摘要:

Direct Determination of Nickel in Heroin and Cocaine by Electrothermal Atomic Absorption Spectrometry Using Deuterium Arc Background Correction Combined with Chemical Modification Journal of Analytical Atomic Spectrometry PILAR BERMEJO-BARRERA ANTONIO MOREDA-PIREIRO JORGE MOREDA-PIfiEIRO AND ADELA BERMEJO-BARRERA Department of Analytical Chemistry Nutrition and Brornatology Faculty of Chemistry University of Santiago de Compostela 15706 Santiago de Compostela Spain A comparative study of the use of nitric acid and magnesium nitrate as chemical modifiers using a deuterium arc background corrector in the direct determination of nickel in cocaine and heroin was developed. Cocaine samples were dissolved in 2 ml of 35.0% nitric acid whereas heroin was dissolved in water. Pyrolysis temperatures of 1500 and 1600 "C were obtained for nitric acid and magnesium nitrate respectively for cocaine and heroin solutions a deuterium lamp being adequate to correct the background signals obtained.The sensitivity was better with the use of magnesium nitrate the limit of detection being 32.1 pg kg-' compared with 34.3 pg kg-' with nitric acid. The within-batch precision (n = 11) and the analytical recovery were adequate with the use of both chemical modifiers. The introduction of a cooling step was studied through the within-run precision resulting in an optimum temperature of 200 "C. However the sensitivity obtained by using cooling was poorer than that given by programmes without a cooling step. Similarly the omission of the pyrolysis step did not give satisfactory results.Studies on interferences and on the variation of the analytical performance with the amount of sample were also carried out with the use of magnesium nitrate as the most suitable chemical modifier. Finally the method was applied to determine nickel in heroin and cocaine samples confiscated by the Spanish Police the concentrations found in heroin being between 0.50 and 1.35 mg kg-' and in cocaine between 0.07 and 0.34 mg kg-l. Keywords Nickel determination; magnesium nitrate chemical modifier; cocaine; heroin; electrothermal atomic absorption spectrometry The inorganic composition of illicit drugs such as heroin and cocaine are still far from well known because inorganic constituents such as metallic species are trace components. Therefore only a few concentration levels for a few metallic elements have been reported.' Studies on inorganic species in these samples are important for elucidating characteristic metals profiles for different drug samples which can be applied to obtain information about their geographical sources.In order to determine the inorganic contaminants in this kind of sample electrothermal atomic absorption spectrometry (ETAAS) appears to be the most adequate technique mainly because of the lower limits of detection that can be achieved for many metallic elements in comparison with those obtained by other techniques. However this technique has the disadvan- tage of the loss of analytes during the pyrolysis step which involves carrying out the pyrolysis at lower charring tempera- tures obtaining high background signals that cannot easily be corrected by using deuterium arc background correction when complex samples such biological and environmental materials are analysed.To overcome this problem some workers2-8 extract the nickel into organic solvents such as isobutyl methyl ketone or destroy the samples by digestion procedures mainly by using microwave wet acid d i g e s t i ~ n ~ ~ ~ . ' ~ prior to analysis. Other procedures such as that described by Shengjun and Holcombe," preconcentrate nickel from complex samples on algae a slurry of the algae then being subjected to AAS. In these procedures the deuterium arc background corrector can be efficiently used to correct the background signals generated. However all the procedures described above are tedious and time consuming and moreover accidental contamination or loss of nickel can occur.Therefore the direct analysis of complex samples by dilution of the samples with mineral acids or by using the slurry sampling technique for samples which cannot easily be dissolved appears more attractive. For this purpose to determine directly nickel in complex samples and overcome the problems related to the high background signal ETAAS methods based on the use of a Zeeman-effect background corre~tor'~-'~ have been extensively developed. However according to Patriarca and Fe11,I8 the direct determi- nation of nickel in complex samples by using a deuterium arc background corrector with instrumentation of more recent design has not been reported. The use of chemical modification is an indispensable approach in order to achieve higher pyrol- ysis charring temperatures thus obtaining better matrix volatil- ization the deuterium arc background corrector being adequate to carry out the measurements.The literature on chemical modifiers for nickel is not extensive but the use of ammonium dihydrogenphosphate," magnesium itra rate^'-^^ and palladium and nitric acid23 has been proposed in recent years. The use of zirconium-coated tubes has also been In this work the application of several chemical modifiers such as nitric acid palladium magnesium nitrate and pal- ladium-magnesium nitrate for the direct determination of nickel in complex samples such as cocaine and heroin using the deuterium arc background corrector system is described.EXPERIMENTAL Apparatus A Perkin-Elmer Model 1 lOOB atomic absorption spectrometer equipped with a deuterium arc background corrector a Perkin- Elmer HGA 400 graphite furnace and a Perkin-Elmer AS 40 autosampler was used for all measurements. A nickel hollow- cathode lamp operated at 25 mA which provides a wavelength of 232.0 nm was used the slit width being 0.2 nm. All experi- Journal of Analytical Atomic Spectrometry November 1995 VoZ. 10 101 Iments were run with 20 pl samples and measurements were made based on integrated absorbance. Pyrolytic graphite- coated graphite tubes and L'vov platforms were used throughout. Reagents All solutions were prepared from analytical-reagent grade chemicals using ultra-pure water of resistivity 18 M a cm which was obtained by means of a Milli-Q water purification system (Millipore Bedford MA USA).Magnesium nitrate stock standard solution 2.000 g 1-' . Prepared by dissolving 2 g of magnesium nitrate (BDH Poole Dorset UK) in 11 of ultra-pure water. Nickel nitrate stock standard solution 1.00Og 1-' Ni. Prepared from the analytical-reagent grade chemical supplied by Merck (Darmstadt