Novel solid-phase extraction and preconcentration technique coupled with ICP-AES for the determination of Cr(III), Ni(II), and Zn(II) in various water samples

Newly synthesized 2-propylpiperidine-1-carbodithioate (2-PPC) was used for the extraction of Cr(III), Ni(II), and Zn(II) from various water samples. In the present investigation, the use of a syringe loaded with sorbent for the separation and enrichment of Cr(III), Ni(II), and Zn(II) prior to their determination by inductively coupled plasma-atomic emission spectrometry (ICP-AES) was proposed to substitute the batch and column techniques. The described method was compared with the column technique with respect to fastness, simplicity, recovery, and risk of contamination. The syringe was loaded with 1.0 g of sorbent in order to retain the analyte elements. Next, 7.0 mL of sample solution (pH 5.0 ± 0.2) was drawn into the syringe in 15 s and discharged over 15 s. Then, an eluent (3.0 M HCl) was drawn into the syringe and ejected back to desorb the analyte elements. At the optimum conditions, the percentage recoveries of Cr(III), Ni(II), and Zn(II) were in the range of 94.50 to 99.62% with a standard deviation (S.D.) of 0.03%. The elements could be concentrated by drawing and discharging several portions of sample successively and eluting only one time. The detailed study of various interferences proved the method to be highly selective. The risk of contamination is less than that with the column technique. The method was successfully applied to the determination of Cr(III), Ni(II), and Zn(II) in spiked and natural water samples. The results obtained are in good agreement with those obtained by the reported methods at the 95% confidence level. The text was submitted by the authors in English.     

Preconcentration and separation of copper(II), cadmium(II) and chromium(III) in a syringe filled with 3-aminopropyltriethoxysilane supported on silica gel

In this study, a syringe was filled with silica gel loaded with 3-aminopropyltriethoxysilane, for the separation and preconcentration of copper, cadmium and chromium prior to their determination by graphite furnace atomic absorption spectrometry (GFAAS) in seawater. For this purpose, a syringe was filled with 0.5 g of modified silica gel and the sample solution was drawn into the syringe and ejected back again. The analyte elements were quantitatively retained at pH 5. Then, the elements sorbed by the silica gel were eluted with 2.0 M of HCl and determined by GFAAS. At optimum conditions, the recovery of Cu, Cd and Cr were 96-98%. Detection limits (3delta) were 6.6, 7.5 and 6.0 micro g L(-1) for Cu, Cd and Cr, respectively. The elements could be concentrated by drawing and discharging several portions of sample successively but eluting only once. Cu, Cd and Cr added to a seawater sample were quantitatively recovered (>95%) in the range of the 95% confidence level. The method proposed in this paper was compared with a column technique. Optimum experimental conditions, reproducibility, precision and recoveries of both techniques are the same, but the syringe technique is much faster, easier and more practical than the column technique. It is a portable system and allows one to make the sorption process in the source of sample. In addition, the risk of contamination is less than in the column technique.     

A novel preconcentration method for determination of iron and lead using Chromosorb-103 and flame atomic absorption spectrometry

Syringe connected-minicolumn (SCM) method for the preconcentration/separation of iron and lead prior to their determination by flame atomic absorption spectrometry (FAAS) was developed. The proposed method is substituted for classical batch and column methods. The method proposed was compared with the column method with respect to easiness, rapidness, simplicity and some analytical performance criteria such as recovery, precision, accuracy and risk of contamination. A minicolumn was filled with Chromosorb-103 as a sorbent and connected with a syringe. The experimental conditions such as pH of sample, concentration and amount of eluent, flow rates of sample and eluent, concentration of 8-hydroxyquinoline (oxine) as a complexing agent were optimized. The sample solution with or without oxine was drawn into the syringe and discharged manually passing through the resin. Analytes were retained at pH ≥5 and eluted with 1 M nitric acid in acetone quantitatively. Analyte elements could be concentrated up to 200-fold. The iron and lead in river-water samples and in certified reference Bovine Liver (NIST 1577b) were quantitatively recovered with relative standard deviation lower than 10%.     

