Determination of Cobalt with a Tungsten Tube Atomizer by Electrothermal Atomic Absorption Spectrometry and its use for the Analysis of Biological Materials

 The determination of cobalt by electrothermal atomic absorption spectrometry with a tungsten tube atomizer has been investigated. The absolute characteristic mass (the mass of element giving 0.0044 abs.) of cobalt by the atomizer was 5.8 pg and the detection limit was 0.65 ng/ml (3S/N). The interferences caused by large amounts of interferents like Al, Ca, Cu, Fe, K, Mg, Na, Pb and Zn were evaluated and most of these elements were found to interfere with the cobalt signal. Ammonium thiocyanate as a chemical modifier was found to be efficient to eliminate the severest interferences. The accuracies of the recommended method were found to be almost satisfactory for the determination of cobalt in biological materials, as shown by a comparison with the reported values of NIST materials at the 0.02–0.4 μg/g level. Received August 24, 1998. Revision February 9, 1999.     

Determination of fission products in nuclear samples by capillary electrophoresis-inductively coupled plasma mass spectrometry (CE-ICP-MS)

Determination of fission products and of their isotopic composition is of high importance for the characterisation and complete inventory of nuclear fuels. Radiometric and mass spectrometric methods, which are currently used, enable only an incomplete determination of the fission products. Radiometric methods cannot be applied to all fission products and mass spectrometric methods are hindered by the existence of isobaric interferences, therefore a previous chemical separation is required before mass spectrometric analysis. Capillary electrophoresis (CE) has been coupled with inductively coupled plasma mass spectrometry (both ICP-QMS and ICP-SFMS). Typical detection limits of 6 ng/mL and 4 pg/mL for caesium as well as 8 ng/mL and 7 pg/mL for lanthanides have been obtained by CE-ICP-QMS and CE-ICP-SFMS, respectively. In addition to these very low detection limits, the procedures present the advantages to be fast (6 min for caesium and 13 min for lanthanides, respectively), to require a low microliter range sample volume and a nanoliter range injection volume. The radiation dose for the personnel as well as the volume of nuclear liquid wastes generated during the measurements are consequently reduced.The procedures have been applied to nuclear samples from PUREX process and leachates from MOX fuels.     

Comparison Between Two Sample Emulsification Procedures for the Determination of Mo, Cr, V and Ti in Diesel and Fuel Oil by ICP OES Along with Factorial Design

Inductively coupled plasma optical emission spectrometry (ICP OES) was used to determine Mo, Cr, V and Ti, in diesel and in used fuel oil. Samples were introduced into the ICP as emulsions to reduce interferences and allow the use of inorganic standards for quantification. A comparative study between one Triton X-100 emulsion and one detergentless emulsion was made. A 2³ factorial design was applied to elucidate and establish the relationship between three experimental variables: presence of HNO₃, amount of diesel fuel oil (between 5 and 25%) and the presence or O₂ into the Ar plasma gas flow rate. Results indicated that best performance were achieved using 10% sample (w/w) together with concentrated HNO₃ (0.5 mL) and using O₂ as auxiliary gas (0.047 L min⁻¹). The use of O₂ minimized both carbon deposits at the injector tip and plasma background. The addition of HNO₃ resulted in good correlation between inorganic standards used for calibration, and metallo-organic standards used for sample enrichment. Analyte enriched diesel and SRM 1634b were analyzed using the optimized conditions. Recoveries from 90.1 to 106.5% were achieved, with better results for detergent emulsions which enabled limits of detection at the ng g⁻¹ range for Mo, Cr, V and Ti and at smaller background.     

Evaluation and application of argon and helium microstrip plasma for the determination of mercury by the cold vapor technique and optical emission spectrometry

