Automated at-line solid-phase extraction-gas chromatographic analysis of micropollutants in water using the PrepStation

A fully automated at-line solid-phase extraction-gas chromatography procedure has been developed for the analysis of aqueous samples using the PrepStation. The sample extract is transferred from the sample preparation module to the gas chromatograph via an autosampler vial. With flame-ionization detection, limits of determination (S/N=10) of 0.05–0.13 μg/l were obtained for the analysis of HPLC-grade water when modifying the PrepStation by: (i) increasing the sample volume to 50 ml, (ii) increasing the injection volume up to 50 μl, and (iii) decreasing the desorption volume to 300 μl. The HP autosampler had to be modified to enable the automated “at-once” on-column injection of up to 50 μl of sample extract. The amount of packing material in the original cartridge had to be reduced to effect the decrease of the desorption volume. The total set-up did not require any further optimization after having set up the method once. The analytical characteristics of the organonitrogen and organophosphorus test analytes, i.e. recoveries (typically 75–105%), repeatability (2–8%) and linearity (0.09–3.0 μg/l) were satisfactory. The potential of the system was demonstrated by determining triazines and organophosphorus pesticides in river Rhine water at the 0.6 μg/l level using flame-ionization and mass-selective detection. No practical problems were observed during the analysis of more than 100 river water samples.     

Determination of trace level perchlorate in drinking water and ground water by ion chromatography

Ammonium perchlorate, a key ingredient in solid rocket propellants, has recently been found in ground and surface waters in the USA in a number of states, including California, Nevada, Utah, and West Virginia. Perchlorate poses a health risk and preliminary data from the US Environmental Protection Agency reports that exposure to less than 4–18 μg/l provides adequate human health protection. An ion chromatographic method was developed for the determination of low μg/l levels of perchlorate in drinking and ground waters based on a Dionex IonPac AS11 column, a 100 mM hydroxide eluent, large loop (1000 μl) injection, and suppressed conductivity detection. The method is free of interferences from common anions, linear in the range of 2.5–100 μg/l, and quantitative recoveries were obtained for low μg/l levels of perchlorate in spiked drinking and ground water samples. The method detection limit of 0.3 μg/l permits quantification of perchlorate below the levels which ensure adequate health protection. A new polarizable anion analysis column, the IonPac AS16, and its potential applicability for this analysis is also discussed.     

Silicon determination by inductively coupled plasma atomic emission spectrometry after generation of volatile silicon tetrafluoride

A method for the determination of silicon by inductively coupled plasma atomic emission spectrometry (ICP-AES) is described. The procedure is based on a discontinuous generation of volatile silicon tetrafluoride in concentrated sulphuric acid medium after injecting 125 μl of 0.1%, w/v sodium fluoride solution into 100 μl of the sample. The gaseous silicon tetrafluoride is fed directly into the ICP torch by a flow of 250 ml min⁻¹ Ar carrier gas. The calibration curve was linear up to at least 100 μg ml⁻¹ of Si(IV) and the absolute detection limit was 9.8 ng working with a solution volume of 100 μl. The relative standard deviation for six measurements of 10 μg ml⁻¹ of Si(IV) was 2.32%. The method was applied to the determination of silicon in water and iron ores.     

Speciation of halogenides and oxyhalogens by ion chromatography-inductively coupled plasma mass spectrometry

A speciation method utilizing ion chromatography coupled with inductively coupled plasma mass spectrometry is described for simultaneous analysis of eight halogenides and oxyhalogens: chloride, chlorite, chlorate, perchlorate, bromide, bromate, iodide and iodate. The method was applied for the analysis of drinking water samples collected from water treatment plants in areas in Finland, which are known to have high bromine concentrations in ground water. Water samples collected before and after disinfection were analyzed to get information about potential species conversion as a result of purification. Chloride and chlorate were the chlorine species found in these water samples, and iodine existed as both iodate and iodide. In the case of bromine, species conversion had taken place, since total bromine concentrations were increased during disinfection but bromide concentrations were decreased. No bromate was observed in the samples. The detection limits for all the chlorine species studied were 500 μg/l, for bromine species studied 10 μg/l, for iodide 0.1 μg/l and for iodate 0.2 μg/l.     

