) were required to be reported separately. The Rtx-500 phase separated the bulk matrix interfering compounds from the target analytes in ²D. The instrumental limit of detection (iLODs) was 0.5 pg for 2,3,7,8-TCDD. The calibration curves showed good correlation coefficients for all the compounds investigated in the concentration range of 0.5–200 pg. GC × GC-IDTOFMS results compared favorably to those from conventional isotope-dilution one-dimensional gas chromatography-high resolution mass spectrometry (GC-IDHRMS). The comprehensive mass analysis of the TOFMS further permitted the identification of other contaminants of concern in the samples.     

Three-phase contactor with distributed U-shaped bundles of hollow-fibers for pertraction

Pertraction of dimethylcyclopropanecarboxylic acid (DMCCA) and phenol in a new hollow-fiber (HF) contactor of three liquid phases with distributed U-shaped bundles of microporous HF was studied. Trioctylamine (TOA), Hostarex A327 and Cyanex 471X have been used as carriers. The performance of this contactor is similar to other HF contactors. Its advantage is that both bundles of fibers can elongate without deformation of fibers and maldistribution of fibers in the bundles. Pulsation of the membrane phase increases the transport rate by 35–61% and it reaches a plateau value at the pulsation velocity of about 1.1 mm s⁻¹. With increasing velocity of the feed in the fiber lumen the value of the overall mass-transfer coefficient increases, but soon a steady value is approached at the velocity of about 5.5 cm s⁻¹. The diffusion resistances in a series model do not describe the pertraction of DMCCA well. The suggested new diffusion-reaction model, considering the resistance based on the kinetics of the stripping reaction, fits experimental data well. Not only the rate of the acid–carrier complex decomposition, but also that of the DMCCA dimer decomposition have to be considered. The estimated values of the apparent reaction rate constant were for the membrane with TOA 1.91×10⁻⁶ m s⁻¹ and for pure n-alkanes 5.78×10⁻⁶ m s⁻¹. The kinetic resistance on the stripping interface represents 20–60% of the overall mass-transfer resistance and its value is comparable to the diffusion resistance of the both walls.     

On-line coupling of sequential injection extraction with restricted-access materials and post-column derivatization for sample clean-up and determination of propranolol in human plasma

The presented paper deals with a new methodology for direct determination of propranolol in human plasma. The methodology described is based on sequential injection analysis technique (SIA) coupled with solid phase extraction (SPE) column based on restricted access materials (RAM). Special RAM column containing 30 μm polymeric material—N-vinylacetamide copolymer was integrated into the sequential injection manifold. SIA–RAM system was used for selective retention of propranolol, while the plasma matrix components were eluted with two weak organic solutions to waste.

Due to the acid–basic and polarity properties of propranolol molecule and principles of reversed-phase chromatography, it was possible to retain propranolol on the N-vinylacetamide copolymer sorbent (Shodex MSpak PK-2A 30 μm (2 mm × 10 mm)). Centrifuged plasma samples were aspirated into the system and loaded onto the column using acetonitrile–water (5:95, v/v), pH 11.00, adjusted by triethylamine. The analyte was retained on the column while proteins contained in the sample were removed to waste. Interfering endogenous substances complicating detection were washed out by acetonitrile–water (15:85), pH 11.00 in the next step. The extracted analyte was eluted by means of tetrahydrofuran–water (25:75), pH 11.00 to the fluorescence detector (emission filter 385 nm). The whole procedure comprising sample pre-treatment, analyte detection and column reconditioning took about 15 min. The recoveries of propranolol from undiluted plasma were in the range 96.2–97.8% for three concentration levels of analyte. The proposed SIA–RAM method has been applied for direct determination of propranolol in human plasma.     

Measurement of binary diffusion coefficient from impulse response curve having extremely low absorbance intensity under supercritical condition by noise elimination technique

A noise elimination technique was applied to the determination of binary diffusion coefficients D₁₂ from the response curves having extremely low absorbance intensities in impulse response methods under supercritical conditions of carbon dioxide. The effectiveness of this technique was experimentally examined for the analyses of response curves through both the curve-fitting and the moment methods in two cases: the chromatographic impulse response method for phenol and β-carotene with a polymer-coated capillary column, and the Taylor dispersion method for acetone with an uncoated capillary column. Unreliable D₁₂ values were obtained from the moment method of the response curves at lower absorbance intensities, even treated with noise elimination. The curve-fitting method with the noise elimination treatment was quite effective for determining the D₁₂ values accurately, and was valid at the lowest absorbance intensities, on the order of 10⁻⁴ absorbance unit of UV-Vis multi-detector, corresponding to the smallest quantity of the solute, i.e. 6×10⁻⁵, 6×10⁻⁶, and 5×10⁻² μ mol for phenol, β-carotene, and acetone, respectively, under conditions studied. Infinite dilution regions for binary diffusion coefficients were obtained by injecting various amounts: the binary diffusion coefficients showed constant values at concentrations less than 0.6, 0.004, and 0.08 mol m⁻³ for phenol, β-carotene, and acetone, respectively, in supercritical carbon dioxide at 313.2 K and 16–18 MPa.     