The suitability of a 2.45-GHz atmospheric pressure, low-power microwave microstrip plasma (MSP) operated with Ar and He for the determination of Hg by continuous-flow cold vapor (CV) generation, using SnCl₂/HCl as the reducing agent, and optical emission spectrometry (OES) using a small CCD spectrometer was studied. The areas of stability for a discharge in the Ar and in the He MSP enclosed in a cylindrical channel in a quartz wafer were investigated. The excitation temperatures as measured for discharge gas atoms (Ar I, He I), and the electron number densities at 35–40 W and 15–400 mL min⁻¹ were found to be at the order of 3,200–5,500 K and 0.8 × 10¹⁴–1.6 × 10¹⁴ cm⁻³, respectively. The relative intensity of the Hg I 253.6-nm line and the signal-to-background ratio as a function of the forward power (35–40 W) as well as of the flow rate of the working gas (15–400 mL min⁻¹) were evaluated and discussed. For the selected measurement conditions, the Ar MSP was established to have the lower detection limit for Hg (0.6 ng mL⁻¹) compared with the He MSP. The linearity range is up to 300 ng mL⁻¹ and the precision is on the order of 1–3%. With the optimized CV Ar MSP-OES method a determination of Hg in spiked domestic and natural waters at concentration levels of 20–100 μg L⁻¹ and an accuracy of 1–4% could be performed. In an NIST domestic sludge standard reference material, Hg (3.64 μg g⁻¹) could be determined with a relative standard deviation of 4% and an agreement better than 4%.     

Selective determination of ubiquinone in human plasma by HPLC with chemiluminescence reaction based on the redox cycle of quinone

Ubiquinone is an important biologically active compound in the living body. The determination of ubiquinone in human plasma is useful for the investigation of bioavailability of ubiquinone and for early diagnosis of several diseases. Therefore, we developed a high-performance liquid chromatography (HPLC) with chemiluminescence detection method for the analysis of ubiquinone in plasma samples. The method is based on luminol chemiluminescence detection of super oxide anion that is generated by the redox cycle reaction between ubiquinone and dithiothreitol. The HPLC system involved an octyl column with a mobile phase of methanol. Ubiquinone eluted from the column was mixed with dithiothreitol and luminol solutions simultaneously, and generated chemiluminescence was monitored by chemiluminescence detector. The calibration curve for standard ubiquinone solution was linear from 0.09 to 43.2 μg/mL (0.45–216 ng on column) with the correlation coefficient of 0.999, and the detection limit (S/N = 3) was 26 ng/mL (130 pg on column). Using the proposed HPLC method, the peak of ubiquinone in human plasma could be clearly detected on the chromatogram without any interference from plasma components.     

Validation of a high-performance chromatographic method for determination of cefotaxime in biological samples

An analytical method for detecting and quantifying cefotaxime in plasma and several tissues is described. The method was developed and validated using plasma and tissues of rats. The samples were analyzed by reversed phase liquid chromatography (HPLC) with UV detection (254 nm). Calibration graphs showed a linear correlation (r > 0.999) over the concentration ranges of 0.5–200 μg/mL and 1.25–25 μg/g for plasma and tissues, respectively. The recovery of cefotaxime from plasma standards prepared at the concentrations of 25 μg/mL and 100 μg/mL was 98.5 ± 3.5% and 101.8 ± 2.2%, respectively. The recovery of cefotaxime from tissue standards of liver, fat and muscle, prepared at the concentration of 10 μg/g was: 89.8 ± 1.2% (liver), 103.9 ± 6.5% (fat) and 97.8 ± 2.1% (muscle). The detection (LOD) and quantitation (LOQ) limits for plasma samples were established at 0.11 μg/mL and 0.49 μg/mL, respectively. The values of these limits for tissues samples were approximately 2.5 times higher: 0.3 μg/g (LOD) and 1.25 μg/g (LOQ). For plasma samples, the deviation of the observed concentration from the nominal concentration was less than 5% and the coefficient of variation for within-day and between-day assays was less than 6% and 12%, respectively. The method was used in a pharmacokinetic study of cefotaxime in the rat and the mean values of the pharmacokinetic parameters are given. Received: 25 May 1998 / Revised: 27 July 1998 / Accepted: 1 August 1998     

Performance of ICP-MS with Hexapole Collision Cell and Application for Determination of Trace Elements in Bio-Assays

 The application of a hexapole collision cell in quadrupole based inductively coupled plasma mass spectrometry (HEX-ICP-QMS) was studied systematically in order to characterize the performance of this approach for multielement analysis. Additionally, the performance of different solution introduction systems was studied for use with HEX-ICP-QMS (Platform, Micromass Ltd., Manchester, UK) and the developed methods were applied in the multielement analysis of biological tissue. A significant improvement in detection limits was achieved for ions affected by interferences with argon ions and argon-based molecular ions due to the removal of Ar⁺ and ArX⁺ ions (X = H, C, N, O, Ar) as well as for heavy elements (m > 100 u) due to better ion transmission through the hexapole ion guide. Metal oxide ion formation was reduced by up to ten times by the application of hexapole bias potential. Accuracy of the HEX-ICP-QMS measurements was assessed by analyzing standard reference material SRM 1566a Oyster Tissue and comparison with neutron activation analysis. Good agreement of measured trace element concentrations with certified values was found for most elements. Received December 7, 2000. Revision April 3, 2001.     

