Ion-imprinted polymethacrylic microbeads as new sorbent for preconcentration and speciation of mercury

Metal ion-imprinted polymer particles have been prepared by copolymerization of methacrylic acid as monomer, trimethylolpropane trimethacrylate as cross-linking agent and 2,2′-azobisisobutyronitrile as initiator, in the presence of Hg(II)-1-(2-thiazolylazo)-2-naphthol complex. The separation and preconcentration characteristics of the Hg-ion-imprinted microbeads for inorganic mercury have been investigated by batch procedure. The optimal pH value for the quantitative sorption is 7. The adsorbed inorganic mercury is easily eluted by 2 mL 4 M HNO₃. The adsorption capacity of the newly synthesized Hg ion-imprinted microbeads is 32.0 μmol g⁻¹ for dry copolymer. The selectivity of the copolymer toward inorganic mercury (Hg(II)) ion is confirmed through the comparison of the competitive adsorptions of Cd(II), Co(II), Cu(II), Ni(II), Pb(II), Zn(II)) and high values of the selectivity and distribution coefficients have been calculated. Experiments performed for selective determination of inorganic mercury in mineral and sea waters showed that the interfering matrix does not influence the extraction efficiency of Hg ion-imprinted microbeads. The detection limit for inorganic mercury is 0.006 μg L⁻¹ (3σ), determined by cold vapor atomic adsorption spectrometry. The relative standard deviation varied in the range 5-9 % at 0.02-1 μg L⁻¹ Hg levels. The new Hg-ion-imprinted microbeads have been tested and applied for the speciation of Hg in river and mineral waters: inorganic mercury has been determined selectively in nondigested sample, while total mercury e.g. sum of inorganic and methylmercury, has been determined in digested sample.     

Non-chromatographic speciation analysis of mercury by flow injection on-line preconcentration in combination with chemical vapor generation atomic fluorescence spectrometry

A novel non-chromatographic approach for direct speciation of mercury, based on the selective retention inorganic mercury and methylmercury on the inner wall of a knotted reactor by using ammonium diethyl dithiophosphate and dithizone as complexing agents respectively, was developed for flow injection on-line sorption preconcentration coupled with chemical vapor generation non-dispersive atomic fluorescence spectrometry. With the sample pH kept at 2.0, the preconcentration of inorganic mercury on the inner walls of the knotted reactor was carried out based on the exclusive retention of Hg–DDP complex in the presence of methylmercury via on-line merging the sample solution with ammonium diethyl dithiophosphate solution, and selective preconcentration methylmercury was achieved with dithizone instead of ammonium diethyl dithiophosphate. A 15% (v/v) HCl was introduced to elute the retained mercury species and merge with KBH₄ solution for atomic fluorescence spectrometry detection. Under the optimal experimental conditions, the sample throughputs of inorganic mercury and methylmercury were 30 and 20 h^− 1 with the enhancement factors of 13 and 24. The detection limits were found to be 3.6 ng l^− 1 for Hg²⁺ and 2.0 ng l^− 1 for CH₃Hg⁺. The precisions (RSD) for the 11 replicate measurements of each 0.2 μg l^− 1 of Hg²⁺ and CH₃Hg⁺ were 2.2% and 2.8%, respectively. The developed method was validated by the analysis of certified reference materials (simulated natural water, rice flour and pork) and by recovery measurements on spiked samples, and was applied to the determination of inorganic mercury and methylmercury in biological and environmental water samples.     

Microcolumn sorption of antimony(III) chelate for antimony speciation studies

Selective sorption of the Sb(III) chelate with ammonium pyrrolidine dithiocarbamate (APDC) on a microcolumn packed with C₁₆-bonded silica gel phase was used for the determination of Sb(III) and of total inorganic antimony after reducing Sb(V) to Sb(III) by l-cysteine. A flow injection system composed of a microcolumn connected to the tip of the autosampler was used for preconcentration. The sorbed antimony was directly eluted with ethanol into the graphite furnace and determined by AAS. The detection limit for antimony was significantly lowered to 0.007 μg l⁻¹ in comparison to 1.7 μg l⁻¹ for direct injection GFAAS. This procedure was applied for speciation determinations of inorganic antimony in tap water, snow and urine samples. For the investigation of long-term stability of antimony species a flow injection hydride generation atomic absorption spectrometry with quartz tube atomization (FI HG QT AAS) and GFAAS were used for selective determination of Sb(III) in the presence of Sb(V) and total content of antimony, respectively. Investigations on the stability of antimony in several natural samples spiked with Sb(III) and Sb(V) indicated instability of Sb(III) in tap water and satisfactory stability of inorganic Sb species in the presence of urine matrix.     