Determination of lead and nickel in Apple-Leaves and sea-water by electrothermal atomic absorption spectrometry after solid-phase extraction using Chromosorb-107 filled in a syringe

Lead and nickel were concentrated and separated after sorbing on Chromosorb-107 filled in a syringe prior to their determination by electrothermal atomic absorption spectrometry. To retain the analytes, the sample solution treated with or without ammonium pyrolidine dithiocarbamate (APDC) was drawn into the syringe filled with Chromosorb-107 and discharged back manually. Then the analyte elements were eluted by drawing and discharging the suitable eluent. The optimum experimental conditions for quantitative recoveries of analytes (amount of sorbent, pH of sample, concentration and kind of eluent, flow rates of sample and eluent and adding of the effect of complexing agent) were investigated. Nickel was quantitatively retained at pH6 irrespective of whether it was complexed with APDC while the quantitative sorption of lead was achieved at pH8 only if it was treated with APDC prior to passing through the sorbent. The lead and nickel retained on the sorbent were eluted drawing and discharging 4.5 M HNO(3), respectively. If the concentration of analyte elements in the sample were too low to be detected, then they were concentrated by increasing the ratio of sample value to eluent volume as needed. The lead and nickel in spiked sea-water samples and in certified reference Apple-Leaves (NIST SRM 1515) standards were quantitatively (95%) recovered with R.S.D. of around +/-2%.     

Determination of bismuth and cadmium after solid-phase extraction with chromosorb-107 in a syringe

The determination of bismuth and cadmium by graphite furnace atomic absorption spectrometry (GFAAS) after solid-phase extraction (SPE) on Chromosorb-107 filled in a syringe was described. To retain the analytes, the sample solution treated with and without ammonium pyrolidine dithiocarbamate (APDC) was drawn into the syringe filled with Chromosorb-107 and discharged back manually. Bismuth and cadmium were quantitatively sorbed at pH ≥ 6 irrespective of whether the analyte was complexed with APDC prior to passing through the Chromosorb-107. Analyte elements sorbed on the resin were quantitatively eluted with 3.0 M of HNO₃ again drawing and discharging the eluent into the syringe and ejected it back. Optimum flow rates of sample or eluent for sorption and elution processes were 20 ml min⁻¹ for drawing and 20 ml min⁻¹ for discharging in all cases. Bismuth and cadmium were analyzed by graphite furnace atomic absorption spectrometry. The elements could be concentrated by drawing and discharging several portions of sample successively but eluting only one time. The validity of the proposed method was checked with standard reference materials (NIST SRM 1515 Apple-Leaves, CWW-TM-E Waste Water and CRM-SW Sea Water). The analyte elements were quantitatively (>95%) recovered from different matrices irrespective of treated samples with APDC. Detection limits (δ) were 0.8 and 1.2 μg l⁻¹ for Bi and Cd, respectively. The method can be characterized with fastness, simplicity, quantitative recovery and high reproducibility.     

Determination of some trace metals in waters by flame atomic absorption spectrometry after preconcentration on Amberlite XAD-16 resin with sodium tetraborate

A method was described for the determination of the elements Cr, Mn, Fe, Co, Ni, Cu, Cd, Pb, and Bi in waters by flame atomic absorption spectrometry (FAAS) after separation and preconcentration on Amberlite XAD-16 resin with sodium tetraborate using a chromatographic column. Parameters influencing the analytical performance, including pH and the volume of sample, amount of analyte and interfering effect of co-existing ions, were studied in detail. The recovery values were quantitative (> or = 95%), and the relative standard deviation (RSD) and detection limit (DL) varied in the range of 1.1-2.4% (n=10) and 0.002-0.177 microg m(-1) (3s, n=20), respectively. After being optimized, the proposed method was applied to the drinking water, waste water and artificial sea water samples. Recovery values of the elements investigated, were quantitative for tap water and synthetic sea water, except for Mn, Co and Ni (including also Cd for synthetic sea water). Recovery values of Cd, Pb, Cu and Co were found to be 95, 102, < or = 87, and < or = 83%, respectively, for the waste water samples.     