Germany). Nitric acid stock standard solution 35%. Prepared from Suprapur acid (69.0-70.5% with a maximum nickel content of 0.005 mg 1-I; BDH) by appropriate dilution with ultra-pure water and used without standardization. Palladium stock standard solution 3.000 g 1 - ' . Prepared according to Welz et al.25 by dissolving 300mg of palladium (99.999% Aldrich Milwaukee WI USA) in 1 ml of concen- trated nitric acid and diluting to 100 ml with ultra-pure water.If the dissolution was incomplete 10 pl of hydrochloric acid (Suprapur 35.0% with a maximum nickel content of 0.005mg1-' BDH) were added to the cold nitric acid and heated to gentle boiling in order to volatilize the excess of chloride. Argon N50. Argon of purity 99.999% was supplied by SEO (Madrid Spain). Procedure Cocaine samples (0.5 g) were dissolved in 2 ml of 35.0% nitric acid and diluted to 10ml with ultra-pure water. For heroin 0.25 g of sample was dissolved in ultra-pure water adding 0.4ml of 35.0% nitric acid to dissolve some inert substances that could not be dissolved in water. The solutions were then diluted to 10ml with ultra-pure water. All samples were kept in polyethylene vials at 4 "C.For measurements 0.5 or 0.25 ml Table 1 Graphite furnace temperature programmes related to the use of nitric acid palladium palladium-magnesium nitrate and magnesium nitrate as chemical modifiers Temperature/ Ramp rate/ Hold time/ Ar flow/ Stage "C S S ml min-l Drying 150 15 15 300 Atomization 25002 0 5 0 (Read) Cleaning 2600 1 2 300 Pyrolysis 1600* 10 lot 300 * 1500 "C with the use of nitric acid and palladium. t 15 s with the use of nitric acid palladium and palladium- 22300°C with the use of nitric acid and 2400°C with the use of magnesium nitrate. palladium-magnesium nitrate. of the cocaine or heroin solutions respectively were transferred into autosampler cups adding a suitable volume of magnesium nitrate used as a chemical modifier to give a concentration of 5 mg I-' after dilution to 1 ml.Calibration was performed over the range 0-30 pg 1-' injecting 20 pl into the atomizer and running the sequential drying pyrolysis atomization cleaning programme of the graphite furnace shown in Table 1. RESULTS AND DISCUSSION Optimization of the Graphite Furnace Temperature Programme Experiments were performed in order to obtain the optimum temperatures ramp and hold times for the drying pyrolysis and atomization stages of the graphite furnace temperature programme related to the use of nitric acid palladium pal- ladium-magnesium nitrate and magnesium nitrate as chemical modifiers. First the optimum pyrolysis temperatures were obtained Table 2 for a cocaine sample solution spiked with 20 pg 1-' of Ni2+ and an aqueous standard solution of 30 pg I-' of Nizf and for the chemical modifiers cited at a concentration of 10 mg 1-' for palladium and magnesium nitrate and 3.5% for nitric acid which is the concentration obtained before cocaine sample preparation and dilution into autosampler cups.The atomization temperature used for all chemical modifiers was 2400 "C. The background signals obtained were gradually reduced when the pyrolysis temperatures were increased obtaining the largest background signal attenuation when magnesium nitrate was used as the chemical modifier. Thus for the pyrolysis temperatures shown in Table 2 related to cocaine solutions background signals lower than 0.08 A s were achieved with the use of all chemical modifiers tested. Optimum pyrolysis temperatures related to each chemical modifier were also obtained for a heroin sample solution (Table 2).The difference in pyrolysis temperatures obtained for cocaine and heroin sample solutions can be explained through the effect of the organic matter on the stabilizing power of the chemical modifiers investigated. As reported by Hinds and Jackson,26 with the use of palladium as chemical modifier in the determination of lead the presence of organic matter reduces the stabilizing effects of palladium higher pyrolysis temperatures being obtained for aqueous solutions. Therefore for heroin sample solutions which were prepared from a smaller amount of drug sample (0.25 g) than for cocaine sample solutions (0.50 g) the organic matter content is lower than for cocaine sample solutions so the stabilizing effect of palladium is greater obtaining a higher pyrolysis temperature of 1700°C.This can also be observed with nickel aqueous standards. As can be seen with the use of nitric acid or magnesium nitrate no effect of the organic matter on the stabilizing power of these two chemical modifiers is observed hence the same pyrolysis temperatures 1600 and 1500"C are achieved for magnesium nitrate and nitric acid respectively for aqueous solutions and for cocaine and heroin sample solutions. However there are no data on the atomization mechanisms of nickel in the presence of the chemical modifiers Table 2 Optimum pyrolysis and atomization temperatures related to nitric acid palladium palladium-magnesium nitrate and magnesium nitrate as chemical modifiers Charring temperaturePC Atomization temperaturePC Cocaine Heroin Ni2+ aqueous Chemical modifier solution solution standard solution HN03 1500 1500 1500 Pd 1500 1700 1700 Pd-Mg ( NO3 12 1600 1700 1700 Mg(NOd2 1600 1600 1600 Cocaine Ni2+ aqueous solution standard solution 2300 2500 2500 2500 2400 2500 2500 2500 101 2 Journal of Analytical Atomic Spectrometry November 1995 Vol.10investigated which could explain the difference between the pyrolysis temperatures achieved related to the different chemi- cal modifiers and the effect of the organic matter for each chemical modifier although the work of Akman et Pupysher and Nagdeev28 and MacAllister2' led to the eluci- dation of the atomization of nickel in the absence of matrix modification. It can be concluded that magnesium nitrate appears to be the most suitable chemical modifier owing to the higher pyrolysis temperature obtained 1600 "C for both cocaine and heroin sample solutions and also for nickel aqueous standards the same effect of magnesium nitrate on nickel being obtained for all organic matter contents.The optimum pyrolysis tem- peratures chosen for the use of the other chemical modifiers tested were related to the cocaine sample solution because at the pyrolysis temperature related to heroin sample and nickel aqueous solutions 1700 "C considerable loss of nickel occurs for cocaine sample solutions. As a larger sample volume is needed for cocaine samples than for heroin samples (involving an increase in the back- ground signal) the following studies were