Determination of substituted benzenes in water samples by fiber-in-tube liquid phase microextraction coupled with gas chromatography

A method for determination of toluene, ethylbenzene, p-xylene, o-xylene, 1,3,5-trimethylbenzene and 1,2,4-trimethylbenzene in water samples was developed by a fiber-in-tube liquid phase microextraction technique (fiber-in-tube LPME) coupled with GC-flame ionization detector (FID). The method used a tube packed with polytetrafluoroethylene (PTFE) fibers as an extraction medium, improving the stableness of the solvent and the performance of extraction. Certain amounts of curled PTFE fibers were packed into a section of PTFE tube. Because the fibers were curled, they formed network structure in the tube. The fiber packed tube was firstly immersed into organic solvent to be filled with organic solvent and then was exposing to an aqueous solution to extract the target compounds. The extract was then retracted by a conventional GC microsyringe and analyzed by GC-FID. Extraction of the analytes in 8 ml aqueous solution for 15 min yielded enrichment factors of 224-361. The precision (R.S.D., n = 5) was 3.6-8.1% for peak area. The limit of detection (LOD, S/N = 3) for the six substituted benzenes were in the range of 0.3-5.0 μg l⁻¹.     

Electrothermal vaporization inductively coupled plasma mass spectrometry for the determination of trace amount of lanthanides and yttrium in soil with polytetrafluroethylene emulsion as a chemical modifier

A method of electrothermal vaporization inductively coupled plasma mass spectrometry (ETV-ICP-MS) for the determination of trace lanthanides and yttrium in soil samples with a polytetrafluorethylene (PTFE) emulsion as chemical modifier to promote the vaporization of the analytes from the graphite furnace was developed in this paper. The analytical characteristics, spectral interference and matrix effect of the analytical method were evaluated and critically compared with those of pneumatic nebulization inductively coupled plasma mass spectrometry (PN-ICP-MS). Under the optimized operation conditions, the relative detection limits of lanthanides (La-Lu) and yttrium for ETV-ICP-MS and PN-ICP-MS were 0.4-20 ng l⁻¹ and 1.0-21 ng l⁻¹, respectively, the absolute detection limits for ETV-ICP-MS were 4-200 fg, which were improved by 1-2 orders of magnitude compared with PN-ICP-MS. While the analytical precision of ETV-ICP-MS is worse than that of PN-ICP-MS, with the R.S.D.s (%) of 4.1-10% for the former and 2.9-7.8% for the latter. Regarding to the matrix effect, both conventional method and stepwise dilution method were employed to observe the effect of matrix and the very similar results were obtained. It was found that the highest tolerance concentration of the matrix is 1000 mg l⁻¹ and 800 mg l⁻¹ for ETV-ICP-MS and PN-ICP-MS, respectively. To assess the accuracy, the proposed method was applied to the determination of trace lanthanides and yttrium in three different soil standard reference materials and one soil sample, and the determined values are in good agreement with the certified values or reference values.     

Novel preconcentration technique using bis(2-ethylhexyl) hydrogen phosphate (HDEHP) loaded porous polytetrafluoroethylene (PTFE) filter tube as a sorbent: Its application to determination of In(III) in seawater by ICP-MS with air segmented discrete sample introduction

Novel preconcetration method involving porous polytetrafluoroethylene (PTFE) filter tube impregnated bis(2-ethylhexyl) hydrogen phosphate (HDEHP) as a sorbent was studied to establish it as a practical preconcentration method for ultra trace analysis with ICP-MS. A 1 ng portion of In(III) in 1000 ml of matrix free solution or 700 ml of synthetic seawater was quantitatively complexated with HDEHP adsorbed onto porous PTFE filter tube by passing the solution through the micro pore of the filter tube. Preconcentrated In(III) was then quantitatively recovered provided that the elution, which consists of a cyclical filtering 0.1 ml of 8 mol dm⁻³ hydrochloric acid through the filter tube for 1 min; therefore, up to 10000- and 7000-fold of enrichment was attained for matrix free solution and synthetic seawater, respectively. To introduce a 0.1 ml of the eluted solution to ICP-MS, flow injection method with air segmented discrete sample introduction (ASDI) was also studied by using manually operated simple valve system. By using ASDI, good linearity of calibration curve (r = 0.99997) was observed from 0.01 to 5.0 ng ml⁻¹ of In(III). Good reproducibility was also shown in measurements of 0.1 ml of 5 ng ml⁻¹ of In(III) (R.S.D. = 1.9%, n = 5). The average recovery and R.S.D. of the results for the five duplicates determination of 0.1 ng of In(III) spiked to 200 ml of synthetic seawater were 99 and 2.4%, respectively. The method was applied to the determination of In(III) in coastal seawater sampled at north east of Hachijyo Island, Japan; using 200 ml of sample, 2000-fold preconcentration of In(III) was performed within 30 min for five samples. The analytical detection limit and the blank were 9.8 and 21 pg l⁻¹, respectively. The average concentration was determined to be 96 pg l⁻¹, and R.S.D. of the results was 3.7% (n = 3).     