X-ray fluorescence determination of As, Bi, Co, Cu, Fe, Ni, Pb, Se, V and Zn in natural water and soil extracts after preconcentration of their pyrrolidinedithiocarbamates on cellulose filters

A simple method for the elements preconcentration on thin-layer paraffin-treated cellulose filters was proposed. It was found that pyrrolydinedithiocarbamates of As(III), Bi, Cd, Co, Cu, Fe(III), Ni, Pb, Se(IV), V(V) and Zn obtained after mixing of sample (3-5 ml min⁻¹) and reagent (0.7-1.0 ml min⁻¹) streams were quantitatively recovered from 100 ml sample. The sample acidity was adjusted to pH 4.8-5.2 for preconcentration of Cd, Co, Cu, Fe(III), Ni, Pb, V(V) and Zn, and to 2 M HCl for preconcentration of As, Bi and Se. The optimum reagent concentration was found to be 0.1%. The elements were determined on the filters by X-ray fluorescence spectrometry. The detection limits achieved were 0.1-4.0 μg of element on the filter. Relative standard deviation (R.S.D.) was not higher than 0.08 while determining 5-50 μg of elements on filter. Accuracy and precision of the technique proposed were evaluated by the analysis of spiked natural samples.     

Sol–gel/nanoclay composite as a sorbent for microextraction in packed syringe combined with corona discharge ionization ion mobility spectrometry for the determination of diazinon in water samples

In this work, the microextraction in packed syringe technique combined with corona discharge ion mobility spectrometry was used for determining diazinon in water samples. A new porous composite of nanoclay and polysiloxane was prepared using a sol–gel process. An amount of 2.0 mg of the sorbent was packed in a 250 μL syringe and used for extraction. A volume of 2 mL of the sample was passed through the sorbent bed, and the entrapped analyte was eluted by 25 μL of methanol. Important parameters influencing the extraction performance were investigated. Under optimum experimental conditions, the detection limit for diazinon was 0.07 ng/mL. The intra‐ and inter‐day relative standard deviations were 5.0 and 12.3%, respectively. The calibration curve was linear in the concentration range from 0.2 to 20.0 ng/mL (r² = 0.999). The applicability of the method was demonstrated by analyzing spiked real water samples and the spiking recoveries were in the range of 95 to 106%.     

Study of the factors affecting the performance of microextraction by packed sorbent (MEPS) using liquid scintillation counter and liquid chromatography-tandem mass spectrometry

Microextraction by packed sorbent (MEPS) is a new technique for sample preparation that can be connected on-line with LC or GC. In MEPS, approximately 1-2 mg of the solid packing material is inserted into a syringe (100-250 μL) as a plug. Sample preparation takes place on the packed bed. The bed can be packed or coated to provide selective and suitable sampling conditions. The new method is very promising for extraction of drugs and metabolites from biological samples.In this paper, some factors affecting the performance of MEPS such as recovery, carry-over, leakage, washing volume and elution volume were studied using C18 and hydroxylated polystyrene-divinylbenzene copolymer (ENV+) as sorbents. Radioactively labelled bupivacaine in plasma samples was used as test analyte. For the extraction of this drug, using methanol/water 95:5 (v/v) (0.25% ammonium hydroxide) was used as elution solvent. The analyte response increased with increasing the elution volume and it was linear upp up to 100 μL utilizing liquid scintillation counter. Further, for concentrating the sample, we found that MEPS may be used such that the sample can be drawn through the needle, up and down, several times. The analyte leakage increases as the volume washing increases, though higher washing volumes may also result in cleaner extracts. To eliminate analyte carry-over, the sorbents were washed first with 3 × 250 μL elution solution and then with 3 × 250 μL washing solution. In addition, the reproducibility measurements show relatively good relative standard deviation (RSD) % values concerning analyte recovery and analyte leakage. The present study provides an understanding of basic aspects when optimizing methods for MEPS. In this study, MEPS was used off-line with liquid scintillation counter and on-line with LC-MS/MS.     

Molybdenum(VI) Oxide-Modified Silica Gel as a Novel Sorbent for the Simultaneous Solid-Phase Extraction of Eight Metals with Determination by Flame Atomic Absorption Spectrometry

A silica-based inorganic sorbent was synthesized by the thermal decomposition of ammonium heptamolybdate on silica and applied for the preconcentration and simultaneous determination of Cd, Co, Cr, Cu, Fe, Mn, Ni, and Pb in river water samples using a column system with flame atomic absorption spectrometry. Attenuated total reflection-Fourier transformation infrared spectroscopy, scanning electron microscopy, and electron dispersive spectroscopy were used for sorbent characterization. The effects of pH, sample volume, eluent type, eluent concentration, eluent volume, sample flow rate, and matrix ions (Al, Bi, Ca, Mg, and Zn) on the recovery of the metals in model solutions were investigated. The adsorption capacities (µmol g^?1) of SiO₂-MoO₃ were 88.96 (Cd), 169.69 (Co), 153.85 (Cr), 188.88 (Cu), 179.05 (Fe), 163.81 (Mn), 136.31 (Ni), and 38.61 (Pb). The detection limits of the method were 9.09, 10.82, 10.77, 49.57, 31.64, 6.40, 8.86, 19.15?µg L^?1 for Cd, Co, Cr, Cu, Fe, Mn, Ni, and Pb, respectively, with a preconcentration factor of 25. The developed method was used for the determination of the target metals in real samples and the recoveries for spiked samples were found to be from 91.2% to 102.9%.     