Direct analysis of reference biofluids by coupled in situ electrodeposition-electrothermal atomic absorption spectrometry

The application of coupled in situ electrodeposition-electrothermal atomic absorption spectrometry (ED-ETAAS) to the determination of Pb in biological standard reference materials is described. In situ electrodeposition at a cell voltage of 3.0 V from 25-μl samples onto electrodeposited Pd is used to quantitatively separate the analyte from blood and urine matrices. With subsequent withdrawal of spent electrolyte, this overcomes the atomisation problems inherent with high salt and organic contents. ED-ETAAS is applied with minimal sample pre-treatment (acidification). The electrolysis process aids decomposition of the organic matrix, and the release of trace elements. Evolution of H₂ at the cathode counters fouling of the Pd modifier surface. The palladium deposit is renewed in situ for each determination. For AMI certified lyophilised blood, diluted 1+3 with 0.1 M HCl (18.1 μg/l Pb), the R.S.D. was 3.0% (peak height; n=5) and the detection limit (3 σ_blank; n=5) was 1.5 μg/l. Results for certified blood samples were AMI 72.3±4.3 μg/l (certified 76.2±7.6 μg/l) and Seronorm 34.2±2.0 μg/l (36±4 μg/l). The result for NIST SRM 2670 normal urine acidified to 1% HNO₃ was 8.1±0.6 μg/l (recommended value 10 μg/l).     

Sensitive and selective ion chromatographic method for the determination of trace beryllium in water samples

A selective and sensitive ion chromatographic method has been developed for the determination of beryllium in a number of water samples at low-μg/l concentrations. The separation was performed on a 250×4.0 mm I.D. iminodiacetic acid functionalised silica gel column. Chromatographed Be(II) was detected using visible detection at 590 nm following post-column reaction with chrome azurol S (CAS). The optimum separation and derivatisation conditions were studied in detail. The optimum eluent conditions were found to be 0.4 M KNO₃, adjusted to pH 2.5 using HNO₃, with optimum post-column detection being achieved using a solution containing 0.26 mM CAS, 2% Triton X-100, 50 mM 2-(N-morpholino)ethanesulfonic acid, pH 6.0. Under the above conditions, the concentration detection limit for Be(II) was found to be 3 μg/l in a standard solution and 4 μg/l in a typical tap water sample, using a 250 μl injection. The method was linear over the investigated range of 10 μg/l to 10 mg/l and highly reproducible. The method was successfully applied to a number of water samples of varying matrix complexity, including simulated seawater, and also to a natural freshwater certified reference material NIST 1640.     

Quantitative analysis of fluid inclusions in evaporites by ion chromatography

Natural salt minerals often contain inclusions of saturated salt solutions with diameters from 1 to> 100 μm. With the quantification of the composition of the fluid inclusions, the origin and metamorphism of the salt rocks can be interpreted. Hence, these data are important concerning the long-term safety of underground repositories in salt rocks [1]. For the extraction of the solutions in fluid inclusions with diameters 300 μm, an optical precision instrument was developed. For the simultaneous determination of Cl⁻, Br⁻, SO₄²⁻, Li⁺, Na⁺, K⁺, Mg²⁺ and Ca²⁺ two ion chromatographic systems with conductivity detection for cations and anions and additional photometric detection for Br⁻ were used. To prevent column overload, the Cl concentration must be less than 50 μg/ml in the measuring solution. The extracted samples (volumes> 0.1 μl) are diluted with demineralized water by a factor of 1 · 10⁴ (20-μl sample loops). The practical limit of determination for the measured elements is 0.01–0.3 μg/ml in the measuring solutions. By calculation of the anion and cation charge balance (molar equivalence), a relative error of <5% for the analysis of fluid inclusions was found.     

