Determination of perfluorooctane sulfonate and perfluorooctanoic acid in human plasma by large volume injection capillary column switching liquid chromatography coupled to electrospray ionization mass spectrometry

Rapid, selective, and sensitive methodology for the quantification of perfluorooctanoic acid (PFOA) and perfluorooctane sulfonate (PFOS) in human plasma using packed capillary liquid chromatography coupled to electrospray ionization ion-trap mass spectrometry has been developed. Plasma proteins were precipitated using acetonitrile and the resulting supernatant was diluted 1+1 with water containing 10 mM ammonium acetate (NH4Ac) prior to injection. Sample volumes of 250 microL were loaded onto a 30 mm x 0.32 mm ID 10 microm Kromasil C18 precolumn by a carrier solution consisting of 10 mM NH4Ac in ACN/H2O (5/95, v/v) at a flow rate of 100 microL/min, providing on-line analyte enrichment and sample clean-up. Backflushed elution onto a 100 mm x 0.32 mm ID 3.5 microm Kromasil C18 analytical column was conducted using an ACN/H2O solvent gradient containing 10 mM NH4Ac. In order to improve the robustness and performance of the method, perfluoroheptanoic acid (PFHA) was used as internal standard. Separation and detection of PFOA, PFHA, and PFOS were achieved within 10 minutes. Ionization was performed in the negative mode in the m/z range 250-550. The method was validated over the concentration range 1-200 ng/mL for PFOA and over the range 5-200 ng/mL untreated plasma for PFOS, yielding correlation coefficients of 0.997 (PFOA) and 0.996 (PFOS), respectively. The within-assay (n = 6) and between-assay (n = 6) precisions were in the range 2.1-9.2 and 5.6-12%, respectively. The concentration limits of detection (cLOD) of PFOA was 0.5 ng/mL while the cLOD of PFOS was estimated to be 0.2 ng/mL in untreated plasma.     

Online trace enrichment to determine pyrethroids in river water by HPLC with column switching and photochemical induced fluorescence detection

The potential of online trace enrichment on a highly apolar short column in LC was evaluated for the determination of pyrethroids in river water. Twelve millilitres of water samples, modified with 8 mL ACN (ACN/water 40:60, v/v), were passed through 50 x 4.6 mm ID first separation column packed with 5 microm Hypersil Elite C18. Pesticides were preconcentrated in this column while the matrix background was eluted to waste. Separation of pesticides was performed on a 3.5 microm symmetric C18 column (250 x 4.6 mm ID) with an ACN step gradient as mobile phase and fluorescence detection was used after postcolumn derivatization by using UV light. The use of photochemically induced fluorescence for detection improved sensitivity and selectivity. Quantification limits ranged from 0.05 to 0.1 microg/L and pesticide recoveries at two concentration levels (0.1 and 0.5 microg/L) were between 93.1 and 118.6%, with RSD between 2.5 and 7.5% (n = 3) in river water samples. No matrix effect was detected.     

Temperature-promoted large-volume solute enrichment in column-switching miniaturized liquid chromatography: determination of an antioxidant

A two-valve sub-ambient temperature-promoted reversed-phase packed-capillary liquid-chromatography column-switching system has been tailored for sensitive determination of hydrophobic compounds. Such compounds are not easily dissolved in solvent mixtures of non-eluting properties that traditionally are used for solute enrichment in reversed-phase liquid chromatography. Enrichment-column solute focusing of large sample volumes was promoted by use of sub-ambient temperatures only, allowing the use of sample solvents that were stronger or equal to the mobile phase solvent strength. Subsequent column switching and enrichment-column temperature increment provided efficient low-dispersion back-flushed enrichment-column solute desorption onto the analytical column, where the solute was subjected to temperature-programmed gradient action. The antioxidant, Irganox 1076 (octadecyl 3-(3,5-di-tert-butyl-4-hydroxyphenyl)propionate) extracted from low density polyethylene with 100% acetonitrile served as a hydrophobic model compound. The mobile phase consisted of acetonitrile containing 10 mM triethylamine and formic acid, and the 0.25 mm id enrichment-column and analytical column in lengths of 27 and 250 mm, respectively, were packed with 3.5 microm Kromasil C18 particles. Sample volumes of up to 500 microL were successfully focused on the enrichment column at 5 degrees C using loading flow rates of up to 40 microL min(-1) prior to temperature programming to 90 degrees C. The concentration limit of detection of Irganox 1076 was 6 ng mL(-1) when using an injection volume of 500 microL. The within-assay precision was in the range 3.5-6.8% (n = 6) while the between-day precision was 7.5% (n = 3) relative standard deviation. The method was linear within the investigated mass range 3-100 ng (R2 = 0.9993).     