Leachability and analytical speciation of antimony in coal fly ash

A new procedure was described with multiwalled carbon nanotubes as solid phase extraction packing material for the trace analysis of nicosulfuron, thifensulfuron and metsulfuron-methyl in water samples. The possible parameters influencing the enrichment were optimized and the optimal conditions were as followed: eluent, sample pH, flow rate and sample volume were acetonitrile containing 1% acetic acid, pH 3, 8 mL min⁻¹ and 500 mL, respectively. Under the optimal chromatographic separation and SPE conditions, the linear range, detection limit (S/N = 3) and precision (R.S.D., n = 6) were 0.04-40 ng mL⁻¹, 6.8 ng L⁻¹ and 2.5% for nicosulfuron, 0.04-40 ng mL⁻¹, 11.2 ng L⁻¹ and 5.4% for thifensulfuron, 0.02-20 ng mL⁻¹, 5.9 ng L⁻¹, 2.1% for metsulfuron-methyl, respectively. The established method was well employed to determine nicosulfuron, thifensulfuron and metsulfuron-methyl in tap water, seawater, reservoir water and well water samples, and satisfactory results were obtained, the spiked recoveries in the range of 87.2-100.7%, 96.5-105.6% and 83.7-111.1% for them each, respectively.     

Determination of trace elements in natural waters by inductively coupled plasma time of flight mass spectrometry after flow injection preconcentration in a knotted reactor

A flow injection on-line sorption system was developed for the separation and preconcentration of traces of Ag, Cd, Co, Ni, Pb, U and Y from natural water samples with subsequent detection by ICP TOF MS. Simultaneous preconcentration of the analytes was achieved by complexation with the chelating reagent 1-phenyl-3-methyl-4-benzoylpyrazol-5-one immobilized on the inner walls of a (200 cm × 0.5 mm) PTFE knotted reactor. The analytes were eluted and transported to an axial ICP TOF MS system with 1% (v/v) HNO₃ containing 0.3 μg l⁻¹ of Rh as an internal standard using ultrasonic nebulization. The detection limits (3σ) varied from 0.3 ng l⁻¹ for Y to 15.2 ng l⁻¹ for Ni and the precision (R.S.D.) was better than 4%. Using a loading time of 90 s and a sample flow rate of 4.5 ml min⁻¹, enhancement factors of 3-14 were obtained for the different analytes in comparison with their direct determination by ICP TOF MS with ultrasonic nebulization without preconcentration. The accuracy of the method was demonstrated by analysis of water based certified reference materials.     

Ultrasensitive and selective determination of trace amounts of nitrite ion with a novel fluorescence probe mono[6-N(2-carboxy-phenyl)]-β-cyclodextrin

A novel fluorescence probe, mono[6-N(2-carboxy-phenyl)]-β-cyclodextrin (OACCD), has been developed for the determination of trace nitrite, In dilute HCl medium, the fluorescence intensity of the newly synthesized fluorescence probe OACCD was quenched in presence of trace nitrite at room temperature. Based on this, a simple, sensitive, and selective method for rapid determination of nitrite was described. Furthermore, common ions do not interfere the determination of trace amounts of nitrite. The fluorescence quenching intensity was linear over a nitrite concentration of 0.02-1.7 μmol l⁻¹ with a detection limit of 0.2 nmol l⁻¹ (S/N = 3). The method was applied to the determination of nitrite in different water samples, soil samples, and food samples with satisfactory results.     