Mercury speciation by coupling cold vapour atomic absorption spectrometry with flow injection on-line preconcentration and liquid chromatographic separation

A fully automated system for the direct determination of methylmercury (MeHg), ethylmercury (EtHg), phenylmercury (PhHg), and inorganic mercury (Hg(II)) at the ng/L level is described. It is based on solid phase extraction preconcentration incorporated in a flow injection (FI) system, high performance liquid chromatography (HPLC) separation, reduction combined with thermolysis and determination by cold vapour atomic absorption spectrometry (CVAAS). For preconcentration a microcolumn of bonded silica with octadecyl functional groups (C18 reversed phase material) was used as a sorbent for the mercury complexes formed on-line with ammonium pyrrolidine dithiocarbamate. Retained mercury species are eluted with a methanol-acetonitrile-water mixture and subjected to separation on an octadecylsilane (ODS) column before determination by CVAAS. The sensitivity of organo-mercury determination could be improved by using NaBH₄ as a reductant combined with a thermolysis step. In order to perform on-line measurements the preconcentration microcolumn was mounted in a pressure-tight casing. Limits of detection for MeHg, EtHg, PhHg and Hg(II) employing a sample volume of 58.5 mL were 9, 6, 10 and 5 ng/L, respectively. The relative standard deviation (RSD) calculated from 9 repeated measurements was found to be 3.6%, 5.5%, 10.4% and 7.6% for MeHg, EtHg, PhHg and Hg(II), respectively. Finally, the application of this method for speciation of mercury in fish and human urine is described. Received: 10 March 1997 / Revised: 29 January 1998 / Accepted: 5 February 1998     

Sequential cloud point extraction for the speciation of mercury in seafood by inductively coupled plasma optical emission spectrometry

A novel nonchromatographic speciation technique for the speciation of mercury by sequential cloud point extraction (CPE) combined with inductively coupled plasma optical emission spectrometry (ICP-OES) was developed. The method based on Hg²⁺ was complexed with I⁻ to form HgI₄²⁻, and the HgI₄²⁻ reacted with the methyl green (MG) cation to form hydrophobic ion-associated complex, and the ion-associated complex was then extracted into the surfactant-rich phase of the non-ionic surfactant octylphenoxypolyethoxyethanol (Triton X-114), which are subsequently separated from methylmercury (MeHg⁺) in the initial solution by centrifugation. The surfactant-rich phase containing Hg(II) was diluted with 0.5 mol L^− 1 HNO₃ for ICP-OES determination. The supernatant is also subjected to the similar CPE procedure for the preconcentration of MeHg⁺ by the addition of a chelating agent, ammonium pyrrolidine dithiocarbamate (APDC), in order to form water-insolvable complex with MeHg⁺. The MeHg⁺ in the micelles was directly analyzed after disposal as describe above. Under the optimized conditions, the extraction efficiency was 93.5% for Hg(II) and 51.5% for MeHg⁺ with the enrichment factor of 18.7 for Hg(II) and 10.3 for MeHg⁺, respectively. The limits of detection (LODs) were 56.3 ng L^− 1 for Hg(II) and 94.6 ng L^− 1 for MeHg⁺ (as Hg) with the relative standard deviations (RSDs) of 3.6% for Hg(II) and 4.5% for MeHg⁺ (C = 10 μg L⁻¹, n = 7), respectively. The developed technique was applied to the speciation of mercury in real seafood samples and the recoveries for spiked samples were found to be in the range of 93.2–108.7%. For validation, a certified reference material of DORM-2 (dogfish muscle) was analyzed and the determined values are in good agreement with the certified values.     

Preconcentration speciation method for mercury compounds in water samples using solid phase extraction followed by reversed phase high performance liquid chromatography

A simple and rapid method for in situ preconcentration of inorganic and organic mercury compounds in water samples, based on solid phase extraction using dithizone immobilised on a reversed-phase C18 cartridge, has been developed. The adsorbed complexes were stable on the cartridge for at least 2 weeks. The speciation analysis of methylmercury (MeHg), phenylmercury (PhHg) and inorganic mercury (Hg (II)) were done by reversed-phase high performance liquid chromatography. The calibration graphs of MeHg, PhHg and Hg (II) were linear (r>0.999) from the detection limits (0.58, 0.66 and 0.54 ng) to 38, 25 and 26 ng of Hg, respectively. The average recoveries of MeHg, PhHg and Hg (II) from spiked samples (0.3-48.0 mug l(-1) Hg) were 98+/-3, 99+/-1 and 100+/-7%, respectively. By applying SPE procedure a 200-fold concentration of the sample was obtained.     

