Michael Sperling

On-line microwave sample pretreatment for the determination of mercury in water and urine by flow-injection cold-vapour atomic absorption spectrometry

Abstract A system for on-line treatment of liquid samples in a microwave oven and determination of mercury by cold-vapour atomic absorption spectrometry was designed and evaluated. The system consisted of an atomic absorption spectrometer, equipped with a mercury-hydride system and amalgamation accessory, a flow-injection system, an autosampler and a mircowave digestor. Urine and environmental water samples were stabilized with potassium dichromate-nitric acid and were mixed with a bromination reagent. The recoveries of eight mercury compounds from aqueous solutions and five compounds from dilute urine were studied. At an applied microwave power of 75 W, the recoveries of mercury(II) nitrate, methylmercury chloride, amidomercury chloride, phenylmercury chloride and diphenylmercury were between 92 and 102% for 1 + 2-diluted urine without amalgamation and between 94 and 111% for 1 + 5-diluted urine with amalgamation, respectively. The sample throughput was 30–40 h −1 without amalgamation and 24 h −1 in the amalgamation mode. Good agreement with certificate values was obtained for urine samples. A limit of detection (3σ) of 10 ng l −1 was obtained using 10-ml sample volumes of environmental waters (river, lake, rain) and the amalgamation technique. The results compared well with those from an external laboratory with correlation coefficients of 0.9302 and 0.9028 ( n = 22) for integrated absorbance and peak-height absorbance, respectively.

Spatially and temporally resolved gas phase temperature measurements in a Massmann-type graphite tube furnace using coherent anti-Stokes Raman scattering

Abstract The temperature of the nitrogen gas phase in a graphite tube furnace for atomic absorption spectrometry has been determined using coherent anti-Stokes Raman scattering (CARS). Subtle details of the temperature evolution at various locations in the tube have been identified. Under steady-state conditions, the temperatures of the tube wall and of the gas phase near the tube centre are essentially identical. The longitudinal gradient of the gas phase temperature between the tube centre (heated to 2700 K)andthetube ends is around 1200 K. This is less than that predicted by model calculations. During rapid heating, typically used for atomization of the analyte, the gas follows the wall temperature very closely and with essentially the same heating rate. Irregularities in this heating pattern, such as an intermediate slowing of the heating rate 0.3 s after start of heating, are most probably caused by gas expansion during the period of rapid tube heating. A pronounced radial temperature gradient was observed in the gas phase of tubes with inserted platform during the rapid heating phase, but not in tubes without a platform. The gradient in the gas phase disappears within about 0.5 s after the tube wall has reached the preset temperature. When the platform technique is used and the temperature program selected with care, volatilization of the analyte can be delayed until the tube wall and the gas phase have almost reached their final temperatures and are close to thermal equilibrium.

Differential determination of arsenic(III) and total arsenic using flow injection on-line separation and preconcentration for graphite furnace atomic absorption spectrometry

Abstract Arsenic(III) can be quantitatively extracted using sodium diethyldithiocarbamate (NaDDTC) as the complexing agent and C18 reversed phase packing as the column material for solid phase extraction. Arsenic(V) must be reduced to its trivalent oxidation state prior to extraction. A mixture of sodium sulphite, hydrochloric acid, sodium thiosulphate and potassium iodide was found to be optimum for on-line reduction. When the sorbent extraction is carried out without and with the addition of the reduction mixture, arsenic(III) and total arsenic can be determined sequentially by graphite furnace atomic absorption spectrometry with detection limits (3 σ) of 0.32 ng for As(III) and 0.43 ng for total arsenic. A 7.6-fold enhancement in peak area compared to direct injection of 40 μl samples was obtained after 60 s preconcentration. Results obtained for sea water standard reference materials, using aqueous standards for calibration, agree well with certified values. A precision of 5.5% RSD was obtained for total arsenic in a sea water sample (1.65 solμg l As). Results obtained for synthetic mixtures of trivalent and pentavalent arsenic agreed well with expected values.