Preconcentration of trace elements by partial dissolution of the matrix—multielement preconcentration from manganese

Trace elements such as Ag, Au, Bi, Cd, Co, Cu, Fe, Ga, In, Ni, Pb, Pd and Tl, can be preconcentrated with recoveries of better than 95% from high-purity manganese if the sample is coated with a thin layer of mercury before its dissolution in HCl to a small residue. For determination of the trace elements, the residue is completely dissolved in aqua regia. After separation of the mercury by reductive precipitation, Bi, Cd, Co, Cu, Fe, Ga, In, Ni, Pb, and Tl are determined by flame a.a.s. ('injection method'). Ag, Au and Pd are determined in the mercury-containing solution. The relative standard deviation was usually about 5%; Cu, Fe and Pb were proved to be inhomogeneously distributed in the sample (electrolytically produced manganese), and so the standard deviations were considerably greater. The detection limits for the different elements were between 0.6 and 0.004 ppm, depending on the sensitivity of their a.a.s. determination. The theoretical basis of this preconcentration method and its applicability to multielement analysis of different high-purity metals are discussed.     

Liquid chromatographic analysis of oxcarbazepine and its metabolites in plasma and saliva after a novel microextraction by packed sorbent procedure

A rapid and reliable analytical method suitable for the simultaneous determination of the antiepileptic drug, oxcarbazepine and its metabolites in human plasma and saliva by means of liquid chromatography with diode array detection (DAD) has been developed. Oxcarbazepine and its metabolites (10,11-dihydro-10-hydroxycarbamazepine, trans-10,11-dihydro-10,11-dihydroxycarbamazepine and 3-hydroxycarbamazepine) were baseline separated within 6.5 min on a reversed-phase C18 column with a phosphate buffer-acetonitrile-triethylamine mixture as the mobile phase. The DAD detector was set at 240 nm. A sample preparation method for biological samples using a microextraction by packed sorbent technique has been implemented, employing a C18 sorbent inserted into a microvolume syringe and using only a small volume (25 μL) of plasma or saliva. The extraction yield values were satisfactory for all analytes (>86.5%) as well as the precision data, which were always in the low percentage of relative standard deviation values (<4.6%). The method was successfully applied to both plasma and saliva samples drawn from psychiatric and neurological patients undergoing treatment with oxcarbazepine (Tolep^®) tablets.     

Semiquantitative Analysis of Biological Materials by Inductively Coupled Plasma–Mass Spectrometry

Semiquantitative analysis with accuracy of ±30 to 50% is a valuable tool for rapid screening of samples prior to quantitative determination of trace metals. In this study semiquantitative analysis software available with commercial inductively coupled plasma–mass spectrometry (ICP-MS) instrumentation is applied for rapid multielemental analysis, and the accuracy and precision of this semiquantitative analysis approach is evaluated with biological certified reference materials. Samples were prepared by high-pressure, high-temperature nitric acid vapor-phase digestion. For most elements the measured semiquantitative results are in the range of the certified values. With appropriate analyte solution dilution, the measured concentrations of the major elements (e.g., Ca) also agree with certified values. The accuracy is within ±10% for 28 element determinations that include 16 individual elements (Ag, As, Cd, Co, Cr, Cu, Fe, Mn, Mo, Ni, Pb, Rb, Sb, Sr, Tl, and Zn) and ±20% for 54 element determinations that include three more elements (Mg, V, and U) in eight certified reference materials including water. The method precision is 11 ± 11% (relative standard deviation,n= 65).     