Mercury determination and speciation analysis in surface waters

A sorbent L-cysteine grafted silica gel has been evaluated for separation and enrichment of dissolved inorganic i-Hg(II) and methylmercury CH₃Hg(I) from surface waters at sub-μg L⁻¹ concentrations. Chemical parameters for mercury species enrichment and separation have been optimized. Analytical schemes for the determination of Hg species, using selective column solid phase extraction (SPE) with continuous flow chemical vapor generation atomic absorption spectrometry (CF-CVG-AAS) or inductively coupled plasma-mass spectrometry (ICP-MS) were developed. Possibilities for on-site SPE enrichment were demonstrated as well. The limits of quantification were 1.5 and 5 ng L⁻¹ for dissolved i-Hg(II) and CH₃Hg(I) by CF-CVG-AAS and 1 and 2.5 ng L⁻¹ by ICP-MS with relative standard deviations between 7–12% and 7–14%, respectively. The chemically modified SPE sorbent has demonstrated high regeneration ability, chemical and mechanical stability, acceptable capacity and good enrichment factors. Results for total dissolved mercury were in reasonable agreement with those from independent analyses by direct ICP-MS determinations for river waters and for estuarine water certified reference material.      

Voltammetric quantification and speciation of mercury compounds

The use of a carbon paste electrode modified with a thiolic resin for the determination of inorganic mercury and organomercury compounds, present simultaneously in a sample, is described. The compounds are first preconcentrated at the electrode surface by means of a purely chemical reaction with the modifier on the electrode surface. The high affinity of the modifier for the mercury compounds ensures low limits of detection and determination. Differentiation between several mercury species is possible by control of the reduction potential applied to the working electrode. This selective reduction results in the formation of atomic mercury at the electrode surface which can be determined with a very high sensitivity by means of its re-oxidation wave in cyclic voltammetry. Optimization of the instrumental parameters and evidence for the reduction processes are discussed. Analysis of inorganic mercury in the presence of methylmercury, with a detection limit of 4 μg Hg 1⁻¹, and of methylmercury in the presence of inorganic mercury, with a detection limit of 2 μg Hg 1⁻¹, is described in detail. In both cases the preconcentration time is 6 min. Other organomercury species can also be quantified. Application of the method to environmental aquatic samples is discussed.     

Mercury speciation by coupling cold vapour atomic absorption spectrometry with flow injection on-line preconcentration and liquid chromatographic separation

A fully automated system for the direct determination of methylmercury (MeHg), ethylmercury (EtHg), phenylmercury (PhHg), and inorganic mercury (Hg(II)) at the ng/L level is described. It is based on solid phase extraction preconcentration incorporated in a flow injection (FI) system, high performance liquid chromatography (HPLC) separation, reduction combined with thermolysis and determination by cold vapour atomic absorption spectrometry (CVAAS). For preconcentration a microcolumn of bonded silica with octadecyl functional groups (C18 reversed phase material) was used as a sorbent for the mercury complexes formed on-line with ammonium pyrrolidine dithiocarbamate. Retained mercury species are eluted with a methanol-acetonitrile-water mixture and subjected to separation on an octadecylsilane (ODS) column before determination by CVAAS. The sensitivity of organo-mercury determination could be improved by using NaBH₄ as a reductant combined with a thermolysis step. In order to perform on-line measurements the preconcentration microcolumn was mounted in a pressure-tight casing. Limits of detection for MeHg, EtHg, PhHg and Hg(II) employing a sample volume of 58.5 mL were 9, 6, 10 and 5 ng/L, respectively. The relative standard deviation (RSD) calculated from 9 repeated measurements was found to be 3.6%, 5.5%, 10.4% and 7.6% for MeHg, EtHg, PhHg and Hg(II), respectively. Finally, the application of this method for speciation of mercury in fish and human urine is described.     