Determination of trimethylselenonium iodide, selenomethionine, selenious acid, and selenic acid using high-performance liquid chromatography with on-line detection by inductively coupled plasma mass spectrometry or flame atomic absorption spectrometry

An analytical method has been developed for the determination of selenious acid, selenic acid, trimethylselenonium ion, and selenomethionine. The four selenium compounds were separated by HPLC on a column (25 cm×4 mm I.D.) of the anion-exchanger ESA Anion III with a mobile phase (1.5 ml/min) of 0.0055 M ammonium citrate (pH 5.5). Detection was carried out using an on-line inductively coupled plasma mass spectrometer (ICP-MS) or a flame atomic absorption spectrometer (FAAS) as the selenium-specific detector. The chromatographic parameters and the chemical factors affecting the separation of the selenium species were optimized. The four selenium compounds could be separated within 8 minutes. The detection limits of the coupled HPLC–FAAS system were approximately 1 mg Se/l for each compound (100 μl injection), estimated as three times the base-line noise of the chromatograms. More powerful selenium detection was achieved with an ICP-MS. Selenium was measured at m/z 78. To increase the nebulization efficiency, the Meinhard concentric glass nebulizer was replaced by an ultrasonic nebulizer. The ICP-MS signal intensity was increased with the ultrasonic nebulization by a factor of 7 times for selenious acid and 24 to 31 times for trimethylselenonium ion, selenomethionine, and selenic acid compared to that with the Meinhard nebulization. The detection limits achieved by the HPLC–ICP-MS with the ultrasonic nebulization were 0.08 μg Se/l for trimethylselenonium ion, 0.34 μg Se/l for selenious acid, 0.18 μg Se/l for selenomethionine, and 0.07 μg Se/l for selenic acid, respectively.     

Determination of fenpyroximate in apples by supercritical fluid extraction and packed capillary liquid chromatography with UV detection

A method using off-line supercritical fluid extraction (SFE) and micro liquid chromatography (μLC) with UV detection at 260 nm, was developed for selective determination of fenpyroximate in apple samples. The packed capillary liquid chromatography method utilises 20 μl injection volumes with on-column focusing. A 350×0.32 mm capillary column packed with Kromasil 100-C₁₈ of 5 μm particle size was used with a mobile phase of acetonitrile–10 mM ammonium acetate (85:15, v/v) at a flow of 5 μl/min. A two-step SFE procedure was used to extract fenpyroximate selectively in 2 g apple samples, with Hydromatrix (HMX) added as a water absorbent at a 1:1 (w:w) ratio. Fenpyroximate was extracted at 200 bar and 90°C for 15 min using carbon dioxide at a flow of 2 ml/min, and solvent trapping collection in 10 ml acetonitrile. The volume of the acetonitrile extract was reduced by evaporation and water was added to a final composition of acetonitrile–water (40:60, v/v). The resulting 2.0 ml solution was filtered using a 0.45 μm poly(vinylidene difluoride) syringe filter before μLC analysis. Validation of the method was accomplished with apple samples spiked with fenpyroximate, covering the range of 0.1 to 1.0 μg/kg. The within-day and between-day repeatabilities were in the range 4–18% relative standard deviation. Accuracy, measured as recovery, was found to be approximately 60%. Apple samples from a field treated with fenpyroximate were analysed. None of the samples contained fenpyroximate above the quantification level.     

Application of colloidal palladium modifier for the determination of As, Sb and Pb in a spiked sea water sample by electrothermal atomic absorption spectrometry

Colloidal palladium was used as a chemical modifier for analysis of complex samples by electrothermal atomic absorption spectrometry. In order to demonstrate high potential of the modifier, optimization of the time–temperature program of the atomizer was limited with only pyrolysis and atomization temperatures. Fixed palladium modifier masses were applied (6 μg for pure analyte solutions and 15 μg for matrix-containing solutions). It was shown that in the presence of colloidal palladium, interference-free determinations of As, Sb and Pb are possible up to at least 450 μg of chloride ion, or 40 μg of sulfate ion (as their sodium salts) in the atomizer. Colloidal palladium was used for the direct determination of As, Sb and Pb in a spiked sea water sample (from Bosphorus channel near Istanbul) by means of the calibration graphs prepared with pure analyte solutions. The detection limits for As, Sb and Pb in a sea water matrix calculated according to 2σ criteria are 5.4, 3.6 and 1.1 ng ml⁻¹, respectively (for sample volume 10 μl). In unspiked sea water, the contents of As, Sb and Pb were found to be below the detection limits. Recoveries of spiked analytes (25 and 50 ng ml⁻¹) were in the region of 98–112% depending on the nature of analyte and the concentration of spike.     