Ultratrace determination of phosphorus in ultrapurified water by a slope comparison method

The analytical method for the determination of phosphorus in ultrapurified water was developed. Ultrapurified water was evaporated to concentrate phosphorus and the final sample volume for analysis was 10 ml. In 0.55 mol l⁻¹ HCl, orthophosphate forms molybdophosphate, and then the molybdophosphate forms ion associate with Malachite Green (MG), which can be collected on a tiny membrane filter (diameter: 5 mm, and effective filtering diameter: 2 mm). After the ion associate on the membrane filter is dissolved together with the membrane filter in 1 ml of methyl cellosolve (MC), the absorbance of MC solution is measured at 627 nm by a flow injection-spectrophotometric detection technique. When 10 ml of the sample solution was used for the procedures and absorbance measurement, the calibration graph is linear up to about 500 ng l⁻¹ of phosphorus and the detection limit was 8 ng l⁻¹ (S/N=3). For the determination of phosphorus in an ultrapurified water, 10-40 ml of sample solutions were transferred into poly(tetrafluoroethylene) (PTFE) beaker and evaporated to 5 ml or to dryness. To them, 0.003 mol l⁻¹ HCl was added to get 10 ml of final solution, which was used as sample. Phosphate is determined by comparing the slope of the varied sample volume after evaporation/concentration with a slope of the standard calibration graph (a slope comparison method: SCM). The SCM enables to evaluate the concentration of phosphate in ultrapurified waters more sensitively and accurately.     

Admicelle-enhanced synchronous fluorescence spectrometry for the selective determination of polycyclic aromatic hydrocarbons in water

A simple and rapid method for the highly sensitive determination of polycyclic aromatic hydrocarbons (PAHs) in water was developed. Benzo[a]pyrene, benzo[k]fluoranthene, perylene, and pyrene in water were concentrated into sodium dodecyl sulfate (SDS)-alumina admicelles. The collection was performed by adding SDS and alumina particles into the sample solution at pH 2. After gentle mixing, the resulting suspension was passed through a membrane filter to collect the SDS admicelles containing highly concentrated PAHs. The filter was placed on a slide glass and then covered admicellar layer with a fused silica glass plate before setting in a fluorescence spectrometer. Benzo[a]pyrene, benzo[k]fluoranthene, perylene, and pyrene were selectively determined by the synchronous fluorescence scan (SFS) analysis with keeping wavelength intervals between excitation and emission to 98, 35, 29, and 45 nm, respectively. Because of the minimum spectral overlapping, 1-40 ng l⁻¹ of benzo[a]pyrene, benzo[k]fluoranthene, and perylene as well as 10-150 ng l⁻¹ of pyrene were selectively determined with eliminating the interferences of other 12 PAHs. The detection limits were 0.3 ng l⁻¹ for benzo[a]pyrene, benzo[k]fluoranthene, and perylene, and 1 ng l⁻¹ for pyrene. They were 2-3 orders of magnitude lower than the detection limits in normal aqueous micellar solutions. The application to water analysis was studied.     

Europium-sensitized luminescence determination of oxytetracycline in catfish muscle

An europium-sensitized time-resolved luminescence (TRL) method was developed to determine oxytetracycline (OTC) in cultivated catfish muscle. Extraction of OTC from fish muscle was performed with pH 4.0 ethylenediaminetetraacetic acid (EDTA)-McIlvaine buffer and clean up with hydrophilic-lipophilic balanced copolymer solid phase extraction (SPE) cartridges. The eluate was used without further concentration for TRL measurement in pH 9.0 micellar tris(hydroxylmethyl)aminomethane (TRIS) buffer. Cetyltrimethylammonium chloride (CTACl) was used as surfactant and EDTA as a co-ligand. The excitation and emission wavelengths were set at 388 and 615 nm, respectively. The linear dynamic range was 0-1000 ng g⁻¹ (R²=0.9995). The recovery was 92-112% in the fortification range of 50-200 ng g⁻¹ and the limits of detection (LOD) ranged from 3 to 7 ng g⁻¹. Incurred catfish samples were used to demonstrate the performance of the method around 100 ng g⁻¹, the European Union maximum residue level.     