Flame Atomic Absorption Spectrometric Determination of Cadmium, Cobalt, and Nickel in Biological Samples Using a Flow Injection System with On-Line Preconcentration by Co-Precipitation without Filtration:

Cadmium, cobalt, and nickel at ng/g to μg/g levels in plant and animal tissue reference materials and at μg/L levels in blood and urine were determined by flame atomic absorption spectrometry. The analyte elements were preconcentrated and separated from the bulk of the matrix by on-line co-precipitation with the hexamethylene ammonium hexamethylene dithiocarbamate iron(II) chelate complex in a flow injection system. The precipitate was collected in a knotted reactor made from 150-cm-long, 0.5-mm-i.d. Microline tubing without using a filter. The precipitate was dissolved in methyl isobutyl ketone and introduced directly into the nebulizer-burner system of an atomic absorption spectrometer. Ascorbic acid in an HC1/KC1 buffer was added on-line in order to reduce iron(III) to iron(II) because of its much better efficiency as a collector for trace elements. Reagent concentrations were optimized so that at least 200 mg/L of iron and 15 mg/L of copper could be tolerated in the sample solution without causing significant interferences. The portion of the analyte retained in the collector was about 70% for cadmium and 50% for cobalt and nickel. Enrichment factors of 24, 19, and 20 were obtained for cadmium, cobalt, and nickel, respectively, with the use of a 40-s co-precipitation time, resulting in enhancement factors, including the effect of the organic solvent, of 52, 43, and 52, respectively. The detection limits (3σ) for cadmium, cobalt and nickel were 0.15, 1.3, and 1.5 μg/L, respectively, and the precision was 1.5% RSD for 10 μg/L Cd, 2.7% RSD for 50 μg/L Co, and 1.8% RSD for 50 μg/L Ni. The analytical results obtained for a number of standard reference materials and control samples were in good agreement with the certified or recommended values.

Investigation of on-line coupling electrothermal atomic absorption spectrometry with flow injection sorption preconcentration using a knotted reactor for totally automatic determination of lead in water samples☆

Abstract A flow injection on-line sorption preconcentration electrothermal atomic absorption spectrometric system for fully automatic determination of lead in water was investigated. The discrete non-flow-through nature of ETAAS, the limited capacity of the graphite tube and the relatively large volume of the knotted reactor (KR) are obstacles to overcome for the on-line coupling of the KR sorption preconcentration system with ETAAS. A new FI manifold has been developed with the aim of reducing the eluate volume and minimizing dispersion. The lead diethyldithiocarbamate complex was adsorbed on the inner walls of a knotted reactor made of PTFE tubing (100 cm long, 0.5 mm i.d.). After that, an air flow was introduced to remove the residual solution from the KR and the eluate delivery tube, then the adsorbed analyte chelate was quantitatively eluted into a delivery tube with 50 μl of ethanol. An air flow was used to propel the eluent from the eluent loop through the reactor and to introduce all the ethanolic eluate onto the platform of the transversely heated graphite tube atomizer, which was preheated to 80°C. With the use of the new FI manifold, the consumption of eluent was greatly reduced and dispersion was minimized. The adsorption efficiency was 58%, and the enhancement factor was 142 in the concentration range 0.01–0.05 μg l−1 Pb at a sample loading rate of 6.8 ml min−1 with 60 s preconcentration time. For the range 0.1–2.0 μg l−1 of Pb a loading rate of 3.0 ml min−1 and 30 s preconcentration time were chosen, resulting in an adsorption efficiency of 42% and an enhancement factor of 21, respectively. A detection limit (3σ) of 2.2 ng l−1 of lead was obtained using a sample loading rate of 6.8 ml min−1 and 60 s preconcentration. The relative standard deviation of the entire procedure was 4.9% at the 0.01 μg l−1 Pb level with a loading rate of 6.8 ml min−1 and 60 s preconcentration, and 2.9% at the 0.5 μg l−1 Pb level with a 3.0 ml min−1 loading rate and 30 s preconcentration. Efficient washing of the matrix from the reactor was critical, requiring the use of the standard addition method for seawater samples. The analytical results obtained for seawater and river water standard reference materials were in good agreement with the certified values.