A fully automated system for inorganic antimony preconcentration and speciation in urine

A study was undertaken to ascertain the analytical capabilities of l-methionine immobilized on controlled pore glass for Sb preconcentration and speciation. A fully automated on-line system, implemented with hydride generation (HG) and inductively coupled plasma optical emission spectrometry (ICP OES), was used. Sb(III), at pH 10 was selectively retained in the column containing the immobilized aminoacid, while Sb(V) was not retained at all. A 30% HCl solution was used as eluent agent. Prior to total Sb determination, a pre-reduction step with thiourea was necessary. An on-line pH adjusting and pre-reduction of Sb(V) was achieved in a fully automated system. The detection limit for the preconcentration of 10 mL of an aqueous solution was 70 ng L⁻¹ with a relative standard deviation of 2%. An enrichment factor of 20 was achieved when 10 mL of sample was passed through the system, reaching a throughput of 23 samples per hour. The method was successfully applied to the determination of Sb(III) and total Sb in urine.     

Selective Pre-concentration and Solid Phase Extraction of Mercury（Ⅱ） from Natural Water by Silica Gel-loaded （E）-N-（1-Thien-2＇-ylethylidene）-1,2-phenylenediamine Phase

Silica gel-loaded （E）-N-（1-thien-2＇-ylethylidene）-1,2-phenylenediamine （TEPDA） phase was synthesized based on physical adsorption approaches. The stability of a chemically modified TEPDA especially in concentrated hydrochloric acid that was then used as a recycling and preconcentration reagent allowed the further uses of silica gel-loaded immobilized TEPDA phase. The application of this silica gel-loaded phase to sorption of a series of metal ions was performed by using different controlling factors such as the pH of the metal ion solution and the equilibration shaking time by the static technique. This difference was interpreted on the basis of selectivity incorporated in these sulfur containing silica gel-loaded TEPDA phases. Hg（Ⅱ） was found to exhibit the highest affinity towards extraction by these silica gel-loaded TEPDA phases. The pronounced selectivity was also confirmed by the determined distribution coefficients （Kd） of all the metal ions, showing the highest value reported for mercury（Ⅱ） extraction by the silica gel immobilized TEPDA phase. The potential applications of the silica gel immobilized TEPDA phase to selective extraction of mercury（Ⅱ） from aqueous solution were successfully accomplished and preconcentration of low concentration of Hg（Ⅱ） （30 pg·mL^-1） from natural tap water with a preconcentration factor of 200 for Hg（Ⅱ） off-line analysis was conducted by cold vapor atomic absorption analysis.     

Chemical speciation of mercury in natural waters

Improved sensitivity of the cold-vapour atomic absorption method for mercury can be obtained by equilibrating the reduced sample with a small volume of air at 90°C. An automated system has been developed that has a detection limit of 1 ng Hg l^-1. By changing the reducing conditions three species of mercury can be differentiated and determined, inorganic mercury, arylmercury compounds such as phenylmercury(II) chloride, and alkylmercury compounds such as methylmercury(II) chloride. Speciation of mercury in natural waters is possible.     

Determination of Hg(II) in waters by on-line preconcentration using Cyanex 923 as a sorbent — Cold vapor atomic absorption spectrometry

Using a solid phase extraction mini-column home-made from a neutral extractant Cyanex 923, inorganic Hg could be on-line preconcentrated and simultaneously separated from methyl mercury. The preconcentrated Hg (II) was then eluted with 10% HNO₃ and subsequently reduced by NaBH₄ to form Hg vapor before determination by cold vapor atomic absorption spectrometry (CVAAS). Optimal conditions for and interferences on the Hg preconcentration and measurement were at 1% HCl, for a 25 mL sample uptake volume and a 10 mL min^− 1 sample loading rate. The detection limit was 0.2 ng L^− 1 and much lower than that of conventional method (around 15.8 ng L^− 1). The relative standard deviation (RSD) is 1.8% for measurements of 40 ng L^− 1 of Hg and the linear working curve is from 20 to 2000 ng L^− 1 (with a correlation coefficient of 0.9996). The method was applied in determination of inorganic Hg in city lake and deep well water (from Changchun, Jilin, China), and recovery test results for both samples were satisfactory.     