performed on cocaine rather than heroin sample solutions.The atomization temperature was also optimized for the use of each chemical modifier for cocaine sample solutions spiked with 20 pg 1-1 of Ni2+ and an aqueous standard solution of 30 pg 1-1 of Ni2+. The pyrolysis temperature used was 1500 "C for nitric acid and palladium and 1600°C for magnesium nitrate and palladium-magnesium nitrate. The results are given in Table 2. The ramp and hold times for the charring step were also optimized for each chemical modifier using a cocaine sample solution spiked with 2Opg1-1 of Ni2+. Optimum values corresponding to each chemical modifier are shown in Table 1. The choice of these values was established by considering to be more adequate those ramp and hold times that offer a higher nickel absorbance and lower background signals.Optimum drying conditions temperatures and times for each chemical modifier were determined by inserting the sample into the graphite tube. In order to provide smooth even evaporation of the solvent with no sputtering of the sample an optimum temperature of 150°C with ramp and hold times of 15 and 20s respectively was achieved. Finally a cleaning temperature of 2600°C with ramp time of 1 s and a hold time of 2 s were used to remove possible memory effects. Based on the above the graphite furnace temperature pro- gramme related to each chemical modifier is as shown in Table 1. Optimization of the Amount of Chemical Modifier The amount of chemical modifier nitric acid palladium and magnesium nitrate was optimized.The effects of increasing amount of nitric acid on the nickel absorbance and background signals were studied separately whereas the effects due to palladium and magnesium nitrate were studied for both species in conjunction. Therefore the amount of nitric acid was increased from 3.5% which is the concentration of nitric acid obtained after the cocaine sample preparation and dilution into the auto- sampler cups to 10%. It was found that the increase in the concentration of nitric acid produced a decrease in the nickel absorbance up to a nitric acid concentration of 6.O% increasing smoothly at higher concentrations although the nickel absorbance was not higher than that achieved with a concen- tration of 3.5%. No variation in the background signal was obtained for concentrations of nitric acid higher than 3.5%.The amount of nitric acid was also increased for the aqueous standard solution of Ni2+ (30 pg l-') obtaining similar results. Therefore the optimum amount of nitric acid selected was 0.1 c I & 0.075 a CI - 0.05 ' 30 Fig. 1 Dependence of the nickel absorbance signal on the combined effects of various amounts of Pd and Mg added to (a) a cocaine sample solution spiked with 20 pg 1 - l Ni2+ and (b) an aqueous standard solution of 30 pg 1-1 Ni2+ 3.5% as higher concentrations do not offer better results with respect to nickel absorbance signal and peak shape and in addition an increase in nitric acid concentration damages the graphite tubes faster. In the same way the amounts of palladium and magnesium nitrate were optimized using different combinations between palladium and magnesium nitrate both in the 0-30mg1-1 range.Figs. 1 (a) and (b) show the results obtained for a cocaine sample solution spiked with 20 pg 1-1 of Ni2+ and an aqueous standard solution of 30 pg 1-1 of Ni2+ respectively. As can be seen in Fig. 1 (a) large amounts of palladium depress the nickel absorbance signal. In addition it can be observed that the highest nickel absorbance signal is related to a concentration of magnesium of 5 mg l-l decreasing gradually with larger amounts of magnesium. This is also shown in Fig.2 where different peak shapes related to different amounts of magnesium 0.25 1 0 5 Timds Fig. 2 Effect of zero Pd concentration and different concentrations of Mg on the peaks scheme related to a cocaine sample solution spiked with 2Opg1-1 NiZ+.A Omgl-I (0.132As); B 5mg1-' (0.145 A s); C 10 mg 1-' (0.138 A s); and D 20 mg 1-' (0.131 A s) Journal of Analytical Atomic Spectrometry November 1995 Vol. 10 101 30.25 1 0 5 Timds Fig. 3 Effect of 5 mg 1-1 of Mg and different concentrations of Pd on the peaks scheme related to a cocaine sample solution spiked with 2Opg1-' Ni2+. A Omgl-I (0.145 As); B 5mg1-1 (0.132 As); C 10 mg 1-1 (0.131 A s); and D 20 mg 1-1 (0.128 A s) nitrate in the absence of palladium are represented. In addition Fig. 3 shows several peak shapes corresponding to experiments involving different concentrations of palladium for a fixed amount of magnesium nitrate of 5 mg 1-'. As can be seen the nickel absorbance signal is decreased when palladium is present.On the other hand in Fig. 1 (a) it is also seen that large amounts of palladium lead to lower nickel absorbance signals which could not be compensated for by the addition of large amounts of magnesium nitrate. When these experiments were performed on nickel aqueous standards the results obtained were different [Fig. l(b)]; the addition of small amounts of palladium does not depress the nickel absorbance the maximum nickel absorbance value being obtained for a concentration of 5 mg I-' for each. The differences obtained for the cocaine sample and aqueous solutions is attributed to the organic matter present in the samples which changes the stabilizing power of palladium.26 Therefore we conclude that magnesium nitrate alone at a concentration of 5 mg I-' is better than the use of both species in conjunction the addition of palladium not being advanta- geous.On other hand as noted above with the use of mag- nesium nitrate the optimum pyrolysis temperature reached 160O0C is the same for drug sample solutions and also for aqueous standards of nickel. Study of Omission of the Pyrolysis Step In order to obtain a faster graphite furnace temperature programme the omission of the pyrolysis step has been proposed by several workers in recent years for the determi- nation of different metals such as aluminium arsenic or lead30-34 and also for nickel using a Zeeman-effect background c~rrector.