Sensitive absorption detection for micro-column liquid chromatography by ultraviolet forward-scattering degenerate four-wave mixing

The performance of forward-scattering degenerate four-wave mixing (F-D4WM) in the mid-ultraviolet (UV) region (351 nm) as a detection technique for micro-column liquid chromatography (μLC) is studied, using nitro-substituted polycyclic aromatic hydrocarbons (NO₂-PAHs) and amino-substituted anthraquinones (AAQs) as test compounds. When using round capillaries, the concentration limits of detection (LOD) for the NO₂-PAHs were 20–70-fold better compared with absorption detection using a U-shaped detector cell. The final result is influenced by photochemical degradation during 351 nm F-D4WM detection. Furthermore, it is demonstrated that the use of recently commercially available square capillaries (i.e., capillaries with both a square cross section and a square bore) instead of round ones is quite suitable for F-D4WM detection. The square capillary (internal dimensions, 75×75 μm²; external dimensions 300×300 μm²) did not cause significant chromatographic band broadening in μLC. The angle of incidence of the laser light should be 56°, thus fulfilling the Brewster condition at the air–quartz boundary. Using this approach for the AAQs, compounds not prone to significant photodegradation under the experimental conditions, a further 4-fold improvement was achieved. As a result, the overall gain in concentration LOD for 2-amino-9,10-anthraquinone (molar absorptivity 4000 M⁻¹ cm⁻¹ at 351 nm) was from 4 to 0.05 μM, and baseline irregulations were reduced.     

Rapid pesticide analysis,in post-harvest plants used as animal feed,by low-pressure gas chromatography–tandem mass spectrometry

A wide range of pesticides used to control pests in vegetables have been determined in agricultural plant waste from beans, watermelons, and melons grown in greenhouses located in a predominantly agricultural area in Southeast Spain (Almería). Analysis of the pesticides was carried out by low-pressure gas chromatography (LP-GC) with mass spectrometry in tandem (MS–MS) mode, after extraction of the lyophilized samples with dichloromethane. The influence of the sample matrix on the analysis was avoided by use of matrix-matched standards. Linearity, detection limit (LOD), quantitation limit (LOQ), recovery, and precision for each pesticide were calculated. The most frequently encountered pesticides were endosulfan (>73% of the analyzed samples) and buprofezin (>55% of the samples), followed by cypermethrin, pirimifos-methyl, bifentrin, and chlorpyrifos (>30% of the samples). The pesticide found at the highest concentration level was endosulfan (223.33 mg kg^–1) in a watermelon sample.     

Carbon ion irradiation induced surface modification of polypropylene

Polypropylene was irradiated with ¹²C ions of 3.6 and 5.4 MeV energies in the fluence range of 5×10¹³–5×10¹⁴ ions/cm² using 3 MV tandem accelerator. Ion penetration was limited to a few microns and surface modifications were investigated by scanning electron microscopy. At the lowest ion fluence only blister formation of various sizes (1–6 μm) were observed, but at higher fluence (1×10¹⁴ ions/cm²) a three-dimensional network structure was found to form. A gradual degradation in the network structure was observed with further increase in the ion fluence. The dose dependence of the changes on surface morphology of polypropylene is discussed.     

High-performance liquid chromatographic stationary phases based on poly(methyloctylsiloxane) immobilized on silica. II. Chromatographic evaluation