Comparing monolithic and microparticular capillary columns for the separation and analysis of peptide mixtures by liquid chromatography-mass spectrometry

A mixture of ten proteins was trypsinized and injected onto poly-(styrene-divinylben-zene) monolithic columns (60 x 0.20 or 0.10 mm ID) and a column packed with C18 silica particles (75 x 0.075 mm ID), respectively. The columns were eluted at 200-2000 nL/min with gradients of ACN in 0.050% TFA. Eluting peptides were detected by ESI-MS/MS and subsequently identified by database searching. The 100 microm ID monolithic column showed the highest cumulative Mowse scores based on the highest ion scores for the peptides and the largest number of identified peptides. It is shown that the number of identified peptides strongly depends on the dynamic range within the peptide mixture. In consequence, all proteins were identified in a mixture of relatively balanced analyte amounts (12.5-80 fmol) whereas only peptides for six out of ten proteins were found in a sample of high-dynamic range (0.65-270 fmol). The 100 microm monolithic column showed the highest reproducibility for peptide identifications in three consecutive runs. Depending on sample amount, 57-72% of the identified peptides were detectable in each of the three runs of triplicate analyses. The results demonstrate the high suitability of 100 microm monolithic columns for high-resolution peptide separations in proteomic research.     

Development of a validated HPLC method for the determination of four penicillin antibiotics in pharmaceuticals and human biological fluids

A quantitative method for the determination of four penicillin antibiotics, amoxicillin (AMO), oxacillin (OXA), cloxacillin (CLO), and dicloxacillin (DICLO), has been developed. Separation was achieved on an Inertsil ODS-3 (250 x 4 mm, 5 microm) column after selective extraction of penicillin drugs from biological matrices by means of SPE. Gradient elution with a mobile phase consisting of 0.1% TFA (pH 1) and ACN, and PDA detection with monitoring at 240 nm was applied. Salicylic acid (5 ng/microL) was used as the internal standard. RP-8 Adsorbex Merck cartridges provided high absolute recoveries (98-101%). The developed method was fully validated in terms of selectivity, linearity, accuracy, precision, stability, and sensitivity. Repeatability (n = 8) and between-day precision (n = 8) revealed RSD <10%. Recoveries from biological samples ranged from 91 to 103%. The detection limits were estimated as 3.3 ng for AMO, OXA, and CLO, and 6.6 for DICLO in blood plasma. LOD in whole blood and urine was 6.6 ng. Injection volume was 20 microL. The method was applied to commercially available AMO containing pharmaceuticals and spiked biological matrices. The method was also applied to biological samples after AMO oral administration, where the drug was successfully identified and quantified.     

On-line SPE-Nano-LC-Nanospray-MS for rapid and sensitive determination of perfluorooctanoic acid and perfluorooctane sulfonate in river water

An instrumental set up including on-line solid-phase extraction, nano-liquid chromatography, and nanospray mass spectrometry is constructed to improve the sensitivity for quantitation of perfluorooctanoic acid (PFOA) and perfluorooctane sulfonate (PFOS) in surface water. Sample volumes of 1000 microL are loaded onto a microbore 1.0-mm i.d. x 5 mm, 5 microm Kromasil C(18) enrichment column by a carrier solution consisting of 10mM ammonium acetate in acetonitrile-water (10:90, v/v) at a flow rate of 250 microL/min, providing on-line analyte enrichment and sample clean-up. Backflushed elution onto a 0.1-mm i.d. x 150 mm, 3.5 microm Kromasil C(18) analytical column is conducted using an acetonitrile-10mM ammonium acetate solvent gradient from 30% to 70% acetonitrile. Water samples are added with internal standard (perfluoroheptanoic acid) and filtrated prior to injection. The mass limits of detection of PFOA and PFOS are 0.5 and 1 pg, respectively, corresponding to concentration limits of detection of 500 pg/L and 1 ng/L, respectively. The total time spent on sample preparation, chromatography, and detection is approximately 12 min per sample. The method was employed for the determination of PFOS and PFOA in urban river water.     