Spectrofluorimetric determination of trace nitrite in food products with a new fluorescent probe 1,3,5,7-tetramethyl-2,6-dicarbethoxy-8-(3',4'-diaminophenyl)-difluoroboradiaza-s-indacene

A new fluorescent probe 1,3,5,7-tetramethyl-2,6-dicarbethoxy-8-(3′,4′-diaminophenyl)-difluoroboradiaza-s-indacene (TMDCDABODIPY) has been developed to detect nitrite in meat products and vegetables. The fluorescence of TMDCDABODIPY is very weak, but when it reacts with nitrite, a strong fluorescent triazole forms in aqueous medium at room temperature, which offers the advantage of specificity and sensitivity for the determination of nitrite. The fluorescence intensity was linear over a nitrite concentration of 9-300 nmol l⁻¹ with a detection limit of 0.21 nmol l⁻¹ (S/N = 3). The proposed method has been used for the determination of trace nitrite in food products with the recoveries of 94.62-105.48%.     

Separation/preconcentration of trace amounts of Cr, Cu and Pb in environmental samples by magnetic solid-phase extraction with Bismuthiol-II-immobilized magnetic nanoparticles and their determination by ICP-OES

A new method for separation/preconcentration of trace amounts of Cr, Cu and Pb in environmental samples by magnetic solid-phase extraction (SPE) with Bismuthiol-II-immobilized magnetic nanoparticles and their determination by ICP-OES has been developed. The separation of the target analytes from the aqueous solution containing the target analytes and Bismuthiol-II-immobilized magnetic nanoparticles was simply achieved by applying external magnetic field. Optimal experimental conditions including pH, sample volume, eluent concentration and volume and co-existing ions have been studied and established. Under the optimal experimental conditions, the detection limits for Cr, Cu and Pb with enrichment factors of 96, 95 and 87 were found to be 0.043, 0.058 and 0.085 ng mL⁻¹ and their relative standard deviations (R.S.D.s) were 3.5%, 4.6% and 3.7% (n = 5, C = 2 ng mL⁻¹), respectively. The method was validated with certified reference material (GBW50009-88) of environmental water sample and the analytical results coincided well with the certified values. Furthermore, the method was successfully applied to the determination of target analytes in river and lake water samples. Compared with established methods, the proposed method is characterized with high enrichment factor, fast separation and low detection limits.     

A rapid and sensitive method for the determination of trace proteins based on the interaction between proteins and Ponceau 4R

The interactions between proteins and Ponceau 4R (PR) in aqueous solution have been studied by the techniques of resonance light scattering (RLS) spectroscopy, the absorption spectroscopy, zeta potential assay and circular dichroism (CD) spectrum. The dry PR can assemble on the surface of protein via electrostatic and hydrophobic forces to produce an associated compound of protein-PR, this compound can enhance the RLS of protein. Based on this fact, a simple, rapid, and sensitive method has been developed for the determination of proteins at nanogram level by RLS technique with a common spectrofluorimeter. Under optimum conditions, the linear range is 0.10-39.2 μg mL⁻¹ for the determination of both bovine serum albumin (BSA) and human serum albumin (HSA). The detection limits (S/N = 3) are 6.96 ng mL⁻¹ for BSA and 5.71 ng mL⁻¹ for HSA, respectively. There is almost no interference from amino acids, most of the metal ions, and other coexistent substances. The method has been satisfactorily applied to the direct determination of the total protein in human serum.     

On-line sorption preconcentration of metals based on mixed micelle cloud point extraction prior to their determination with micellar chemiluminescence: Application to the determination of chromium at ng l levels

The on-line incorporation of cloud point extraction (CPE) to flow injection analysis (FIA) is modified to extract and preconcentrate metal species rapidly, avoiding the formation of hydrophobic complexes, using a mixed micellar medium. Coupling the procedure with chemiluminescence (CL) detection based on the catalytic activity of metal species on the luminol-hydrogen peroxide reaction and enhancing the signal with the presence of a micellar carrier containing bromide ions produces a powerful tool for the preconcentration and determination of metal species at ng l⁻¹ levels. As an analytical demonstration ultratrace concentrations of chromium were conveniently detected and quantified in samples with a complex matrix such as seawater and wastewater. The figures of merit for the determination of chromium were: 0.9-1.6% R.S.D. (n=5) with detection and quantification limits 0.5 and 2.0 ng l⁻¹, respectively. The calibration graph was rectilinear from 2 to 200 ng l⁻¹ (r=0.9996, n=6). Several other metal ions were determined in ideal situations, with analogous results.     