Cold vapor generation interface for mercury speciation coupling capillary electrophoresis with electrothermal quartz tube furnace atomic absorption spectrometry: Determination of mercury and methylmercury

A novel on-line coupled capillary electrophoresis (CE) cold vapor generation (CVG) with electrothermal quartz tube furnace atomic absorption spectrometry (EQTF-AAS) system for mercury speciation has been developed. The mercury species (inorganic mercury and methylmercury) were completely separated by CE in a 80 cm length × 100 μm i.d. fused-silica capillary at 20 kV and using a buffer of 100 mM boric acid and 10% (v/v) methanol (pH 8.30). The effects of the inner diameter of quartz tube, the acidity of HCl, the NaBH₄ concentration and N₂ flow rate on Hg signal intensity were investigated. Speciation of mercury was highlighted using CE-CVG-EQTF-AAS. The detection limits of methylmercury and mercury were 0.035 and 0.027 μg mL⁻¹, respectively. The precisions (RSDs) of peak height for six replicate injections of a mixture of 10 μg mL⁻¹ (as Hg) were better than 4%. The interface was used for speciation analysis of mercury in dry goldfish muscle.     

Some aspects of speciation and reactivity of mercury in various matrices

Speciation of mercury compounds in environmental and biological samples requires different techniques and different approaches. This speciation is mandatory to explain the toxicity, the reactivity and the bioavailability of mercury. It is dominated by inorganic mercury species Hg(II) and Hg(0), and the organic mercury species CH₃Hg and (CH₃)₂Hg. In this paper, some aspects of mercury speciation are presented in terms of:- mercury reactivity (Hg(II) complexation and reduction),- mercury species distribution in the main compartments of the environment     

On-line redox speciation analysis of antimony using l-proline immobilized on controlled pore glass and hydride generation inductively coupled plasma optical emission spectrometry for detection

l-proline was immobilized on controlled pore glass to study the ability of this material for the separation and preconcentration of Sb(III) and Sb(V). The substrate was packed in a minicolumn and incorporated in a flow injection system. The effluents of the on-line solid phase extraction (before and after elution) were directly coupled to the hydride generation inductively coupled plasma optical emission spectrometry system. The effect of pH, sample (and eluent) volume, flow rates of sample loading and elution on separation of Sb(III) e Sb(V) were evaluated. Our experiments demonstrated that Sb(V) was not retained and it was selectively determined during the loading step, while retained Sb(III) was determined after elution. The proposed system was also used for the selective preconcentration of Sb(III). In this case, a preconcentration factor of 11 and a limit of detection of 90 ng L⁻¹ for Sb(III) were achieved when 8 mL of sample were loaded into the column. The speciation analysis of inorganic Sb in river water and effluent samples was performed using the proposed method. The values obtained for total Sb (obtained by sum of Sb(III) and Sb(V)) were in good agreement with expected values. Recoveries of Sb(III) and Sb(V) in the river water Standard Reference Material 1640 (from National Institute of Standard and Technology) and spiked river waters were between 83 and 111%.     

Simultaneous speciation of inorganic arsenic and antimony in water samples by hydride generation-double channel atomic fluorescence spectrometry with on-line solid-phase extraction using single-walled carbon nanotubes micro-column

A new method was developed for the simultaneous speciation of inorganic arsenic and antimony in water by on-line solid-phase extraction coupled with hydride generation-double channel atomic fluorescence spectrometry (HG-DC-AFS). The speciation scheme involved the on-line formation and retention of the ammonium pyrrolidine dithiocarbamate complexes of As(III) and Sb(III) on a single-walled carbon nanotubes packed micro-column, followed by on-line elution and simultaneous detection of As(III) and Sb(III) by HG-DC-AFS; the total As and total Sb were determined by the same protocol after As(V) and Sb(V) were reduced by thiourea, with As(V) and Sb(V) concentrations obtained by subtraction. Various experimental parameters affecting the on-line solid-phase extraction and determination of the analytes species have been investigated in detail. With 180 s preconcentration time, the enrichment factors were found to be 25.4 for As(III) and 24.6 for Sb(III), with the limits of detection (LODs) of 3.8 ng L^− 1 for As(III) and 2.1 ng L^− 1 for Sb(III). The precisions (RSD) for five replicate measurements of 0.5 μg L⁻¹ of As(III) and 0.2 μg L⁻¹ of Sb(III) were 4.2 and 4.8%, respectively. The developed method was validated by the analysis of standard reference materials (NIST SRM 1640a), and was applied to the speciation of inorganic As and Sb in natural water samples.     