Short communication performance of octadecylsilylated monolithic silica capillary columns of 530 microm inner diameter in HPLC

Monolithic silica capillary columns were successfully prepared in a fused silica capillary of 530 microm inner diameter and evaluated in HPLC after octadecylsilylation (ODS). Their efficiency and permeability were compared with those of columns pakked with 5-microm and 3-microm ODS-silica particles. The monolithic silica columns having different domain sizes (combined size of through-pore and skeleton) showed 2.5-4.0-times higher permeability (K= 5.2-8.4 x 10(-14) m2) than capillary columns packed with 3-mm particles, while giving similar column efficiency. The monolithic silica capillary columns gave a plate height of about 11-13 microm, or 11 200-13 400 theoretical plates/150 mm column length, in 80% methanol at a linear mobile phase velocity of 1.0 mm/s. The monolithic column having a smaller domain size showed higher column efficiency and higher pressure drop, although the monolithic column with a larger domain size showed better overall column performance, or smaller separation impedance (E value). The larger-diameter (530 microm id) monolithic silica capillary column afforded a good peak shape in gradient elution of proteins at a flow rate of up to 100 microL/min and an injection volume of up to 10 microL.     

Development and validation of an HPLC method for the determination of benzodiazepines and tricyclic antidepressants in biological fluids after sequential SPE

A novel method for the simultaneous determination of six benzodiazepines (BZDs) and four tricyclic antidepressants (TCAs) in biological fluids by HPLC with UV detection at 240 nm has been developed. After a deproteinization step biological fluids were analyzed by direct injection. SPE on Nexus cartridges was also applied. Since two compounds, namely imipramine and diazepam, were coeluting, a sequential SPE protocol has been developed. BZDs were eluted by a mixture of methanol/ACN(1:1), followed by the elution of TCAs with methanol. Separation was performed on a Kromasil C8 column (250 x 64 mm(2) id, 5 microm) using a mobile phase of 0.05 MCH3COONH4/ACN/methanol (initial composition 55:15:30 v/v/v) at a flow rate of 1.0 mL/min delivered by a gradient program within 15 min. Colchicine was used as the internal standard (4 ng/microL). The method was linear for all analytes up to 20 ng/lL, with coefficients of regression between 0.996 and 0.99996. LODs and LOQs were 0.08-1.17 and 0.28-3.91 ng/lL, respectively. Recovery was in the range of 92.8-108.7% for within-day and 91.9-109.9% for between-day assays, with RSD values lower than 10.0% for all matrices.     

A direct injection high-throughput liquid chromatography tandem mass spectrometry method for the determination of a new orally active alphavbeta3 antagonist in human urine and dialysate

A generic high-throughput liquid chromatography (HTLC) tandem mass spectrometry (MS/MS) assay for the determination of compound I in human urine and dialysate (hemodialysis) was developed and validated. By using the HTLC on-line extraction technique, sample pretreatment was not necessary. The sample was directly injected onto a narrow bore large particle size extraction column (50 x 1.0 mm, 60 microm) where the sample matrix was rapidly washed away using a high flow rate (5 mL/min) aqueous mobile phase while analytes were retained. The analytes were subsequently eluted from the extraction column onto an analytical column using an organic-enriched mobile phase prior to mass spectrometric detection. The analytes were then eluted from the analytical column to the mass spectrometer for the determination. The linear dynamic range was 2.0-6000 ng/mL for the urine assay and 0.1-300 ng/mL for the dialysate assay. Intraday accuracy and precision were evaluated by analyzing five replicates of calibration standards at all concentrations used to construct the standard curve. For the urine assay, the precision (RSD%, n=5) ranged from 1.9 to 8.0% and the accuracy ranged from 87.8 to 105.2% of nominal value. For the dialysate assay, the precision (RSD%, n=5) ranged from 1.1 to 10.0% and the accuracy from 94.5 to 105.2% of nominal value. In-source fragmentation of the acyl glucuronide metabolite (compound III) did not interfere with the determination of parent compound I. The developed HTLC/MS/MS methodology was specific for compound I in the presence of compound III. Column life-time is increased and sample analysis time is decreased over traditional reversed-phase methods when direct injection assays for urine and dialysate are coupled with the technology of HTLC.     