An integrated on-line SPE–HPLC–MS/MS system has been developed for the rapid analysis of various trace level priority pesticides in surface and drinking water. Eleven pesticides were included in this study, with various phenylureas, triazines and organophosphorous species among them. Use of turbulent-flow chromatography columns (TFC, 50×1 mm, 30–50 μm particle size) as extraction cartridges enables fast on-line SPE at high sampling flow-rate (5 ml/min). Polymeric and carbon based TFC columns (Oasis HLB, Cyclone, Hypercarb) allow complete extraction with good recoveries from water volumes up to 50 ml. On-line coupling to HPLC is performed with re-mixing of the organic TFC eluate with water in front of the analytical column to ensure efficient band focussing. For fast HPLC analysis, a short monolithic column is applied in combination with highly selective API–MS/MS detection. Matrix effects on the APCI–MS/MS signal were found to be reduced by the system to an acceptable minimum. Limits of detection, determined for 10-ml samples of river water were in the range between 0.4 and 13 ng/l typically, except trifluralin (approximately 280 ng/l), which is less susceptible to ionization under atmospheric pressure conditions. At an enriched water volume of 10 ml, the whole SPE–HPLC–MS/MS procedure requires less than 14 min. The method was successfully applied to the analysis of drinking and surface water samples taken from several sampling sites around the city of Leipzig, Germany. Concentrations measured (maximum: 16 ng/l simazine in river water) were far below the concentration limits scheduled by law.     

Flow-injection in-line complexation for ion-pair reversed phase high performance liquid chromatography of some metal-4-(2-pyridylazo) resorcinol chelates

Flow injection (FI) was coupled to ion-pair reversed phase high performance liquid chromatography (IP-RPHPLC) for the simultaneous analysis of some metal-4-(2-pyridylazo) resorcinol (PAR) chelates. A simple reverse flow injection (rFI) set-up was used for in-line complexation of metal-PAR chelates prior to their separation by IP-RPHPLC. The rFI conditions were: injection volume of PAR 85 μL, flow rate of metal stream 4.5 mL min⁻¹, concentration of PAR 1.8 × 10⁻⁴ mol L⁻¹ and the mixing coil length of 150 cm. IP-RPHPLC was carried out using a C₁₈ μBondapak column with the mobile phase containing 37% acetonitrile, 3.0 mmol L⁻¹ acetate buffer pH 6.0 and 6.2 mmol L⁻¹ tetrabutylammonium bromide (TBABr) at a flow rate of 1.0 mL min⁻¹ and visible detection at 530 and 440 nm. The analysis cycle including in-line complexation and separation by IP-RPHPLC was 16 min, which able to separate Cr(VI) and the PAR chelates of Co(II), Ni(II) and Cu(II).     

Calorimetric comparison of the interactions between salivary proteins and Streptococcus mutans with and without antigen I/II

Antigen I/II can be found on streptococcal cell surfaces and is involved in their interaction with salivary proteins. In this paper, we determine the adsorption enthalpies of salivary proteins to Streptococcus mutans LT11 and S. mutans IB03987 with and without antigen I/II, respectively, using isothermal titration calorimetry. In addition, protein adsorption to the cell surfaces was determined spectrophotometrically. S. mutans LT11 with antigen I/II, yielded a much higher, exothermic adsorption enthalpy at pH 6.8 (ranging from −2073 × 10⁻⁹ to −31707 × 10⁻⁹ μJ per bacterium) when mixed with saliva than did S. mutans IB03987 (−165 × 10⁻⁹ to −1107 × 10⁻⁹ μJ per bacterium) at all bacterial concentrations studied (5 × 10⁹, 5 × 10⁸, and 5 × 10⁷ ml⁻¹), largest effects per bacterium being observed for the lowest concentration. However, the enthalpy of salivary protein adsorption to S. mutans LT11 became smaller at pH 5.8. Adsorption isotherms for the S. mutans LT11 showed considerable protein adsorption at pH 6.8 (1.2–2.1 mg/m²), that decreased only slightly at pH 5.8 (1.1–1.6 mg/m²), with the largest amount adsorbed at the lowest bacterial concentration. This suggests that the protein(s) in the saliva with the strongest affinity for antigen I/II is (are) readily depleted from saliva. In conclusion, antigen I/II surface proteins on S. mutans play a determinant role in adsorption of salivary proteins through the creation of enthalpically favorable adsorption sites.     

Automatic in-tube SPME and fast liquid chromatography: a cost-effective method for the estimation of dibuthyl and di-2-ethylhexyl phthalates in environmental water samples

A 80-cm length commercially available capillary coated with 95% polydimethylsiloxane and 5% polydiphenylsiloxane (TBR-5) was employed to carry out on-line extraction and preconcentration of dibuthyl phthalate (DBP) and di-2-ethylhexyl phthalate (DEHP) in the chromatographic system. The coated capillary was placed between the sample injection loop and the injection needle of an autosampler. Variables affecting the automatic in-tube solid-phase microextraction (SPME) were optimized. A Genesis C₁₈ (5 cm × 4.6 mm i.d., 4 μm particle size) was employed as analytical column. The achieved limits of detection by use of diode array detection were 1 and 2.5 μg L⁻¹, respectively. The proposed conditions have been applied to determine those compounds at low ppb levels (≤250 μg L⁻¹) in aqueous samples. No matrix effect was found, and recoveries between 85 and 115% were obtained. The precision of the method was good, and the achieved intra- and inter-day variation coefficients were between 5 and 20%. The analysis time per sample was 20 min and any off-line pre-treatment of the samples was needed. The taken sample volume was 100 μL. Data on the application of the described method to the analysis of different water samples are presented.     