Utility of Certain σ and π-Acceptors for the Spectrophotometric Determination of Sildenafil Citrate (Viagra)

 The molecular interaction between sildenafil citrate as electron donor and each of iodine; 7,7,8,8-tetracyanoquinodimethane (TCNQ); 2,3-dichloro-5,6-dicyano-1,4-benzoquinone (DDQ); tetracyanoethylene (TCNE); 2,4,7-trinitro-9-fluorenon (TNF); chloranilic acid (CLA); chloranil (CL) and bromanil (BL) as acceptors have been investigated spectrophotometrically. Different variables affecting the reaction were studied and optimized. Beer’s law was obeyed in a concentration limit of 10–260 μg/mL for sildenafil citrate. For more accurate analysis, Ringbom optimum concentration range was found to be between 20–240 μg/mL. The limits of detection and determination were calculated and found to be 1.5 and 5.2 μg/mL, respectively. The standard deviations were calculated for different concentrations of sildenafil citrate using various acceptors. A Job’s plot of the absorbance versus the molar ratio of the sildenafil citrate to each of acceptors under consideration indicated (1:1) ratio. The proposed methods were found to be rapid, accurate, precise and sensitive and could be applied for determination of sildenafil citrate in pharmaceutical dosage forms (Viagra) without interferences from common additives encountered. Received August 30, 2000. Revision January 5, 2001.     

Development of automated amperometric detection of antibodies against <Emphasis Type="Italic">Bacillus anthracis</Emphasis> protective antigen

Picogram levels of antibodies against the protective antigen (PA) of Bacillus anthracis were detected in an automated electrochemical sandwich-type enzyme-linked immunosorbant assay. The antibodies were captured and detected using an 8 × 3 array of 50-μm-diameter cavities. The reagent and sample volumes were as low as 200 nL in a less than 25-min assay from capture to signal generation. The electrochemical detection of the antibodies was demonstrated at 0.05–10 μg/mL containing only 10–5,000 pg antibodies. The limit of detection is 10 fg for a 200-nL sample. Detection of anti-PA immunoglobulin G performed in spiked normal human serum and fresh whole human blood did not show a significant difference from detection in a buffer. The initial automation of the assay involved the use of a digital syringe pump for the delivery of reagents to the capture surface.     

A simplified hollow-fibre supported liquid membrane extraction method for quantification of 2-amino-1-methyl-6-phenylimidazo[4,5-<Emphasis Type="BoldItalic">b</Emphasis>]pyridine (PhIP) in urine and plasma samples

A simple and easy-to-use extraction procedure has been optimised, validated, and applied for extraction of 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) in urine and spiked plasma samples. PhIP is a carcinogenic and mutagenic heterocyclic aromatic amine that is formed during cooking of meat and fish. The novelty of the extraction procedure lies in using a short piece of narrow capillary-like microporous hollow-fibre (HF) membrane as extraction device. The HF membrane was filled with a few microlitres of acidic solution and the membrane pores were impregnated with an organic extraction solvent. Therefore, the technique was called hollow-fibre supported liquid membrane (HF-SLM) extraction. The HF extraction device was then supported by a syringe needle and directly immersed in urine (1.4 mL) or plasma (0.3 mL) previously made alkaline by adding 0.5 mol L⁻¹ NaOH solution to give a final volume of 1.6 mL. The operation of the HF-SLM extraction at the optimal conditions resulted in a PhIP extraction efficiency of 74% from both spiked urine and plasma, corresponding to enrichment factors of 126 and 27, respectively. For 90 min extraction time, limits of detection and quantification were, respectively, 8 and 25 pg mL⁻¹ for urine and 6 and 11 pg mL⁻¹ for plasma. Within-day repeatability (n = 6) and between-day reproducibility (n = 3) were, respectively, 5% and 13% for urine and 6% and 7% for plasma. Analysis of urine samples collected for 12 h after a volunteer had eaten 250 g well-done chicken showed the PhIP concentration was 124 ± 21 pg mL⁻¹, calculated assuming an extraction efficiency of 74%.     