Determination of rotenone in river water utilizing packed capillary column switching liquid chromatography with UV and time-of-flight mass spectrometric detection

Fast and sensitive packed capillary column switching liquid chromatography methodology has been developed for the determination of the pesticide rotenone in river water. Sample volumes of up to 1 ml are loaded onto a 23 x 0.25 mm, 5 microm Kromasil C18 packed capillary precolumn using a noneluting solvent composition of water-acetonitrile (99:1, v/v) at flow-rates up to 100 microl/min prior to solute backflushing onto a 200 x 0.32 mm, 3.5 microm Kromasil C18 packed capillary analytical column using a mobile phase of water-acetonitrile (30:70, v/v) at a flow-rate of 5 microl/min. The method was evaluated using river water samples spiked with rotenone in the concentration range 0.5-50 ng/ml using UV detection. The within-assay precision was between 5.0 and 7.7% relative standard deviation (RSD, n = 6) and the between assay precision was between 7.5 and 8.9% RSD (n = 6). The method was linear within the investigated mass range displaying a calibration curve correlation factor of 0.997. The mass limit of detection was 10 pg corresponding to a concentration limit of detection of 10 pg/ml, using time-of-flight mass spectrometry.     

Direct determination of four fluoroquinolones,enoxacin, norfloxacin,ofloxacin, and ciprofloxacin,in pharmaceuticals and blood serum by HPLC

A rapid, accurate and sensitive method has been developed for the quantitative determination of four fluoroquinolone antimicrobial agents, enoxacin, norfloxacin, ofloxacin and ciprofloxacin, with high in-vitro activity against a wide range of Gram-negative and Gram-positive organisms.A Kromasil 100 C(8) 250 mm x 4 mm, 5 microm analytical column was used with an eluting system consisting of a mixture of CH(3)CN-CH(3)OH-citric acid 0.4 mol L(-1) (7:15:78 %, v/v). Detection was performed with a variable wavelength UV-visible detector at 275 nm resulting in limits of detection: 0.02 ng per 20 microL injection for enoxacin and 0.01 ng for ofloxacin, norfloxacin and ciprofloxacin. Hydrochlorothiazide (HCT) was used as internal standard at a concentration of 2 ng microL(-1). A rectilinear relationship was observed up to 2 ng microL(-1) for enoxacin, 12 ng microL(-1) for ofloxacin, 3 ng microL(-1) for norfloxacin, and 5 ng microL(-1) for ciprofloxacin. Separation was achieved within 10 min. The statistical evaluation of the method was examined by performing intra-day (n=8) and inter-day precision assays (n=8) and was found to be satisfactory with high accuracy and precision. The method was applied to the direct determination of the four fluoroquinolones in human blood serum. Sample pretreatment involved only protein precipitation with acetonitrile. Recovery of analytes in spiked samples was 97+/-6% over the range 0.1-0.5 ng microL(-1).     

Evaluation of several solid phase extraction liquid chromatography/tandem mass spectrometry on-line configurations for high-throughput analysis of acidic and basic drugs in rat plasma

Several configurations using 6- and 10-port switching valves were studied for high flow, on-line extraction of rat plasma coupled to an electrospray triple quadrupole mass spectrometer. Each plasma sample was diluted 1:1 with an aqueous internal standard solution. The sample was injected into a 2.1 x 20 mm cartridge column packed with 25 microm divinylbenzene/N-vinylpyrrolidone packing using 100% aqueous mobile phase at 4 mL/min. After sample loading and sample cleanup, the analytes were eluted from the extraction column with a 1.0-min gradient at 0.4 mL/min. The samples were either analyzed directly after elution from the extraction column or after additional separation using a short high performance liquid chromatography (HPLC) column. The different configurations were tested using an acidic drug (diflunisal) and a basic drug (clemastine) in rat plasma. On-line analysis was performed by injecting 200 microL of diluted plasma. The mass spectrometer was operated in the multiple reaction monitoring (MRM) mode. All calibration standards gave relative standard deviations (RSDs) below 5%. The total time per sample was 3 min.     