~' Therefore a study of the omission of the pyrolysis stage was carried out using a cocaine sample solution spiked 0.25 1 2 0.25 0 2 5 5 with 20 pg 1-' of Ni2+.First the drying temperature was increased from 150 "C which is the drying temperature found to be optimum for the former programme (Table l) to 400 "C the atomization temperature being 2500 "C and magnesium nitrate used as the chemical modifier. The results are shown in Fig. 4 where different peak shapes related to each drying temperature and also to the former programme (Table l) are presented. As can be seen with omission of the pyrolysis step when lower drying temperatures are used (peak B) a distortion of the nickel atomic absorption signal occurs owing to the water v a p o ~ r . ~ ' 9 ~ ~ This distortion could not be eliminated with higher drying temperatures (peaks C and D) owing to the insufficient matrix volatilization which generates background signals that cannot be corrected by using the deuterium arc background corrector.In addition in Fig. 4 it can be also seen that the nickel absorbance signals related to the different drying temperatures are lower than those obtained by using the former programme. These results are in accord with our experience on the omission of the pyrolysis step in the determi- nation of lead in organic samples36 and with other workers who recommended the omission of the pyrolysis step when the organic matter content is Study of the Introduction of a Cooling Stage The introduction of a cooling step has been proposed as a way to increase the sensitivity of analytical methods and it has been successfully applied to the determination of several However there are no data on the introduction of a cooling step in nickel determination. Therefore the introduc- tion of a cooling step just before the atomization stage was studied using ramp and hold times of 10s each and varying the cooling temperatures between 50 and 300°C.The within- run precision for seven replicate injections of a cocaine sample solution spiked with 20 pg 1-' of Ni2+ the nickel absorbance values for an aqueous standard solution of 30 pg I-' of Ni2+ and a cocaine sample solution spiked with 20 pg 1-l of Ni2+ related to each cooling temperature and to the former pro- gramme (without cooling) were investigated. No improve- ments in the within-run precision and nickel absorbance (integrated and peak height) were obtained.Therefore further studies on the introduction of a cooling stage were abandoned. Calibration and Standard Addition Graphs In order to compare the convenience of using nitric acid and magnesium nitrate as chemical modifiers for the determination 0.25 0 0.25 0 Timds 17 Fig.4 Peaks scheme related to the omission of the pyrolysis stage for various drying temperatures A nickel absorbance signal and B background signal. (a) Former programme (Table 1); (b) 150 "C; (c) 300 "C; and (d) 400 "C 101 4 Journal of Analytical Atomic Spectrometry November 1995 Vol. 10of nickel in illicit drugs the linear range with the use of each chemical modifier was studied the results being 0.85-40 and 0.80-40 pg l-l respectively. The standard addition method was used over the same range of concentrations obtaining the following equations calibration graph QA = 0.003 + 0.0053 [Nil; r = 0.999 standard addition graph QA = 0.010 + 0.0057 [Nil; r = 0.999 for the use of nitric acid and calibration graph QA = 0.003 + 0.0054 [Nil; r = 0.999 standard addition graph QA = 0.016 + 0.0057 [Nil; r = 0.999 for the use of magnesium nitrate where QA is integrated absorbance and [Ni2+] is in pg 1-l.As can be seen slopes of calibration and standard addition graphs are similar for both chemical modifiers no statistical difference being found between the two when the F-test was applied for a confidence level of 99.5Y0.~~ This means that matrix effects can be assumed not to be important and hence calibration graphs can be used to develop the measurements for the use of both chemical modifiers.The slopes of the calibration graphs and also the standard addition graphs related to the use of nitric acid and magnesium nitrate are similar no variation in the slopes being observed when magnesium nitrate is added. This may mean that although the use of magnesium nitrate as a chemical modifier stabilizes nickel at a higher temperature 16OO0C than nitric acid 1500 "C the analytical performances such as sensitivity and precision with the use of nitric acid are also similar to those found with the use of magnesium nitrate. In addition the use of nitric acid as chemical modifier is advantageous over the use of magnesium nitrate in the sense that the optimum concentration of nitric acid is obtained after cocaine sample dissolution and dilution into autosampler cups and hence it is only necessary to dilute the samples without addition of other species.Therefore in the following sections the use of magnesium nitrate and nitric acid as chemical modifiers will be compared. Figures of Merit Sensitivity was studied through the limit of detection (LOD) the limit of quantification (LOQ) and the characteristic mass (mo) which are defined as follows S LOD=3 - S C LOQ= ig s c VC x 0.0044 QA - QA m = where s is the standard deviation of eleven measurements of the blank Sc is the slope of the calibration graph C is the nickel concentration in pg 1-1 of Ni V is the volume of sample in pl and QA and QA are the integrated absorbances of the sample and blank respectively. Blanks were solutions of 3.5% nitric acid for the use of nitric acid as chemical modifier and 3.5% nitric acid and 5 mg 1-1 magnesium nitrate for the use of magnesium nitrate as chemical modifier.The mean values obtained for eleven replicate injections of the blanks were 0.006 & 0.001 and Table3 Figures of merit for the use of nitric acid and magnesium nitrate as chemical modifiers Chemical modifier LOD/pg kg-' LOQ/pg kg-' rn,/pg 16.0 f 0.7 107.1 16.2k0.7 HN03 34.3 114.4 Mg(N03)z 32.1 0.002 f 0.001 A s for nitric acid and magnesium nitrate respectively. The results referred to 0.5 g of drug sample are given in Table 3. Precision and Accuracy Precision was studied through the within-batch precision obtained for eleven replicate injections of four solutions of the same cocaine sample which was spiked with 0 10 20 and 30 pg 1-1 of Ni2+.Results expressed as relative standard deviation s (YO) were 7.41 4.74 2.62 and 1.83% for 0 10 20 and 30 pg 1-l of Ni2+ respectively for the use of nitric acid and 6.96 2.21 1.51 and 1.48% also for 0 10 20 and 30 pg 1-l of Ni2+ respectively for the use of magnesium nitrate. As can be seen good precision is achieved by using both chemical modifiers although with the use of magnesium nitrate the s values are lower than those with nitric acid. As no reference material for samples such cocaine and heroin is available the accuracy of methods was studied through the analytical recovery using nitric acid and magnesium nitrate as chemical modifiers. Solutions of a cocaine sample were spiked with 0 10 20 and 30 pg 1-l of Ni2+ and were injected eleven times into the atomizer