A new electrolytic dissolution of aluminum alloys and determination of some constituents by inductively coupled plasma optical emission spectrometry

The new electrolytic dissolution in batch of aluminum alloys samples as grains or turns and the determination of Fe, Cu, Mn, Mg, Cr, Ni, Zn, Pb and Ti by ICP OES was investigated. In on-line electrodissolution procedures described in the literature, samples were restricted to be in the form of solid blocks or plates with one polished flat face. Here, the sample was loaded in the barrel of a modified disposable syringe (the anodic semi-cell) and pressed with a modified plunger fitted with a platinum disk to establish electrical contact with the analyte. This arrangement was introduced in a beaker containing the electrolyte (1 mol L⁻¹ HNO₃) and a platinum wire as the cathode. The resulting solution from electrodissolution (0.6 A) was used for the ICP OES determinations. The influence of the aluminum concentration increase on the determination of the elements was evaluated. Electrodissolution of certified reference materials and commercial samples revealed relative errors lower than 10% for the elements Fe, Cu, Mg, Ni, Cr, Zn and Ti (when their content is above 0.1%). Higher inaccuracies (>10%) were observed for Mn and for Fe in B.C.S. 268/1 reference material certified. The proposed method presented a relative standard deviations (R.S.D.) lower or circa 10% to all of the elements (except Pb). In comparison with traditional acid dissolution, the proposed electrodissolution method is relatively fast (about 30 min), it is clean (there is no projection of solution) and simple (heating and fumes exhaust system were not necessaries).     

Column solid-phase extraction with Chromosorb-102 resin and determination of trace elements in water and sediment samples by flame atomic absorption spectrometry

A column, solid-phase extraction (SPE), preconcentration method was developed for determination of Bi, Cd, Co, Cu, Fe, Ni and Pb ions in drinking water, sea water and sediment samples by flame atomic absorption spectrometry. The procedure is based on retention of analytes in the form of pyrrolidine dithiocarbamate complexes on a short column of Chromosorb-102 resin from buffered sample solution and then their elution from the resin column with acetone. Several parameters, such as pH of the sample solution, amount of Chromosorb-102 resin, amount of ligand, volume of sample and eluent, type of eluent, flow rates of sample and eluent, governing the efficiency and throughput of the method were evaluated. The effects of divers ions on the preconcentration were also investigated. The recoveries were >95%. The developed method was applied to the determination of trace metal ions in drinking water, sea water and sediment samples, with satisfactory results. The 3σ detection limits for Cd, Cu, Fe, Ni and Pb and were found to be as 0.10, 0.44, 11, 3.6, and 10 μg l⁻¹, respectively. The relative standard deviation of the determination was <10%. The procedure was validated by the analysis of a standard reference material sediment (GBW 07309) and by use of a method based on coprecipitation.     

全谱直流电弧发射光谱法同时测定钼样品中17种杂质元素

为了解决现有测定方法中样品前处理复杂、测定过程繁琐的问题,研究了直流电弧发射光谱法测定钼样品中Al、Bi、Ca、Cd、Co、Cr、Cu、Fe、Mg、Mn、Ni、Pb、Sb、Si、Sn、Ti、V等17种元素的方法。用石墨粉(98%)、碳酸钠(1%)和氟化钠(1%)作为光谱缓冲剂抑制基体元素蒸发、稳定弧焰、促进杂质元素的蒸发;采用样品装满电极后下压2 mm和滴加蔗糖乙醇水溶液(2%)的方式,防止喷溅的发生,稳定激发过程,减小基体干扰;对挥发难易程度不同的元素采用信号分段采集的方式,提高各元素信噪比。方法灵敏度高、分析速度快、操作简便,加标回收率为85%~120%,各元素分析精密度均小于15%。     

The Enrichment/Separation of Fe, Co, Pb, Cd, and Cr on Ambersorb 563 Prior to Their Flame Atomic Absorption Spectrometric Determinations

A sorbent extraction method for the separation/preconcentration of Fe, Co, Pb, Cd, and Cr was developed. The analyte metal ions were retained on a column of Ambersorb 563 from a buffered sample solution. The flow rates of the sample and eluent solution were controlled by a peristaltic pump. The analyte ions were quantitatively retained at pH 9 by using an ammonia/ammonium chloride buffer solution, and were then eluted with 5 mL of 0.25 M HNO₃ at 5 mL/min flow rate. The detection limits were in the range of 0.33 and 72 g/L for Cd and Pb, respectively. The relative standard deviations were less than 10%. Recoveries of spike addition to drinking water and seawater were quantitative. The method presented was applied for the determination of Fe, Co, Pb, Cd, and Cr ions in drinking and seawater samples with satisfactory results (recoveries >95%).     