Pharmaceutical analysis by supercritical fluid chromatography: Optimization of the mobile phase composition on a 2-ethylpyridine column

The separation of neutral, acidic, and basic pharmaceuticals with diverse physicochemical properties by packed column supercritical fluid chromatography (pSFC) on a 2-ethylpyridine column (25 cmx4.6 mm id, 3 mum particles) is presented. The optimization strategy involved separations at 100% methanol (MeOH) and at 50% MeOH/50% ACN while keeping the peak symmetry additives formic acid (FA) and isopropylamine (IPA) at constant levels of 0.25% v/v. By plotting the adjusted retention times as a function of the MeOH/ACN ratio, an optimal modifier ratio composition of 65% MeOH/35% ACN was found. The total set of 26 neutral, acidic, and basic pharmaceuticals was analyzed and the optimal composition experimentally verified. This mobile phase composition is currently used in pharmaceutical method development and open-access generic screening environments.     

Evaluation of monolithic and sub 2 microm particle packed columns for the rapid screening for illicit drugs--application to the determination of drug contamination on Irish euro banknotes

A study comparing recently available 100 x 3 mm id, 200 x 3 mm id monolithic reversed-phase columns with a 50 x 2.1 mm id, 1.8 microm particle packed reversed-phase columns was carried out to determine the most efficient approach (using traditional van Deemter analysis and a modern kinetic plot approach) for the rapid screening of samples for 16 illicit drugs and associated metabolites. A plot of column backpressure versus plate number (N) showed a significant advantage of using the monolithic phases, with the 20 cm monolithic column exhibiting a maximum 15,000 plates at a column backpressure of approximately 70 bar, compared to approximately 7000 plates at 150 bar for the 5 cm 1.8 microm particle packed column. Optimum linear velocities were found to be 0.40 mm s(-1), 0.52 mm s(-1) and 0.98 mm s(-1) for the three above columns, respectively. The 20 cm monolithic column was subsequently applied to the separation and determination of illicit drug contamination on Irish euro banknotes, using methanol extraction followed by LC-MS/MS. Method performance data showed that the new LC-MS/MS method was significantly more sensitive than previous GC-MS/MS based methods for this application, with detection limits in the pg note(-1) region, based upon a 20 microL standard injection. All of the notes examined tested positive for trace quantities of cocaine, with benzoylecgonine detected on 12 of the 45 notes sampled. Traces of heroin were also detected on three of the 45 notes.     

填充毛细管液相色谱-毛细管气相色谱在线联用分析八角茴香挥发油

首次将填充毛细管高效液相色谱-毛细管气相色谱在线联用技术(μ-HPLC-CGC)用于分离分析八角茴香果实的挥发油成分。液相色谱选用氰基分析柱(250 mm×0.32 mm i.d.),正己烷-乙腈-二氯甲烷(体积比为80∶8∶12)为流动相,对挥发油样品做族组分分离,得到的5个族组分被依次存放在多位储存接口内,然后不分流分别转入毛细管气相色谱仪做详细分析。气相色谱柱由10 m×0.53 mm i.d.保留间隔柱和30 m×0.53 mm i.d.×1.0 μm SE-54分析柱组成。采用了不分流柱内进样模     

Quantification of myrislignan in rat plasma by solid-phase extraction and reversed-phase high-performance liquid chromatography

A rapid and simple reversed-phase high-performance liquid chromatographic (RP-HPLC) method has been developed for determination of myrislignan in rat plasma after intravenous administration. The analytes extracted from plasma samples by solid-phase extraction were successfully carried out on a Diamonsiltrade mark ODS C(18) column (250 x 4.6 mm i.d., 5 microm) with an RP(18) guard column (8 x 4.6 mm i.d., 5 microm) and a mobile phase of MeOH-H(2)O (4:1, v/v). The UV detector was set at a single wavelength of 270 nm. The linear ranges of the standard curves were 0.5-30.0 microg/mL with the correlation coefficients greater than 0.9992. The lower limits of detection and quantification were 0.1 and 0.3 microg/mL for myrislignan. Intra- and inter-day precisions were 2.4-7.5 and 1.3-5.7%, respectively. The extraction recovery from plasma was more than 90%. This assay method has been successfully used to study the pharmacokinetics of myrislignan in rats.     