obtaining the analytical recover- ies given in Table 4.As can be seen adequate analytical recoveries close to loo% were achieved with the use of both chemical modifiers and for all concentration levels studied. As can be observed similar precisions and limits of detection were obtained with the use of both magnesium nitrate and nitric acid chemical modifiers. However we selected mag- nesium nitrate as the chemical modifier in order to volatilize the sample matrix more efficiently. Study of Precision and Analytical Recovery with Variation of the Amount of Sample In order to decrease the limits of detection and quantification the amount of sample can be increased obtaining higher sample concentrations. However an increase in sample concen- tration can impair the precision and even the accuracy of the methods.Therefore a study of the effect of increasing the sample concentration on the precision and analytical recovery of the method was undertaken. Amounts of cocaine sample of 0.25,0.50 1.00 and 2.00 g giving after dilution to 10 ml sample concentrations of 2.5 5.0 10.0 and 20.0% m/v respectively were investigated. With 2.00 g of cocaine sample the volume of nitric acid used (see Experimental) 2 ml was insufficient to dissolve it and for this amount of sample 2.5 ml of 35.0% nitric acid were required to dissolve the sample completely. To study the effect of the sample concentration on precision solutions with amounts of sample less than 2.00 g were mixed Table 4 Analytical recoveries achieved with the use of nitric acid and magnesium nitrate as chemical modifiers Analytical recovery (%) Concentration added/pg 1-l HNO Mg(NO3)z 10 20 30 96.1 _+ 4.2 96.1 _+ 1.9 108.0 k2.2 104.6 f 1.2 96.5 k 1.3 98.8+_ 1.1 Journal of Analytical Atomic Spectrometry November 1995 Vol.10 101 5with nickel aqueous standard solution to give absorbances similar to those achieved for the injection of the sample solution prepared from 2.00 g and the within-run precision of eleven replicate injections of each solution was obtained. The results expressed as s were 2.18 2.35 1.45 and 1.82% for sample concentrations of 20.0 10.0 5.0 and 2.5% m/v respect- ively. These results are acceptable as for all sample concen- trations tested the s values are lower than 10%.Therefore an increase in sample concentration does not impair the precision. To study the effect of the amount of sample on the analytical recovery solutions related to each sample concentration were spiked with 10 and 20 pg 1-l of Ni2+ giving the analytical recoveries shown in Table 5 corresponding also to eleven replicate injections of each solution. As can be seen analytical recoveries close to 100% are achieved for all sample concentrations studied and for both concentration levels. Therefore the amount of sample can be increased to 2.00 g which gives a sample solution concentration of 20.0% m/v without loss of analytical recovery or precision. In addition there are no problems with the background signals. In this way the limits of detection and quantification can be decreased to 8.0 and 26.8 pg kg-' respectively.Finally from these results it can also be concluded that 0.5 g of drug sample to give a sample solution concentration of 5.0% m/v is a representative mass to develop the analysis as the precision and analytical recovery do not vary statistically in the range 0.25-2.00 g. Study of Interferences The effects of different species on the nickel absorbance signal were studied. We assumed that a species is an interferent when at a particular concentration the absorbance signal varies Table 5 Analytical recovery related to different sample concentrations by+ 10% from the signal recorded in its absence. Different volumes of aqueous solutions of the potential interferents were added to cocaine sample solutions spiked with 20 pg 1-l of Ni2 + at concentrations higher than those typically found in this kind of sample. The results obtained are given in Table 6 where the lowest concentration of interferent added in mg l-l the lowest concentration of interferent referred to 0.5 g of drug sample (according to dilutions made) and the percentage nickel absorbance variation are presented.Table 6 also shows the concentration of some species present in cocaine and heroin reported in the literature.' As can be seen for species whose levels are available no interferences are observed. For other species which are not reported in drugs such Ca2+ K+ Na' C1- P043- and SO4'- and which can be considered as major constituents a concentration of 100 mg 1-l (4000 mg kg-l referred to the drug sample) does not produce interferences. Hence we can assume that the addition of magnesium nitrate as chemical modifier allows the determination of nickel in such samples without interferences. Results relative to chloride are important as this species has been extensively reported as a serious interferent in the determination of n i ~ k e l . ~ ~ - ~ ~ Further as indicated in preceding sections there is no matrix effects with the use of magnesium nitrate as the slopes of the calibration and standard addition graphs are statistically the same.Applications The method was applied to the determination of nickel in heroin and cocaine samples from different production areas. One sub-sample was taken from each drug sample and was prepared as described under Experimental then subjected to AAS.The results obtained are given in Table 7 where it can be seen that the nickel levels are higher in heroin than in cocaine samples mainly owing to their lower purity. Analytical recovery (%) Amount of Sample concentration sample/g (YO m/v) + 10 pg 1-1 +20 pg 1-1 0.25 2.5 106.7 & 3.2 98.5 & 1.6 0.50 5.0 100.6 k 3.5 98.9 f 1.6 1 .00 10.0 96.9 k 3.2 103.2 f 2.4 2.00 20.0 100.2 & 3.8 99.6 & 1.5 CONCLUSIONS This work confirms the advantageous application of mag- nesium nitrate as a chemical modifier for nickel determination in comparison with other chemical modifiers such as palladium nitric acid and palladium-magnesium nitrate obtaining adequate nickel stabilization at the pyrolysis stage which Table 6 Effect of different species on nickel absorbance signal for a cocaine sample solution spiked with 20 pg 1-1 of Ni2' Interferent ~ 1 3 + Ba2+ Ca2 + Cd2+ co2 + Cr3 + cu2 + Fe3 + K+ Li + Mn2+ Na+ NH4+ Pb2+ Zn2 + Citrate c1- so,2- Si03'- Lowest concentration added/mg 1-' 100 100 2.5 2.5 2.5 2.5 10 50 100 12.5 25 100 25 80 lo0 100 100 100 25 2.5 Lowest concentration referred to sample/ mg kg-l 4000 100 4000 100 100 100 400 2000 4000 500 1000 4000 1000 100 3200 4000 4000 4000 4000 1000 