Flow-injection hydride generation atomic absorption spectrometric study of the automated on-line pre-reduction of arsenate, methylarsonate and dimethylarsinate and high-performance liquid chromatographic separation of their l-cysteine complexes

An automated on-line pre-reduction of arsenate, monomethylarsonate (MMA) and dimethylarsinate (DMA) using flow injection hydride generation atomic absorption spectrometry (FI-HGAAS) is feasible. The kinetics of pre-reduction and complexation depend strongly on the concentration of l-cysteine and on the temperature in the following increasing order: inorganic As(V)<DMA<MMA. Arsenate is pre-reduced/complexed within less than 50 s at 70-100 degrees C compared to 1 h at room temperature, while MMA and DMA require 1.5-2 min at 70-100 degrees C and up to 1-2 h at room temperature. The characteristic masses and concentrations for 100 mul injections are 0.01 ng and 0.1 mug l(-1) in integrated absorbance and 0.2 ng and 2 mug l(-1) in peak height measurements, and the limits of detection are ca. 0.5 ng and 5 mug l(-1), respectively. In a high-performance liquid chromatography (HPLC)-HGAAS system, the l-cysteine complexes of inorganic As(III), MMA and DMA are best separated within 7 min by HPLC on a strongly acidic cation exchange column such as Spherisorb S SCX 120x4 mm (5 mum) with a mobile phase containing 12 mmol l(-1) phosphate buffer (KH(2)PO(4)/H(3)PO(4))-2.5 mmol l(-1)l-cysteine, pH 3.3-3.5. Upon dilution to l-cysteine levels below 10 mmol l(-1), which are compatible with HPLC separations, the DMA-cysteine complex is unstable on storage. No baseline separations are possible with anion exchange and reverse phase C(18) HPLC columns. The limits of detection with 50 mul injections in peak area mode are ca. 0.5 ng and 10 mug l(-1), respectively.

Flame atomic absorption spectrometric determination of cadmium and copper in biological reference materials using on-line sorbent extraction preconcentration

SummaryCadmium and copper at the μg/g to ng/g level in plant and animal tissue reference materials, and at the μg/l level in urine were determined by flame atomic absorption spectrometry using on-line sorbent extraction preconcentration based on flow injection techniques. Bonded silica reversed phase sorbent with octadecyl functional groups (RP-C 18), packed in a 100 μl column, was used to collect the diethylammonium-N,N-diethyldithiocarbamate (DDTC) complex formed on-line in the sample digests at low pH. Methanol was used to elute the analyte chelates directly into the nebulizer-burner system of the spectrometer. Small air segments introduced before and after elution prevented the eluent from mixing with the sample solution and increased the sensitivity. A sampling frequency of 85/h could be obtained with a sample loading time of 30 s at a flow rate of 4.0 ml/min. The enrichment factor for both elements was 20 and the enhancement factors, including the effect of the organic solvent and with the flow spoiler removed, were 126 and 114 for cadmium and copper, respectively. The detection limits (3σ) were 0.15 μg/l for cadmium and 0.2 μg/l for copper. The precision was 2.3% and 1.4% r.s.d. for 10 μg/l Cd and 45 μg/l Cu, respectively (n=11). Results for the determination of cadmium and copper in various biological reference materials were typically in good agreement with certified values. Low recoveries were observed, however, for cadmium in samples containing high levels of copper and/or iron, such as bovine liver.