Two-Phase Aqueous Extraction of Chromium and its Application to Speciation Analysis of Chromium in Plasma

 This work proposes a new extraction method for chromium based on the two-phase aqueous system isopropyl alcohol-ammonium sulfate-ammonium thiocyanate (i-PrOH-(NH₄)₂SO₄-NH₄SCN), and the related experimental conditions are optimized. The results show that chromium (III) can be quantitatively extracted under the selected conditions: 4 mL of i-PrOH, 200 μL of 2 mol/L sulfuric acid, 1 mL of 4 mol/L NH₄SCN and 3 mL of saturated (NH₄)₂SO₄ solution (V_total=10 mL). Application of the proposed method to speciation analysis of plasma chromium was also investigated and satisfactory results were obtained. Received May 22, 1999. Revision November 8, 1999.     

Determination of lead in sediments and sewage sludge by on-line hydride-generation axial-view inductively-coupled plasma optical-emission spectrometry using slurry sampling

Among the “traditional” hydride-forming elements, lead is probably the most difficult, and its determination in this form has rarely been reported in the literature. In this paper a simple and rapid method, axial-view inductively-coupled plasma optical-emission spectrometry using on-line hydride generation (HG–ICP–OES) from samples prepared as slurry, is proposed for determination of lead in environmental samples. The samples (20–50 mg, particle size ≤120 μm) were treated with 1 mL aqua regia in a 40-kHz ultrasonic bath for 60 min. The slurry was diluted to a final volume of 50 mL with a 10% m/v solution of (NH₄)₂S₂O₈. The concentrations of NaBH₄, tartaric acid, and (NH₄)₂S₂O₈, used for on-line plumbane generation were optimized by means of a complete factorial analysis applied to an aqueous standard solution and to the slurry of a sediment certified reference material (CRM). External calibration against aqueous standards in the concentration range 10–100 μg L⁻¹ was used for analysis of six CRM—three marine sediments, one river sediment, and two sewage sludges. Analysis of the filtered slurry showed that Pb was only partially extracted into the liquid phase. Several major concomitants tested did not affect the Pb signal. The detection limit (3s, n = 10) for 20 mg sample in a final volume of 50 mL was 5.0 μg g⁻¹. Tin was the only other hydride-forming analyte that could be determined satisfactorily with Pb; for tin the detection limit was 1.0 μg g⁻¹. The values obtained for Pb and Sn were not significantly different from the certified concentrations, according to the t-test at the 95% confidence level. Nine river sediments collected locally were also analyzed and the concentrations were in agreement with results obtained after total digestion.     

Determination of vancomycin in human plasma using high-performance liquid chromatography with fluorescence detection

A high-performance liquid chromatographic (HPLC) method with fluorescence detection for the quantification of vancomycin in human plasma was developed and validated. The method includes an extraction of vancomycin by deproteinization with acetonitrile. The analyses were carried out at 258 nm as the emission wavelength while exciting at 225 nm on a reversed-phase column (30 cm × 4 mm i.d. × 10 μm Waters Associates μBondapak C18) using a mobile phase composed of methanol and phosphate buffer at pH 6.3. Vancomycin was quantitatively recovered from human plasma samples (>96%) with high values of precision. The separation was completed within 27 min. The calibration curve was linear over the range from 5 to 1,000 ng/mL with the detection and quantification limits of 2 ng/mL and 5 ng/mL, respectively. This method is suitable for the routine assay of plasma samples. Figure The effect of the deproteinization solvent on the signal of the interference peak at retention time of 15.0 min. The peak which interferes with the peaks of Erythromycin and Vancomycin has been disappeared by using 2 mL acetonitrile as the deproteinization solvent.     

High-throughput and simultaneous analysis of eight central-acting muscle relaxants in human plasma by ultra-performance liquid chromatography–tandem mass spectrometry in the positive and negative ionization modes

In this report, a high-throughput and sensitive method for analysis of eight central-acting muscle relaxants in human plasma by ultra-performance liquid chromatography–tandem mass spectrometry (UPLC-MS/MS) in the positive and negative ionization modes using tolbutamide as internal standard is presented. After pretreatment of a plasma sample by solid-phase extraction with an Oasis HLB cartridge, muscle relaxants were analyzed by UPLC with Acquity UPLC BEH C₁₈ column and Acquity TQD tandem quadrupole mass spectrometer equipped with an electrospray ionization interface. The calibration curves for muscle relaxants spiked into human plasma equally showed good linearities in the nanogram per milliliter order range. The detection limits (signal-to-noise ratio = 3) was as low as 0.1–2 ng/mL. The method gave satisfactory recovery rates, accuracy, and precision for quality control samples spiked with muscle relaxants. To further validate the present method, 250 mg of chlorphenesin carbamate was orally administered to a healthy male volunteer, and the concentrations of chlorphenesin carbamate in plasma were measured 0.5, 1, 2, 4, 6, and 8 h after dosing; their concentrations in human plasma were between 0.62 and 2.44 μg/mL. To our knowledge, this is the first report describing simultaneous analysis of over more than two central-acting muscle relaxants by liquid chromatography–tandem mass spectrometry. This has been realized by the capability of our instrument for simultaneous multiple reaction monitoring of the target compounds in both positive and negative ionization modes. Therefore, the present method seems very useful in forensic and clinical toxicology and pharmacokinetic studies.     