Concentration interval/ mg kg-' 0.5-2700 0.1-26 0.1-2.1 0.6-1.6 0.7-270 2.7-2200 -* -* -* -* 0.5-220 -* -* 0.3-2.1 0.1-2600 -* * -* - * - -* Nickel absorbance variation (%) + 0.0 + 5.32 + 1.49 + 0.0 -5.15 -5.15 -4.08 - 8.74 - 3.98 + 0.52 + 7.81 + 1.56 - 3.09 + 2.66 + 9.47 0.0 - 4.64 - 2.04 -0.51 -4.61 ~~ ~ ~ * Data not available.101 6 Journal of Analytical Atomic Spectrometry November 1995 Vol. 10Table 7 Nickel concentrations found in different heroin and cocaine samples Sample Heroin 1 Heroin 2 Heroin 3 Heroin 4 Heroin 5 Cocaine 1 Cocaine 2 Cocaine 3 Cocaine 4 Nickel concentration,’ mg kg-I 1.35 0.50 0.53 1.48 0.53 0.18 0.18 0.34 0.20 Sample Cocaine 5 Cocaine 6 Cocaine 7 Cocaine 8 Cocaine 9 Cocaine 10 Cocaine 11 Cocaine 12 Nickel concentration/ mg kg-l 0.16 0.26 0.14 0.07 0.08 0.10 0.21 0.1 1 allows an efficient matrix volatilization. In this way the deuterium arc background corrector can be used successfully to correct the background signal generated. Further the omis- sion of the pyrolysis step is not possible when using the deuterium arc background corrector.Finally from the concen- tration levels found in the samples analysed it can be concluded that nickel is a trace contaminant in cocaine and heroin. REFERENCES 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Violante N. Quaglia G. Lopez A. and Caroli S. Microchem. J. 1992 42 79. Pedersen B. Willens M. and Jorgensen S . S. Analyst 1980 105 119. Petrov I. S. Tsalev D. L. and Barsev A. I. At. Spectrosc. 1980 1 47. Brown S. S. Nomoto S. Stoeppler M. and Sunderman F. W. Pure Appl. Chem. 1981 53 773. Ramesh Babu D. and Naidu P. R. Talanta 1991 38 175. Cano-Pavon J. M. Vereda-Alonso E. Bosch-Ojeda C. and Garcia de Torres A. Anal. Lett. 1991 24 153. Vereda-Alonso E. Garcia de Torres A. and Cano-Pavon J. M. Mikrochim. Acta 1993 110 41. Vereda-Alonso E. Cano-Pavon J. M. Garcia de Torres A.and Siles-Cordero M. T. Anal. Chim. Acta 1993 283 224. Jaffe R. Fernandez C. A. and Alvardo J. Talanta 1992,39,113. Alvarado J. and Cristiano A. R. J. Anal. At. Spectrom. 1993 8 253. Shengiun M. and Holcombe J. A. Talanta 1991 38 503. Vollkopf U. Grobenski Z. and Welz B. At. Spectrosc. 1981 2 68. Pruszkowska E. and Barrett P. Spectrochim. Acta Part B 1984 39 485. Sunderman F. W. Chrisostomo M. C. Reid M. C. and Nomoto S. Ann. Clin. Lab. Sci. 1984 14 232. Slavin W. and Carnrick G. R. Spectrochim. Acta Part B 1984 39 271. 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 Andersen J. R. Gammelgaard B. and Reimert S. Analyst 1986 111 721. Nixon D. E. Moyer T. P. Squillace D. P. and MacCarthy J. T. Analyst 1989 114 1671.Patriarca M. and Fell G. S. J. Anal. At. Spectrom. 1994 9 457. Hinderberger E. J. Kaiser M. L and Koirtyohann S. R. At. Spectrosc. 1981 2 1. Slavin W. Manning D. C. and Carnrick G. R. At. Spectrosc. 1981 2 137. Manning D. C. and Slavin W. Appl. Spectrosc. 1983 37 1. Grobenski Z. Lehmann R. Radziuk B. and Vollkopf U. At. Spectrosc. 1984 5 87. Jaganathan J. Ewing K. J. Buckley E. A. Peitersen L. and Aggarwal I. D. Microchem. J. 1990 41 106. Schmidt W. and Diett F. Fresenius’ 2. Anal. Chem. 1980 308 129. Welz B. Schlemmer G. and Mudakavi J. R. J. Anal. At. Spectrom. 1988 3 695. Hinds M. W. and Jackson K. W. J. Anal. At. Spectrom. 1990 5 199. Akman S. Genc 0 Ozdural A. R. and Balkis T. Spectrochim. Acta Part B 1980 35 373. Pupyshev A. A. and Nagdeev U. K. J. Appl. Spectrosc. (USSR) 1979 33 227.MacAllister T. J. Anal. At. Spectrom. 1994 9 427. Bahreyni-Toosi M. H. and Dawson J. B. Analyst 1983,108,225. Halls D. J. Analyst 1984 109 1081. Subramanian K. S. At. Spectrosc. 1987 8 7. Halls D. J. J. Anal. At. Spectrom. 1989 4 149. Bradshaw D. and Slavin W. Spectrochim. Acta Part B 1989 44 1245. Slavin W. Manning D. C. and Carnrick G. R. Spectrochim. Acta Part B 1989 44 1237. Bermejo-Barrera P. Moreda-Piiieiro A. Moreda-Piiieiro J. and Bermejo-Barrera A. Anal. Chim. Acta 1995 310 355. Granadillo V. A. Navarro J. A. and Romero R. A. J. Anal. At. Spectrom. 1993 8 615. Penninchx W. Massart D. L. and Smeyers-Verbeke J. Fresenius’ J. Anal. Chem. 1992 343 526. Navarro J. A. Granadillo V. A. Parra 0. E. and Romero R. A. J. Anal. At. Spectrom. 1989 4 401. Halls D. J. Mohl C. and Stoeppler M. Analyst 1987 112 185. Hinds M. W. Katyal M. and Jackson K. W. J. Anal. At. Spectrom. 1988 3 997. Miller-Ihli N. J. and Greene F. E. JAOAC Int. 1992 75 354. Miller J. C. and Miller J. N. Statistics for Analytical Chemistry Wiley New York 1st edn. 1984 ch. 3. Czobik E. J. and Matousek J. P. Anal. Chem. 1978 50 2. Sedykh E. M. Belyaev Y. I. and Sorokina E. V. J. Anal. Chem. (USSR) 1980,35 2162. Erspamer J. P. and Niemczyk T. M. Anal. Chem. 1982,54,2150. Paper 5/03523F Receiued June 1 1995 Accepted July 27 1995 Journal of Analytical Atomic Spectrometry November 1995 Vol. 10 I01 7

ISSN:0267-9477    

DOI:10.1039/JA9951001011

出版商:RSC    

年代:1995

数据来源: RSC

摘要:

Direct Coupling of High-performance Liquid Chromatography to Microwave= induced Plasma Atomic Emission Spectrometry via Volatile-species Generation and Its Application to Mercury and Arsenic Speciation Journal of Analytical Atomic Spectrometry JOSE M. COSTA-FERNANDEZ FLORIAN LUNZER,* ROSARIO PEREIRO-GARCIA AND ALFRED0 SANZ-MEDEL? Department of Physical and Analytical Chemistry Faculty of Chemistry CIJulihn Claveria 8 University of Oviedo 33006 Oviedo Spain NEREA BORDEL-GARCiA Department of Physics Faculty of Science University of Oviedo Spain The on-line coupling of vesicle-mediated high-performance liquid chromatography (HPLC) to low-power argon microwave-induced plasma (MIP) detection is described. The analytical potential of such a hybrid technique is illustrated for the speciation of mercury and arsenic compounds.Continuous cold vapour (CV) or hydride generation (HG) techniques were used as interfaces between the exit of the HPLC column and the MIP held in a surfatron at reduced pressure. Detection was by atomic emission spectrometry (AES). The effect of different surfactants on mercury CV generation was evaluated using SnCl as the reducing solution instead of sodium tetrahydroborate(1n). Emission signals increased by about 75% by adding the vesicle-forming surfactant didodecyldimethylammonium bromide (employed as the HPLC mobile phase for speciation). Enhancements of around 100% of signals were found in micelles of cetyltrimethylammmonium bromide. The detection limits by vesicular HPLC-HG-MIP-AES for the more toxic arsenic species investigated (namely arseneous arsenic monomethylarsonic and dimethylarsinic acids) were in the range 1-6 ng ml-'.The detection limits for mercury speciation by vesicular HPLC-CV-MIP-AES were 0.15 ng ml-' Hg for inorganic mercury and 0.35 ng ml-' Hg for methylmercury. Both methods have been successfully applied to the speciation of mercury and arsenic in natural waters (sea-water and tap water) and in human urine. Keywords Vesicles; high-performance liquid chromatography; arsenic speciation; mercury speciation; microwave-induced plasma; atomic emission spectrometry Nowadays the potential of plasma-based sources to be the most sensitive and specific atomic detectors for chromatogra- phy is well recognized. Microwave-induced plasmas (MIPs) in particular have aroused great interest as gas chromato- graphic detectors.'