Time-based and volume-based sampling for flow-injection on-line sorbent extraction graphite furnace atomic absorption spectrometry

Abstract Fow-injection on-line preconcentration systems for graphite furnace atomic absorption spectrometry are complicated by the low eluate volume of typically less than 50 μl which can be accommodated in a graphite tube or on a graphite platform. Even when a column with an extra small capacity of 15 μl was used, it was found impossible to elute the sorbed analyte completely with an eluate volume that was compatible with the capacity of the graphite furnace. Two approaches for introducing only the most concentrated fraction of the eluate into the graphite tube while discarding the rest were investigated and compared: controlling the time interval for collection and introduction of the eluate fraction into the furnace tube (time-based sampling), and collection of the eluate fraction of interest in thin tubing of fixed volume, followed by introduction of this fraction into the tube using a low flow of air (volume-based sampling). Cadmium, copper, lead and nickel were the analyte elements investigated. A 15–30% greater enhancement factor was obtained for volume-based sampling because dispersion was interrupted during sample injection by air segmentation. The short- and long-term reproducibility were also better for volume-based sampling because variations in the pump tubing had no influence on the eluate volume introduced. These combined effects resulted in an improvement in detection limits of the four elements by factors of 1.3–2.0. Sample throughput (23 h−1), sample consumption (3 ml min−1 and reagent consumption were the same for both approaches. There were no significant differences in the accuracy and precision of the two techniques in the analysis of sea water, estuarine water and river water standard reference materials.

Flow injection on-line acid digestion and pre-reduction of arsenic for hydride generation atomic absorption spectrometry : a feasibility study

A flow injection (FI) manifold is described which makes possible on-line microwave-assisted acid digestion, followed by pre-reduction of As(V) to As(III) and its determination by hydride generation atomic absorption spectrometry. The merging zone technique is used in order to reduce acid consumption for digestion. The efficiency of acid digestion is increased by pressure which is built up in-line by a flow restrictor. Flows for sample pretreatment and hydride generation can be optimized independently. L-cysteine was found superior to potassium iodide as the pre-reductant because much lower reagent and acid concentrations are required, much harsher conditions can be tolerated for acid digestion, and the integrated absorbance signals for arsenic in blood and standards are essentially identical, making possible the use of the standard calibration procedure. The sampling frequency is 7-10/hr, depending on the conditions chosen, and the limit of detection, i.e. the concentration giving a signal equal to three times the standard deviation of the signal of the blank solution, is 0.25 mug/l for a 500 mul sample volume. The recovery of 10 mug/l As(V) added to a blood sample was 94 +/- 2 and 98 +/- 2% (n = 3) in absorbance and integrated absorbance, respectively.

Analysis of high-purity reagents using automatic on-line column preconcentration-separation and electrothermal atomic absorption spectrometry*

SummaryA fully automatic flow-injection on-line column preconcentration and separation procedure for electrothermal atomic absorption spectrometry is described for the determination of cadmium and lead in high-purity reagents. A microcolumn with 9 μL of solid sorbent, introduction of air prior to elution, and use of methanol as the eluent made it possible to elute the sorbed analyte quantitatively with only 80 μL of solvent. The total eluate volume was transferred into the graphite tube which was pre-heated to 80°C using a flow rate of 0.08 mL min−1 which allowed the solvent to evaporate in part during eluate introduction. The efficiency of the total procedure was 0.63 and 0.65, and the enrichment factor was 62 and 64 for cadmium and lead, respectively, compared with the direct introduction of 30 μL of an aqueous solution. The sample throughput was 13 h−1 for a sample loading time of 60 s and the detection limits (3 σ) were 0.7 and 4.5 ng L−1 for cadmium and lead, respectively. The relative standard deviation of the entire procedure in the optimum working range was typically around 3% (n=6). A number of high purity (suprapure) and analytical grade (pro analysi) reagents were analyzed, and spiking experiments resulted in recoveries of 97–104%.