Automated mass spectrometric analysis of urinary and plasma serotonin

Serotonin emerges as crucial neurotransmitter and hormone in a growing number of different physiologic processes. Besides extensive serotonin production previously noted in patients with metastatic carcinoid tumors, serotonin now is implicated in liver cell regeneration and bone formation. The aim was to develop a rapid, sensitive, and highly selective automated on-line solid-phase extraction method coupled to high-performance liquid chromatography–tandem mass spectrometry (XLC-MS/MS) to quantify low serotonin concentrations in matrices such as platelet-poor plasma and urine. Fifty microliters plasma or 2.5 μL urine equivalent were pre-purified by automated on-line solid-phase extraction, using weak cation exchange. Chromatography of serotonin and its deuterated internal standard was performed with hydrophilic interaction chromatography. Mass spectrometric detection was operated in multiple reaction monitoring mode using a quadrupole tandem mass spectrometer with positive electrospray ionization. Serotonin concentrations were determined in platelet-poor plasma of metastatic carcinoid patients (n = 23) and healthy controls (n = 22). Urinary reference intervals were set by analyzing 24-h urine collections of 120 healthy subjects. Total run-time was 6 min. Intra- and inter-assay analytical variation were <10%. Linearity in the 0–7300 μmol/L calibration range was excellent (R² > 0.99). Quantification limits were 30 and 0.9 nmol/L in urine and plasma, respectively. Platelet-poor serotonin concentrations in metastatic carcinoid patients were significantly higher than in controls. The urinary reference interval was 10–78 μmol/mol creatinine. Serotonin analysis with sensitive and specific XLC-MS/MS overcomes limitations of conventional HPLC. This enables accurate quantification of serotonin for both routine diagnostic procedures and research in serotonin-related disorders.     

Quantification of plasma phospholipids by ultra performance liquid chromatography tandem mass spectrometry

We describe a fast and robust ultra performance liquid chromatography tandem mass spectrometry method for the quantification of phospholipid (PL) species in EDTA-plasma samples. We quantified total phosphatidylcholine (PC), phosphatidylethanolamine (PE), lysophosphatidylcholine (LPC), and sphingomyelin (SM) and several species within these classes using one or two external calibrators and one internal standard for each class. Inter-assay coefficients of variation were <10% for the most abundant species and <20% for all quantified PC, LPC, and SM species and the three most abundant PE species. Coefficients of linear regression were R ² > 0.98. Mean recoveries were between 83% and 123%. The limits of detection were 0.37 μmol/L for PC, 4.02 μmol/L for LPC, 3.75 μmol/L for PE, and 0.86 μmol/L for SM. Quantification was linear over the physiological ranges for PE, LPC, and SM and up to 500 μmol/L for PC. The concentrations of PLs in the plasma of healthy donors yielded results that were comparable with those of previous works.     

Automated on-line in-tube solid-phase microextraction coupled with HPLC/MS/MS for the determination of butyrophenone derivatives in human plasma

This paper describes a fully automated on-line method combining in-tube solid-phase microextraction (SPME) in which sample clean-up and enrichment are conducted through an open tubular fused-silica capillary column and high-performance liquid chromatography (HPLC)/tandem mass spectrometry (MS/MS) detection for the determination of six butyrophenone derivatives (moperone, floropipamide, haloperidol, spiroperidol, bromperidol, and pimozide) in human plasma samples. The six butyrophenones were extracted by repeatedly aspirating and dispensing plasma sample solutions on a DB-17 capillary column (60 cm × 0.32 mm i.d., film thickness 0.25 μm). The analytes retained on the inner surface of the capillary column were then eluted into an acetonitrile-rich mobile phase using a gradient separation technique. Extraction efficiencies ranged from 12.7% to 31.8% for moperone, spiroperidol, and pimozide, and from 1.08% to 4.86% for floropipamide, haloperidol, and bromperidol. The regression equations for all compounds showed excellent linearity, ranging from 0.05 to 50 ng/0.1 mL of plasma, except for moperone and spiroperidol (0.01 to 50 ng/0.1 mL). The limits of detection and quantification in plasma for each drug were 0.03–0.2 and 0.1–0.5 ng/mL, respectively. The intra- and inter-day coefficients of variation for all compounds in plasma were not greater than 13.7%.     