-3 Conversely the coupling and applications of MIPs to high-performance liquid chromatographic (HPLC) separations have proved very since major difficulties arise from the low tolerance of conventional low power MIPs to the introduction of liquid aerosols (or to vapours from organic solvents).Owing to these shortcomings attempts to couple HPLC systems to low power MIPs are * On leave from Graz University of Technology 8010 Graz Austria. t To whom correspondence should be addressed. scarce and they usually involve the design of complicated and inefficient interfaces. For instance complex interfaces using a moving-wheel sample transport desolvation system (in which the solvent is evaporated by a flow of hot inert gas leading the dry analyte to the MIP8) or a direct introduction system involving the vaporization of the liquid mobile phase dried over a heated wire by a cross-stream of He (ref.4) have been published. The direct nebulization of the HPLC column effluent to Ar-0 low power plasmas6 or to He moderate power (500 W) MIPS' has also been described; unfortunately quenching of the discharge and serious spectral and excitation interferences originating from the hydro-organic mobile phases entering the plasma were observed. Hydride generation (HG) has been explored extensively as a means of improving the sensitivity of HPLC-inductively coupled plasma (ICP) instruments,'O~'' being particularly useful for trace metal speciation where only one element has to be detected at a time as required by the chromatographic separation (determinations of the different species of the same element).It is clear that the formation of volatile species at the exit of the HPLC column could be a good interface to avoid the quenching of the MIP discharge by liquid flow aerosols. On the other hand it has been claimed' that MIPs generated at reduced pressure may have more resistance than atmospheric pressure plasmas to the entrance of molecular gases (i.e. H produced in hydride generation); in fact we have already reported detection limits by atomic emission spec- trometry (AES) in the low ng ml-' range for As Sb and Se using continuous HG directly coupled with a 50 Torr Ar-MIPI3 sustained in a surfatron. Moreover considerably lower detec- tion limits for mercury determinations were achieved by coup- ling cold vapour (CV) generation (using SnC1,-HCl as the reducing agent) to the low pressure MIP,14 and could be a favourable specific detector to be coupled with the HPLC column.The analytical performance of the proposed HPLC-HG- MIP-AES and HPLC-CV-MIP-AES systems is evaluated and its application to the speciation of arsenic and mercury in waters (tap or sea water) and human urine is discussed. Interest in using surfactant-based organized media as mobile phases for HPLC separation^'^,^^ has been growing during the last few year^,'^^^^-'^ and Sanz-Medel et have shown very recently the potential of utilization of surfactant vesicles such as didodecyldimethylammonium bromide (DDAB) as mobile phases in HPLC systems coupled with atomic absorption Journal of Analytical Atomic Spectrometry November 1995 Vol.10 101 9spectrometry (AAS) or ICP-AES for metal speciation. Rapid and reliable separations in aqueous media can be performed using these mobile phases and a modified C1,-bonded silica stationary phase both for arsenic" and mercurylg speciation. In the present paper the strategy of on-line interfacing the separation and detection via vesicle-mediated HPLC separa- tions-volatile species generation has been extended to MIP- AES detection. The potential of MIPS as HPLC detectors for metal speciation is therefore evaluated for the first time. Inorganic mercury (Hg") and methylmercury (MeHg) were chosen for mercury speciation while the more toxic arsenic species arseneous (As"') arsenic (As') monomethylarsonic (MMAs) and dimethylarsinic (DMAs) acids were selected as 'model' analytes for arsenic.EXPERIMENTAL Instrument a tion The experimental set-up used for the HPLC-MIP experiments is shown schematically in Fig. 1. A Knauer Model 6400 HPLC pump with an attached sample injection valve equipped with a 100 pl loop was used for eluent delivery and sample introduc- tion. The separation column was a Spherisorb ODS 2 (250 x 4.6 rnm id) packed with 10 pm C,,-bonded silica station- ary phase previously modified by passing DDAB solution as described e1~ewhere.l~ A four-channel peristaltic pump HP4 Minipuls 2 Gilson and a laboratory-made gas-liquid separator constituted the con- tinuous hydride or cold vapour generation systems.Details of the MIP-AES set-up components the low pressure equipment and the optical detection system have been given e1~ewhere.l~~~~ The rate of aspiration of the vacuum pump in the reduced pressure MIP was matched to the total flow rate of the plasma gas with the aid of a throttling clamp located before the plasma region. As Fig. 1 shows the plasma was viewed axially through a silica window fitted in a brass chamber connected to the open end of the plasma tube. An ultrasonic device from Sonics & Materials (CT USA) Model VC (500 W) was used for the preparation of DDAB vesicles from the corresponding surfactant solutions. For non-chromatographic experiments with mercury a con- tinuous CV generation system as shown in Fig. 2 was used. Reagents A lo00 mg 1-1 stock solution of Hg" was obtained from Merck atomic absorption standards.Methylmercury stock solution (lo00 mg 1-I) was prepared by dissolving the appropriate amount of methylmercury chloride salt (Merck) in 50 ml of methanol which was made up to 100ml with ultrapure Milli-Q water. Stock solutions (1000 mg 1-I) of DMAs and MMAs were prepared by dissolving respectively (CH,)2AsOzNa3Hz0 (Sigma) and CH,ASO(ON~)~.