Ligandless cloud point extraction of Cr(III), Pb(II), Cu(II), Ni(II), Bi(III), and Cd(II) ions in environmental samples with Tween 80 and flame atomic absorption spectrometric determination

A cloud point extraction (CPE) procedure has been developed for the determination trace amounts of Cr(III), Pb(II), Cu(II), Ni(II), Bi(III), and Cd(II) ions by using flame atomic absorption spectrometry. The proposed cloud point extraction method was based on cloud point extraction of analyte metal ions without ligand using Tween 80 as surfactant. The surfactant-rich phase was dissolved with 1.0 mL 1.0 mol L⁻¹ HNO₃ in methanol to decrease the viscosity. The analytical parameters were investigated such as pH, surfactant concentration, incubation temperature, and sample volume, etc. Accuracy of method was checked analysis by reference material and spiked samples. Developed method was applied to several matrices such as water, food and pharmaceutical samples. The detection limits of proposed method were calculated 2.8, 7.2, 0.4, 1.1, 0.8 and 1.7 μg L⁻¹ for Cr(III), Pb(II), Cu(II), Ni(II), Bi(III), and Cd(II), respectively.     

Anreicherung und spektrochemische Bestimmung von Spurenelementen in reinsten Thalliumpr?paraten

Zusammenfassung Zur Anreicherung von Spurenelementen aus Lösungen reinster Thalliumpräparate wird die Matrix Thallium als TlJ ausgefällt. Dabei bleiben die Spuren Mg, Zn, Al, Ga, In, Fe, Co, Ni, Mn, Sn, La, Ce, Th und U quantitativ in Lösung, während Bi, Cu, Pb und Cd sowie die Edelmetalle Ag, Hg, Pd, Au und Pt vom Niederschlag mitgerissen werden. Die Adsorption der für die Reinheitsprüfung des Thalliums wichtigen Elemente Bi, Cu, Pb und Cd kann verhindert werden, wenn sie mit Hilfe von ÄDTA in Komplexanionen übergeführt werden. Als Fällungsmittel wird NaJ verwendet. Zur spektrochemischen Bestimmung werden Pb, Cd, Bi, Ni, Co und In als DDTC-Komplexe mit Chloroform extrahiert. Nach Zusatz einer Mischung von Lithiumcarbonat-Kohle und den Referenzelementen Sn und Pd wird der Eindampfrückstand im Gleichstrombogen analysiert. Cu und Fe werden photometrisch bestimmt.

---

Enrichment and spectrochemical determination of trace elements in high-purity thallium preparations

For the enrichment of trace elements in solutions of high-purity thallium samples, thallium is precipitated in form of TlI. In this process the noble metals Ag, Hg, Pd, Au, Pt and also Bi, Cu, Pb and Cd will be co-precipitated, whereas many other elements (Mg, Zn, Al, Ga, In, Fe, Co, Ni, Mn, Sn, La, Ce, Th, U) will remain completely in solution. The adsorption of the four elements Bi, Cu, Pb, Cd, which are important for the purity control of thallium compounds, can be prevented, if they are complexed with EDTA to form complex anions. The precipitating reagent is NaI. For the spectrochemical determination of Pb, Cd, Bi, Ni, Co, and In, the trace elements are extracted as DDTC complexes with chloroform. After the addition of a mixture of Li₂CO₃, carbon, and the reference elements Sn and Pd, the sample is analysed, using the d.c. arc. Cu and Fe are determined spectrophotometrically.

---

---

Wir danken dem Fonds der chemischen Industrie für die finanzielle Unterstützung dieser Arbeit.     

On-line preconcentration and GFAAS determination of trace metals in waters

An on-line separation preconcentration system coupled to electrothermal (graphite furnace) atomic absorption spectrometry was developed. A miniature column packed with iminodiacetic acid ethyl cellulose (IDAEC) was inserted into the loop. A peristaltic pump was used to deliver solutions. A flow of air was driven into the packed column, evacuating it between sample loading, washing and elution. The retained analyte was introduced on-line to graphite furnace using countercurrent elution with HNO₃. The system was applied for the determination of V, Co and Pb in medicinal mineral water samples, and nickel in sea water samples. The detection limits (3σ) were 0.058, 0.022, 0.067, 0.062 μg/l for Co, Pb, V, and Ni, respectively. The R.S.D. (n=5) was <5% at 0.4–1.0 μg/l concentration range.