Spectrophtometric determination of malachite green in fish farming water samples after cloud point extraction using nonionic surfactant Triton X-100

A novel and sensitive cloud point extraction procedure for the determination of trace amounts of malachite green by spectrophotometry was developed. Malachite green was extracted at pH 2.5 mediated by micelles of nonionic surfactant Triton X-100. The extracted surfactant-rich phase was diluted with ethanol and its absorbance was measured at 630 nm. The effect of different variables such as pH, Triton X-100 concentration, cloud point temperature and time and diverse ions was investigated and optimum conditions were established. The calibration graph was linear in the range of 4-500 ng mL⁻¹ of malachite green in the initial solution with r = 0.9996 (n = 10). Detection limit based on three times the standard deviation of the blank (3S_b) was 1.2 ng mL⁻¹ and the relative standard deviation (R.S.D.) for 20 and 300 ng mL⁻¹ of malachite green was 1.48 and 1.13% (n = 8), respectively. The method was applied to the determination of malachite green in different fish farming and river water samples.     

Flow injection wetting-film extraction system for flame atomic absorption spectrometric determination of cadmium in environmental waters

A newly simple flow injection wetting-film extraction system coupled to flame atomic absorption spectrometry (FAAS) has been developed for trace amount of cadmium determination. The sample was mixed on-line with sodium diethyl dithiocarbamate and the produced non-charged Cd(II)-diethyl dithiocarbamate (DDTC) chelate complex was extracted on the thin film of diisobutyl ketone (DIBK) on the inner wall of the PTFE extraction coil. The wetting-film with the extracted analyte was then eluted by a segment of the cover solvent, and transported directly to the FAAS for evaluation. All the important chemical and flow parameters were optimized. Under the optimized conditions an enhancement factor of 35, a sample frequency of 22 h⁻¹ and a detection limit of c_L = 0.7 μg l⁻¹ Cd(II) were obtained for 60 s preconcentration time. The calibration curve was linear over the concentration range 1.5-45.0 μg l⁻¹ Cd(II) and the relative standard deviation, R.S.D. (n = 10) was 3.9%, at 10.0 μg l⁻¹ concentration level. The developed method was successfully applied to cadmium determination in a variety of environmental water samples as well as waste-water sample.     

A carbon nanotubes assisted strategy for insulin detection and insulin proteolysis assay

The carbon nanotubes (CNTs) assisted strategy has been proposed for insulin sensing and insulin proteolysis analysis. Experiments demonstrated that this strategy could be used for trace insulin determination with a low detection limit 7.75 ng mL⁻¹ (S/N = 3) and a detection range from 20 ng mL⁻¹ to 400 ng mL⁻¹. Both biocompatibility and intrinsic conductivity of pristine CNTs enabled them to act an excellent biosensing platform for the realization of direct electrochemistry and electrocatalysis of insulin. Compared with the present methods, the proposed strategy could realize the trace insulin detection without electrode modifications. It is more convenient and simpler than those based on the chemically modified electrodes. This method also made the CNTs as the indicator for insulin proteolysis analysis so that the biological process could be studied by electron microscope, electrochemical methods and digital camera. CNTs obtained after the proteolysis showed the same capabilities as the pristine ones in electrochemical signal enhancement and could participate in the bio-circle repeatedly.     

Determination of amino acids in Sargassum fusiforme by high performance capillary electrophoresis

High performance capillary electrophoresis (HPCE) was developed for quantitative determination of 18 phenylthiohydantoin (PTH)-amino acids. The influence of electrolyte concentration, pH, organic modifier and applied voltage on HPCE performance was investigated. The HPCE separation of a PTH-amino acids mixture was much improved by adding organic modifier and Tris-boric acid buffer to the run buffer. After optimization of the method, 17 PTH-amino acids in a solution containing 18 PTH-amino acids could be separated using 400 mmol l⁻¹ Tris-boric acid, 1.0 mmol l⁻¹ diethylamine at pH 9.5 adjusted with 0.1 mol l⁻¹ NaOH as a run buffer, voltage of 25 kV was applied, temperature was maintained at 25 °C, detection wavelength was 254 nm. The precision (n = 7) of this method is less than 3.2% (peak area) and 1.1% (migration time) of relative standard deviation (R.S.D.). Linearity was established over the concentration range 50-1000 μM of each derivative, with correlation coefficients (r) ranging between 0.9904 and 0.9993. The detection limits (S/N = 3) range from 2 to 48 μmol l⁻¹. The method was applied to determine amino acids in Sargassum fusiforme, a marine algae collected from Tongtou County of Zhejiang Province in China with satisfactory results.     