Capillary high-performance liquid chromatography/electrospray ion trap time-of-flight mass spectrometry using a novel nanoflow gradient generator

A type of high-performance liquid chromatography (HPLC) based on a novel nanoflow gradient generator (Asymptotic-Trace-10-Port-Valve (AT10PV) nanoGR generator) was developed and coupled with an electrospray ion trap time-of-flight mass spectrometer (ESI-IT-TOF MS). Stability of the nanoflow GR HPLC system was tested at flow rates of 20 and 50 nL/min by using a nanoflow meter. Average flow rates in a 2-h run were 51.2 nL/min with RSD 0.7% and 21.0 nL/min with RSD 1.8%. Repeatability of analysis of the nanoHPLC/ESI-IT-TOF MS system was also tested by injecting 1.0 microL of trypsin digested bovine serum albumin (BSA) (100 fmol) into a monolithic silica-ODS column (30 microm i.d., 150 mm in length) through a packed silica-ODS trapping column (particle size 5 microm, 150 microm i.d., 10 mm in length). At a flow rate of 50 nL/min, the result demonstrated a reasonably good repeatability of peak retention times (RSD: 0.32-1.1%) and base-ion peak areas (RSD: 4.4-6.6%).     

Quantitative analysis of D-24851, a novel anticancer agent, in human plasma and urine by liquid chromatography coupled with tandem mass spectrometry

The development of a liquid chromatography/tandem mass spectrometric assay for the quantitative analysis of the novel tubulin inhibitor D-24851 in human plasma and urine is described. D-24851 and the deuterated internal standard were extracted from 250 microL of plasma or urine using hexane/ether (1:1, v/v). Subsequently, 10-microL aliquots of reconstituted extracts were injected onto an Inertsil ODS analytical column (50 x 2.0 mm i.d., 5 microm particle size). An eluent consisting of methanol/5 mM ammonium acetate, 0.004% formic acid in water (80:20, v/v) was pumped at a flow rate of 0.2 mL/min. An API 365 triple quadrupole mass spectrometer was used in the multiple reaction monitoring mode for sensitive detection. For human plasma a dynamic range of 1-1000 ng/mL was validated, and for human urine a range of 0.25-50 ng/mL. Validation was performed according to the most recent FDA guidelines and all results were within requirements. The assay has been successfully applied to support a phase I clinical trial with orally administered D-24851.     

2D LC Separation and Determination of Bradykinin in Rat Muscle Tissue Dialysate with On-Line SPE-HILIC-SPE-RP-MS

A 2D liquid chromatography (LC) system using hydrophilic interaction chromatography (HILIC) and reversed phase columns has been employed for comprehensive (LC × LC) separation of rat muscle tissue micro-dialysate. Incorporation of an on-line reverse-phase solid phase extraction (SPE) enrichment column in front of the first dimension enabled aqueous samples with high salt concentrations to be injected directly without compromising the chromatographic performance of the HILIC column. Since the SPE enrichment column allowed injection of large sample volumes (e.g. 450 μL), a capillary HILIC column (inner diameter 0.3 mm) could be employed instead of a larger column which is often used in the first dimension to load sufficient amounts of sample. The two chromatographic dimensions were connected using a column selector system with 18, 1.0 mm I.D. C₁₈ “transition” SPE columns. A PLRP C₁₈ column was used in the second dimension. The 2D LC system’s performance was evaluated with a tryptic digest mixture of three model proteins. Good trapping accuracy (HILIC→transition SPE→RP recovery >95%) and repeatability (within-and between day retention time RSDs of first and second dimension chromatography >1%) was achieved. A dialysis sample of rat muscle tissue was separated with the 2D system, revealing complexity and large differences in concentrations of the various compounds present, factors which could potentially interfere with the quantification and monitoring of two target analytes, arg-bradykinin and bradykinin. Subsequently, “Heart-cut” 2D LC-electrospray–mass spectrometry (ESI–MS) with post-column on-line standard injection was employed to monitor arg-bradykinin and bradykinin levels as a function of various muscle conditions. The method’s quantification precision was RSD = 3.4% for bradykinin.     

Determination of major carotenoids in vegetables by capillary electrochromatography

A simple and rapid method for the isocratic separation and determination of carotenoids (carotenes and xanthophylls) in vegetables by CEC is described. The capillary column (100 microm ID, 25 cm effective length) was packed with 3 microm Hypersil ODS particles. The optimized mobile phase contained 60% ACN, 35% THF and 5% of a 5 mM Tris aqueous buffer of pH 8. beta-Carotene, lycopene and lutein were separated with efficiencies of 66 000-128 000 plates/m within a short time (less than 12 min for the last peak eluted, 13/13'-cis-beta-carotene). An excellent resolution of the three carotenoids, as well as partial resolution of their geometrical isomers, was achieved. Application to the determination of the analytes in carrot, tomato, spinach and corn was demonstrated.