Dual-cloud point extraction as a preconcentration and clean-up technique for capillary electrophoresis speciation analysis of mercury

A novel dual-cloud point extraction (dCPE) technique is proposed in this paper for the sample pretreatment of capillary electrophoresis (CE) speciation analysis of mercury. In dCPE, cloud point was carried out twice in a sample pretreatment. First, four mercury species, methylmercury (MeHg), ethylmercury (EtHg), phenylmercury (PhHg), and inorganic mercury (Hg(II)) formed hydrophobic complexes with 1-(2-pyridylazo)-2-naphthol (PAN). After heating and centrifuging, the complexes were extracted into the formed Triton X-114 surfactant-rich phase. Instead of the direct injection or analysis, the surfactant-rich phase containing the four Hg species was treated with 150 microL 0.1% (m/v) l-cysteine aqueous solution. The four Hg species were then transferred back into aqueous phase by forming hydrophilic Hg-l-cysteine complexes. After dCPE, the aqueous phase containing the Hg-l-cysteine complexes was subjected into electrophoretic capillary for mercury speciation analysis. Because the concentration of Triton X-114 in the extract after dCPE was only around critical micelle concentration, the adsorption of surfactant on the capillary wall and its possible influence on the sample injection and separation in traditional CPE were eliminated. Plus, the hydrophobic interfering species were removed thoroughly by using dCPE resulted in significant improvement in analysis selectivity. Using 10 mL sample, 17, 15, 45, and 52 of preconcentration factors for EtHg, MeHg, PhHg, and Hg(II) were obtained. With CE separation and on-line UV detection, the detection limits were 45.2, 47.5, 4.1, and 10.0 microg L(-1) (as Hg) for EtHg, MeHg, PhHg, and Hg(II), respectively. As an analysis method, the present dCPE-CE with UV detection obtained similar detection limits as of some CE-inductively coupled plasma mass spectrometry (ICPMS) hyphenation technique, but with simple instrumental setup and obviously low costs. Its utilization for Hg speciation was validated by the analysis of the spiked natural water and tilapia muscle samples.     

Nanometer‐sized alumina packed microcolumn solid‐phase extraction combined with field‐amplified sample stacking‐capillary electrophoresis for the speciation analysis of inorganic selenium in environmental water samples

In this paper, a new method of nanometer‐sized alumina packed microcolumn SPE combined with field‐amplified sample stacking (FASS)–CE‐UV detection was developed for the speciation analysis of inorganic selenium in environmental water samples. Self‐synthesized nanometer‐sized alumina was packed in a microcolumn as the SPE adsorbent to retain Se(IV) and Se(VI) simultaneously at pH 6 and the retained inorganic selenium was eluted by concentrated ammonia. The eluent was used for FASS–CE–UV analysis after NH₃ evaporation. The factors affecting the preconcentration of both Se(IV) and Se(VI) by SPE and FASS were studied and the optimal CE separation conditions for Se(IV) and Se(VI) were obtained. Under the optimal conditions, the LODs of 57 ng L^?1 (Se(IV)) and 71 ng L^?1 (Se(VI)) were obtained, respectively. The developed method was validated by the analysis of a certified reference material of GBW(E)080395 environmental water and the determined value was in a good agreement with the certified value. It was also successfully applied to the speciation analysis of inorganic selenium in environmental water samples, including Yangtze River water, spring water, and tap water.     

On-line preconcentration of inorganic mercury and methylmercury in sea-water by sorbent-extraction and total mercury determination by cold vapour atomic absorption spectrometry

A comparative study of three mercury chelate forming reagents [diethyldithiocarbamate, pyrrolidin-1-yldithioformate and diphenylthiocarbazone (dithizone)] has been carried out for the preconcentration of ultratrace amounts of inorganic mercury and methylmercury in silica C(18) minicolumns as the solid sorbent. Sample flow injection in-line sorbent extraction was coupled with continuous cold vapour atomic absorption spectrometry (CVAAS) for detection. Results showed the superiority of the carbamate type reagents over the dithizone for the on-line formation and preconcentration of the corresponding mercury chelates. Using diethyldithiocarbamate (DDC) as reagent, aqueous sample volumes of 100 ml can be preconcentrated with 100% efficiency for both inorganic mercury and methylmercury. Quantitative release of the retained DDC chelates was obtained for volumes of eluent (ethanol) of 50 microl. Following the proposed procedure, detection limits of 16 ng/l. of mercury were achieved for sample volumes of 25 ml. The relative standard deviation was +/- 3.4% at 0.5 microg/l. Hg(II) levels. The method has been successfully applied to the determination of low levels of mercury in sea-water.