Study of the electron kinetic processes in laser-induced breakdown of electronegative gases over an extended wavelength range

A theoretical investigation of laser-induced breakdown of electronegative gases is presented. The formulations are based on an electron cascade model previously developed by Evans and Gamal (J. Phys. D: Appl. Phys. 13 (1980) 1447–1458). This model solves numerically the time-dependent Boltzmann equation simultaneously with a set of rate equations describing the population density of the formed excited states. It includes the possible kinds of interactions between electrons, molecules and photons. Calculations are carried out under the experimental conditions of Davis et al. (Appl. Optics 30 (1991) 4358–4364) where molecular oxygen over a pressure range of 20–760 Torr is irradiated with the first four harmonics of a Nd : YAG laser source at wavelengths 1064, 532, 355 and 266 nm of pulse duration 8.5, 7.5, 6.5 and 5.5 ns, respectively, and peak irradiance varies between 3.6×10¹⁰ and 3.7×10¹¹ W/cm². Computations revealed that the dependence of threshold irradiance on gas pressure is in quite close agreement with those measured by Davis et al. (1991) for the four laser wavelengths considered in this analysis. It is also shown that at laser wavelengths 532 and 266 nm oxygen breakdown is mainly governed by the combined effect of two and three-body attachment loss processes, while for the wavelengths 335and 1064 nm, the three-body attachment process dominates. In addition, calculation of the electron energy distribution function and its parameters, viz, electron density, ionization rate and electron mean energy predicted the importance of the photoionization of excited states as the main electron generation process over the short wavelength range.     

Unauthorised antibiotic treatments in beekeeping: Development and validation of a method to quantify and confirm tylosin residues in honey using liquid chromatography–tandem mass spectrometric detection

A liquid chromatographic–tandem mass spectrometric (HPLC-MS/MS) method is proposed for the identification and quantification of tylosin in honey. Sample treatment involves an extraction in a Tris buffer at pH 10.5, followed by a solid-phase clean up step on an Oasis HLB column. Roxithromycin was used as the internal standard. Chromatographic separation of tylosin and roxithromycin was performed on an XTerra MS C₁₈ column (100 mm × 2.1 mm i.d., 5 μm) using a gradient of aqueous 0.01 M ammonium acetate pH 3.5 and acetonitrile as the mobile phase, at a flow rate of 0.25 ml min⁻¹. The method was validated according to the guidelines laid down by the Commission Decision 2002/657/EC. Tylosin residues were confirmed by MS/MS experiments considering the appropriate identification points. All validation parameters such as Cc (lower than 3 ng g⁻¹), Ccβ (lower than 5 ng g⁻¹), recovery and precision were assessed on the basis of the “critical ion” (less intense ion permitting unambiguous identification of the analyte).     

Multiple enzyme linked immunosorbent assay system on a capillary-assembled microchip integrating valving and immuno-reaction functions

Gastrodin is a bioactive constituent of rhizome in Gastrodia elata Blume (Orchidaceae) The aim of this study is to develop a rapid and sensitive liquid chromatographic method coupled to microdialysis sampling system to measure the unbound of gastrodin in rat blood, brain and bile. Microdialysis probes were simultaneously inserted into the jugular vein, brain striatum and bile duct of each anesthetized rat for sampling after the administration of gastrodin (100 or 300 mg kg⁻¹) through the femoral vein. Separation of unbound gastrodin from various biological fluids was applied to an RP-select B column (250 mm × 4.6 mm i.d., 5 μm). The mobile phase consisted of acetonitrile–50 mM potassium dihydrogen phosphate buffer–triethylamine (5:95:0.1, v/v/v, adjusted to pH 2.5 with orthophosphoric acid) with a flow rate of 1 mL min⁻¹. The UV detector wavelength was set at 221 nm. Fifteen minutes after the administration, the gastrodin reached the peak concentration in brain and bile. In addition, the results indicate that gastrodin penetrates the blood-brain barrier (BBB) and goes through hepatobiliary excretion.