Evaluation of a magnetic polysulfone microcapsule containing organic modified montmorillonite as a novel solid-phase extraction sorbent with chlorophenols as model compounds

A porous polysulfone microcapsule containing organic modified montmorillonite and magnetic nanoparticles (OMMT-Fe₃O₄@PSF) has been successfully prepared by a phase-inversion method and evaluated as a magnetic solid-phase extraction (MSPE) sorbent for clean-up and enrichment of 4-chlorophenol (4-CP) and 2-chlorophenol (2-CP) in aqueous samples. Compared with a microcapsule containing the conventional extraction sorbent C18 (C18-Fe₃O₄@PSF), OMMT-Fe₃O₄@PSF had much lower cost, a faster adsorption rate, and superior uptake amounts for the investigated analytes. The proposed microcapsule has been developed for the extraction of 4-CP and 2-CP from environmental water samples and their analysis by high-performance liquid chromatography with UV detection (HPLC-UV). Various parameters, such as pH, extraction time, the mass of sorbent, and the desorption conditions, have been evaluated and the calibration curves of the chlorophenols were linear (R² ≥ 0.9985) in the range from 1.01 to 104.5 ng mL⁻¹. The limits of detection at a signal-to-noise (S/N) ratio of 3 were 0.22 and 0.17 ng mL⁻¹ and the limits of quantification calculated at S/N = 10 were 1.52 and 1.07 ng mL⁻¹ for 2-CP and 4-CP, respectively. The recoveries of 2-CP and 4-CP from natural water and the treated wastewater samples were in the range of 84.4-115% with relative standard deviations (RSDs) lower than 7.0%. The results have demonstrated the suitability of the MSPE approach for the analysis of trace chlorophenols in aqueous samples.     

Improvement of analytical performances for mercury speciation by on-line derivatization, cryofocussing and atomic fluorescence spectrometry

A modified automated on-line hyphenated system for simultaneous inorganic ionic mercury (Hg²⁺) and monomethylmercury (MeHg⁺) analysis by hydride generation (HG) or ethylation (Eth), cryofocussing, gas chromatography (GC) separation and atomic fluorescence spectrometry (AFS) detection has been improved. Both derivatization methods are investigated with respect to the chromatographic and analytical performances. They can be both affected by interferences when the AFS detection system is used. Water vapor removal using a soda lime moisture trap improves significantly the chromatographic performances, the reproducibility and the detection limits for Hg²⁺ and MeHg⁺ analyzed with both methods. For ethylation (Eth) derivatization, a scattering interference generated from low-quality ethylation reagent has also been eliminated. For HG, improved detection limits are 0.13 ng l⁻¹ and 0.01 ng l⁻¹ for Hg²⁺ and MeHg⁺, respectively (0.1 l water sample), and reproducibility are 5% for Hg²⁺ (20 ng l⁻¹) and MeHg⁺ (5 ng l⁻¹). Improved detection limits for Eth are 0.22 ng g⁻¹ for Hg²⁺ and 0.02 ng g⁻¹ for MeHg⁺ (1 g dry sediment sample) and the reproducibility are 5-6% for Hg²⁺ and MeHg⁺ (1-2 ng g⁻¹).     

Preconcentration and speciation of chromium in drinking water samples by coupling of on-line sorption on activated carbon to ETAAS determination

An on-line flow injection (FI) preconcentration-electrothermal atomic absorption spectrometry (ETAAS) method is developed for trace determination of chromium in drinking water samples by sorption on a conical minicolumn packed with activated carbon (AC) at pH 5.0. The chromium was removed from the minicolumn with 1.0% (v/v) nitric acid. An enrichment factor (EF) of 35-fold for a sample volume of 10 ml was obtained. The detection limit (DL) value for the preconcentration method proposed was 3.0 ng l⁻¹. The precision for 10 replicate determinations at the 0.5 μg l⁻¹ Cr level was 4.0% relative standard deviation (R.S.D.), calculate with the peak heights obtained. The calibration graph using the preconcentration system for chromium was linear with a correlation coefficient of 0.9992 at levels near the detection limits up to at least 50 μg l⁻¹. The method was successfully applied to the determination of Cr(III) and Cr(VI) in drinking water samples.