Simultaneous determination of inorganic disinfection by-products and the seven standard anions by ion chromatography

For the first time, an ion chromatographic method for the simultaneous determination of the disinfection by-products bromate, chlorite, chlorate, and the so-called seven standard anions, fluoride, chloride, nitrite, sulfate, bromide, nitrate and orthophosphate is presented. The separation of the ten anions was carried out using a laboratory-made high-capacity anion-exchanger. The high capacity anion-exchanger allowed the direct injection of large sample volumes without any sample pretreatment, even in the case of hard water samples. For quantification of fluoride, chloride, nitrite, sulfate, bromide, nitrate, orthophosphate and chlorate, a conductivity detection method was applied after chemical suppression. The post-column reaction, based on chlorpromazine, was optimized for the determination of chlorite and bromate. The method detection limit for bromate measured in deionized water is 100 ng/l and for chlorite, it is 700 ng/l. In hard drinking water, the method’s detection limits are 700 ng/l (bromate) and 3.5 μg/l (chlorite). The method’s detection limits for the other eight anions, determined by conductivity detection, are between 100 μg/l (nitrite) and 1.6 mg/l (chlorate).     

Determination of arsenic(III) and arsenic(V) by electrothermal atomic absorption spectrometry after complexation and sorption on a C-18 bonded silica column

A flow injection procedure for the separation and pre-concentration of inorganic arsenic based on the complexation with ammonium diethyl dithiophosphate (DDTP) and sorption on a C-18 bonded silica gel minicolumn is proposed. During the sample injection by a time-based fashion, the As³⁺-DDTP complex is stripped from the solution and retained in the column. Arsenic(V) and other ions that do not form complexes are discarded. After reduction to the trivalent state by using potassium iodide plus ascorbic acid, total arsenic is determined by electrothermal atomic absorption spectrometry (ETAAS). Arsenic(V) concentration can be calculated by difference. After processing 6 ml sample volume, the As³⁺-DDTP complexes were eluted directly into the autosampler cup (120 μl). Ethanol was used for column rinsing. Influence of pH, reagent concentration, pre-concentration and elution time and column size were investigated. When 30 μl of eluate plus 10 μl of 0.1% (w/v) Pd(NO₃)₂ were dispensed into the graphite tube, analytical curve in the 0.3–3 μg As l⁻¹ range was obtained (r=0.9991). The accuracy was checked for arsenic determination in a certified water, spiked tap water and synthetic mixtures of arsenite and arsenate. Good recoveries (97–108%) of spiked samples were found. Results are precise (RSD 7.5 and 6% for 0.5 and 2.5 μg l⁻¹, n=10) and in agreement with the certified value of reference material at 95% confidence level.     

Integrated system for on-line gas and liquid chromatography with a single mass spectrometric detector for the automated analysis of environmental samples