Direct Analysis of Coffee and Tea for Aluminium Determinationby Electrothermal Atomic Absorption Spectrometry

 A method for direct analysis of tea and coffee samples by using electrothermal atomic absorption spectrometry is described. Coffee and tea from different sources were analyzed without digestion step. For slurry analyses the samples were ground, sieved at 105 μm and then suspended in 0.2% v/v HNO₃ and 10% v/v Triton X-100 medium. For liquid phase aluminium determination the samples were prepared in the same way and only the liquid is introduced directly into the graphite furnace. Calibration was performed by aqueous standards for both cases and the determinations were carried out in the linear range between 50 and 250 μg L⁻¹. The characteristic mass of aluminium and the detection limit were 45 pg and 2 μg L⁻¹, respectively. Using a typical 0.1% m/v coffee slurry sample, the relative standard deviation of measurements (n=15) for repeatability was about 8.2%. Received December 27, 1998. Revision March 18, 1999.     

A critical evaluation of the graphite furnace conditions for the direct determination of chromium in urine

A critical study of the conditions for the direct determination of Cr in urine using GFAAS was carried out. A chemometric approach showed that the most important parameters that influence the efficiency of atomic Cr cloud formation were the atomization from the tube wall and the proper implementation of a pyrolysis step. The established procedure does not require any sample pre-treatment thus minimizing the risk of contamination. The Cr content in urine was determined using a calibration curve prepared with Cr spiked urine, easily correcting all potential matrix interferences. The heating program proposed took 68 s for a 30 μL urine sample. The procedure was applied for the determination of Cr in urine to investigate the absorption rate of Cr picolinate. The limit of detection and the characteristic mass for a 30 μL urine sample were 0.18 μg/L and 5.4 pg, respectively. Received: 22 June 1998 / Revised: 13 January 1999 / Accepted: 18 January 1999     

Determination of antibacterials in animal feed by pressurized liquid extraction followed by online purification and liquid chromatography-electrospray tandem mass spectrometry

A fully automated method has been developed for analysis of eighteen antibacterial compounds, including penicillins, cephalosporins and sulfonamides, in animal feed with limits of quantification in the range 0.25–5.79 μg kg⁻¹. The method is based on pressurized liquid extraction of 3 g homogenized feed with water and online clean-up of 500 μL of the extract with C₁₈HD cartridges. The purified sample was directly analysed by liquid chromatography–electrospray tandem mass spectrometry (SPE–LC–ESI-MS–MS). Chromatographic separation was achieved within 10 min by use of a C₁₂ Phenomenex Hydro-RP reversed-phase analytical column and a mobile phase gradient (water + 0.1% formic acid–methanol + 0.1% formic acid). The method was validated, revealing capability for detection of concentrations as low as 0.09 μg kg⁻¹, decision limits (CCα) and detection capabilities (CCβ) in the range 10–174 μg kg⁻¹ and 22–182 μg kg⁻¹, respectively, and inter-day precision ranging from 0.7 to 8.3%. Recovery, with internal standard correction, was in the range 93–134% for all analytes. The method was then applied to analysis of fifteen feed samples, nine of which contained at least one antimicrobial at concentrations between 0.006 and 1.526 mg kg⁻¹. The performance data and results from the method were compared with those from a previous method developed by our group, using offline SPE, by analyzing the same set of samples by both methods. The online SPE approach resulted in slightly improved sensitivity, with LODs of 0.09–2.12 μg kg⁻¹ compared with 0.12–3.94 μg kg⁻¹ by the offline approach. In general, better recovery was achieved by use of online purification (for 72% of the analytes) and the correlation between the two methods was good. The main advantages of the new online method are rapid and automated sample pre-treatment, and reduction of sample manipulation, enabling high-throughput analysis and highly accurate results. Because of all these characteristics, the proposed method is applicable and could be deemed necessary within the field of food control and safety.