An integrated system has been developed which combines liquid (LC) and gas (GC) chromatographic separation with a single mass spectrometer (MS). On-line solid-phase extraction (SPE) of 10–200 ml aqueous samples on a short (10 × 2.0 mm I.D.) precolumn packed with a styrene-divinylbenzene copolymer is used for analyte enrichment. The trace-enrichment procedure was automated by means of a PROSPEKT cartridge-exchange/solvent-selection/valve-switching unit. After sample loading, the precolumn is eluted on-line in two subsequent runs, first onto the GC-MS system and, next, onto the LC-MS system using a particle beam (PB) interface. Prior to entering the PB-MS, the LC eluent passes through the flow cell of a UV diode-array detector (DAD). Both GC-MS and LC-PB-MS generate classical electron ionisation (EI) and chemical ionisation (CI) spectra which are useful for the identification of low- and sub-μg/l concentrations of environmental pollutants covering a wide polarity and volatility range. The LC-DAD data provide additional means for quantitation and yield complementary spectral information. All three detection systems (GC-MS, LC-DAD, LC-PB-MS) and the trace-enrichment procedure are fully automated and controlled from the keyboard of the central computer. With such a ‘MULTIANALYSIS’ system GC-MS, LC-DAD and LC-MS data of the same sample can be obtained within 3 h. The system was optimised with nine chlorinated pesticides in drinking water as test mixture. With 100-ml samples detection limits in GC-MS were 0.0005−0.03 μg/l, and in LC-PB-MS 0.5–7 μg/l, both in the full-scan (EI) mode. Negative chemical ionisation (NCI) with methane as reagent gas improved the sensitivity of six halogenated compounds 3- to 30-fold and provided relevant information for structural elucidation of unknown compounds in real-world samples. LC-DAD detection limits varied from 0.01 to 0.05 μg/l. Relative standard deviations (R.S.D.) of retention times were less than 0.2% in all systems, R.S.D.s of peak areas were 5–15% for GC-MS and LC-PB-MS and less than 5% for LC-DAD. The ‘MULTIANALYSIS’ system was used to analyse surface water samples and river sediment extracts; several pollutants were detected and identified.     

Analysis of 500-ng/l levels of bromate in drinking water by direct-injection suppressed ion chromatography coupled with a single, pneumatically delivered post-column reagent

In July 1997, the US Environmental Protection Agency (EPA) began sampling and analyzing drinking water matrices from US municipalities serving populations greater than 100 000 for low-level bromate (>0.20 μg/l) in support of the Information Collection Rule (ICR) using the selective anion concentration (SAC) method. In September 1997, EPA published Method 300.1 which lowered the Method 300.0 bromate method detection limit (MDL) from 20.0 to 1.4 μg/l. This paper describes the research conducted at the EPA’s Technical Support Center laboratory investigating a single post-column reagent, o-dianisidine (ODA), which has been successfully coupled to EPA Method 300.1 to extend the MDL for bromate. Initial studies indicate that this method offers a MDL which approaches the EPA’s SAC method with the added benefit of increased specificity, shortened analysis time and reduced sample preparation. The method provides excellent ruggedness and acceptable precision and accuracy with a bromate MDL in reagent water of 0.1 μg/l, and a method reporting limit of 0.50 μg/l.     

Use of ion chromatography with post-column reaction for the measurement of tribromide to evaluate bromate levels in drinking water

A user-friendly ion chromatography method in conjunction with a post-column reaction (PCR) achieves practical quantitation limits for the oxyhalides bromate and chlorite of 0.05 μg/l and 0.10 μg/l, respectively. This level of measurement allows for the accurate assessment of bromate contributed to finished drinking waters that have been chlorinated using sodium hypochlorite. The target sensitivity of oxyhalides in the presence of other major ion species typically found in drinking water is achieved by PCR using excess bromide under acidic conditions to form a tribromide species that is detected by ultraviolet spectrometry. The method setup involves non-hazardous materials, as opposed to other recently developed methods that employ somewhat hazardous chemicals for generating the reaction necessary for the detection of bromate at sub-μg/l levels. No pretreatment of the samples is required, other than filtration and quenching of oxidant residual.     

Uranium determination at ppb levels by X-ray fluorescence after its preconcentration on polyurethane foam

A sensitive method based on the preconcentration of uranium on powdered polyurethane foam (PUF) has been developed to determinate this element in water samples by X-ray florescence. Uranium at ppb levels was sorbed as the salicylate complex on powdered PUF at pH 4.0. The resulting PUF was filtered through a filter paper and used for X-ray fluorescence measurements. For 50 μg/l of uranium the coefficient of variation for five measurements is 5% and the detection limit is 5.5 μg/l. The interference level of various ions and ligands was studied and optimum conditions were developed to determine uranium in reference materials, waste water, mine drainage, and sea water.     

Determination of phenols in water by on-line solid-phase disk extraction and liquid chromatography with electrochemical detection

Poly(styrene-divinylbenzene) (PS-DVB) membrane extraction disks were used as sorbents for the on-line solid phase extraction of 13 phenols (nitro and chlorophenols) in river and tap waters. Determination was performed by liquid chromatography with electrochemical detection (LC-ED). An acetate buffer-acetonitrile-methanol mixture as mobile phase and amperometric detection at +1100 mV were used. High water volumes, up to 250 ml, can be preconcentrated without loss of phenols (recoveries between 80% and 100%) except for the more polar ones. Moreover, detection limits between 0.01 and 0.1 μg l⁻¹ in tap water and between 0.1 and 1.0 μg⁻¹ in river water were obtained. The method has been applied to the analysis of two river water samples.     

Determination of glyphosate by ion chromatography

An ion chromatography system for the determination of glyphosate was described. Ion chromatograph was carried out by suppressed conductivity detection (DX-100). The eluent contained 9 mmol l⁻¹ Na₂CO₃ and 4 mmol l⁻¹ NaOH. The detection limit was 0.042 μg ml⁻¹ (S/N=3). The relative standard deviation was 1.99% and the correlation coefficient of the calibration curve for area was 0.9995. The linear range was 0.042100 μg ml⁻¹. Common inorganic ion and organic acids did not interfere. The recovery was 96.4103.2%. The method was simple, rapid, reliable and inexpensive.     

Trace analysis of chlorobenzenes in water samples using headspace solvent microextraction and gas chromatography/electron capture detection

In the present work, a rapid method for the extraction and determination of chlorobenzenes (CBs) such as monochlorobenzene, 1,2-dichlorobenzene, 1,3-dichlorobenzene, 1,4-dichlorobenzene, 1,2,3-trichlorobenzene and 1,2,4-trichlorobenzene in water samples using the headspace solvent microextraction (HSME) and gas chromatography/electron capture detector (ECD) has been described. A microdrop of the dodecane containing monobromobenzene (internal standard) was used as extracting solvent in this investigation. The analytes were extracted by suspending a 2.5 μl extraction drop directly from the tip of a microsyringe fixed above an extraction vial with a septum in a way that the needle passed through the septum and the needle tip appeared above the surface of the solution. After the extraction was finished, the drop was retracted back into the needle and injected directly into a GC column. Optimization of experimental conditions such as nature of the extracting solvent, microdrop and sample temperatures, stirring rate, microdrop and sample volumes, the ionic strength and extraction time were investigated. The optimized conditions were as follows: dodecane as the extracting solvent, the extraction temperature, 45 °C; the sodium chloride concentration, 2 M; the extraction time, 5.0 min; the stirring rate, 500 rpm; the drop volume, 2.5 μl; the sample volume, 7 ml; the microsyringe needle temperature, 0.0 °C. The limit of detection (LOD) ranged from 0.1 μg/l (for 1,3-dichlorobenzene) to 3.0 μg/l (for 1,4-dichlorobenzene) and linear range of 0.5–3.0 μg/l for 1,2-dichlorobenzene, 1,3-dichlorobenzene and from 5.0 to 20.0 μg/l for monochlorobenzene and from 5.0 to 30 μg/l for 1,4-dichlorobenzene. The relative standard deviations (R.S.D.) for most of CBs at the 5 μg/l level were below 10%. The optimized procedure was successfully applied to the extraction and determination of CBs in different water samples.     

Gas chromatographic determination of nitrite in water by precolumn formation of 2-phenylphenol with flame ionization detection

Nitrite can be determined by its reaction with 2-aminobiphenyl in acidic medium to produce 2-phenylphenol which is quantified by gas chromatography with flame ionisation detection using biphenyl as an internal standard. The hydrolysis of the intermediate diazonium ion avoids many of the problems encountered in the conventional determination of nitrite by the diazotization of an aromatic amine (usually sulphanilamide) and coupling with N-(1-napthyl)ethylenediamine dihydrochloride to yield an azo dye followed by spectrophotometry. Unlike this method, the proposed reaction is rapid and does not suffer from interferences by copper(II), iron(III) and lead(II). The calibration graph was linear over the range 5–1000 μg/l NO₂-N and the limit of detection found to be 0.5 μg/l NO₂-N. A single analysis can be completed within 20 min. The method was not affected by coloured or turbid analyte solutions and has been used to determine nitrite in natural waters.