~H,O HPLC SYSTEM INJECTION HPLC PUMP SEPARATION COLUMN VESICULAR 6 MOBILE PHASE INTERFACE detector Peristaltic ' r l l Pump Gas-liquid separator Fig.2 Continuous cold vapour generation system for the study of the effect of surfactants (Carlo Erba) into Milli-Q water. Arsenic oxide (Merck) was dissolved in 0.5 mol 1-1 NaOH solution and then diluted to a final concentration of lOOOmgl-' with 0.6moll-' of HC1.Arsenic(v) stock solution (1000 mg 1-') was obtained from Merck. All stock solutions were stored in dark glass bottles at 4°C. Working standard solutions of the compounds were freshly prepared daily by appropriate dilution of the stock solutions with ultrapure Milli-Q water. A 3% m/v SnCl (Merck) solution was prepared in 2.4 moll- NaOH. Possible mercury impurities were elimin- ated by constant purging with argon for 4 h at 0.2 1 min-'. An oxidant solution of 5% m/v K2S208 (Merck) was prepared in 0.5 moll-' H2S04 (Merck) containing 1.6 mmol 1-l of CuSO (Merck). The NaBH solution (0.5% m/v) was prepared by dissolving tetrahydroborate(n1) powder (Probus) in water stab- ilized by 0.5% NaOH m/v (final concentration); this solution was prepared weekly and filtered before use.moll-') were prepared by dis- solving the appropriate amount of DDAB (Fluka) in 500 ml of Milli-Q water. Surfactant-vesicles of DDAB were prepared from such solutions by dilution with Milli-Q water. Surfactant- vesicles of DDAB were prepared from such solutions by dilution with Milli-Q water and sonication with a power output of 60 W for 10 min. Other surfactant solutions Le. Triton X- 100 sodium dodecyl sulfate (SDS) and cetyltrimethylammonium bromide (CTAB) were prepared by dissolving the appropriate amount of the solid compound in Milli-Q water. Methanol (HPLC grade from Romile Chemicals) and 0.005% v/v mercaptoethanol (Merck) were also used. Argon (99.998%) was used as the plasma gas. All other chemicals were of analytical-reagent grade and distilled deionized (Milli-Q system Millipore) water was used throughout.DDAB solutions (1 x Procedures Table 1 gives all the relevant experimental conditions and MIP-AES operating parameters finally selected for separation/ REDUCED PRESSURE-MIP To " O d A .... PERISTALTIC PLASMA PUMP SURFATRON To the vacuum DESICCATOR - DETECTOR 1 Fig. 1 Schematic diagram of the experimental systems GLS gas-liquid separator; CL clamp; A oxidant mixture (in the cold vapour system) or HCI (in hydride generation); B reducing solution (SnCl in the cold vapour system or NaBH in the hydride generation system). A detailed diagram of the MIP-AES system is shown in a previous p~blication'~ 1020 Journal of Analytical Atomic Spectrometry November 1995 Vol.10Table 1 Experimental conditions for As and Hg speciation HPLC mobile phase Mercury Arsenic CV (mercury speciation) Oxidant mixture Reducing solution HG (arsenic speciation) HCI NaBH MIP-AES 10 mmol 1-l ammonium acetate buffer + 0.005 % 2-mercaptoethanol-tO.2 mmol 1-l vesicle of DDAB pH = 5 at 1.5 ml min-' 5 mmol 1- ' NaH,PO + 1 % methanol + 0.01 mmol 1- vesicles of DDAB pH = 5.75 at 1.3 ml min-l 5% K,S,OB (in 0.5 mol I-' H2SO,+O.OOl6 moll-' CuS04); flow rate 0.4 ml min-' 3% SnCI (in 2.4 mol 1-' NaOH); flow rate 0.8 ml min-l 15% at a flow rate of 0.4 ml min-' 0.5% at a flow rate of 0.4 ml min-' wavelength 253.6 nm for Hg 228.8 nm for As; Ar flow 120ml min-l; pressure 50 Torr; forward power 75 W for Hg and 115 W for As detection of Hg and As species (after the corresponding optim- ization procedures).Since better detection limits were reported for argon than for helium low pressure discharges in the determination of arsenic by hydride generati~n'~ and mercury by cold v a p o ~ r ~ ~ Ar was chosen as the plasma gas for our experiments. Mercury speciation The mobile phase was prepared by dilution of 1 x lo-' moll-' vesicular DDAB solution to obtain a concentration of 2 x lo- mol I-' of DDAB in water and 0.005% v/v of 2-mercaptoethanol was added. The solution was buffered with ammonia acetate (10 mmol 1-') at pH = 5 and de-gassed by ultrasonication for 30 min prior to use. The mobile phase was continuously pumped through the column at a flow rate of 1.5 ml min-'. The eluent at the exit of the column was first mixed with an oxidant solution of K2S208 in H2S04 and CuSO which acts as a catalystz0 for the destruction of the possible mercury organic compounds.Then the flow was mixed on-line with a solution of SnC1 in NaOH medium for the mercury CV generation. Volatile species were continuously swept through the gas-liquid separator and mixed with the plasma gas which thus carried the gaseous analyte to the MIP (after passing through a H,S04 desiccator as shown in Fig. 1). Arsenic speciation The mobile phase employed and the flow rate used through the column for arsenic speciation is shown in Table 1. The HPLC eluent was first mixed at the exit of the column on-line with HCl solution and then mixed with a sodium tetrahydrobo- rate solution (see Fig. 1). As in the mercury speciation manifold the volatile species formed are separated from the liquid and directed to the low pressure MIP after passing through a desiccator (H,PO,).All separations were performed at room temperature under isocratic conditions. Peak heights from the chromatograms were used in all quantifications. Sample collection and pre-treatment Sea-water samples were collected from a Spanish coastal region (Gij6n) of the Cantabric Sea and immediately acidified (by the addition of ultrapure nitric acid) to a final pH of 2 for storage in pre-cleaned poly( propylene) bottles. The samples were fil- tered through a Millipore 0.45 pm membrane. The pH was adjusted to 7 with dilute ammonia solution before the analysis. The tap water samples were filtered through a Millipore 0.45 pm membrane before their direct introduction into the HPLC system.The human urine samples were first filtered through a Whatman grade 4 filter and then the filtered liquid was passed through a Millipore 0.45 pm membrane before the injection into the HPLC system. Since all the samples under study contained no detectable mercury or arsenic by our techniques they were spiked with the two relevant mercury species or alternatively with the four toxic arsenic compounds at different concentration levels. Such spiked samples were used for the validation of the proposed methods. Surfactant media for CVgeneration Fig. 2 shows schematically the continuous flow system used for the study of the effect of several surfactants on the generation of mercury CV. The Hg" solution (2 ng ml-') prepared in different surfactant media is continuously pumped through one of the channels by a peristaltic pump at a flow rate of 1.5 ml min-' while an oxidant solution (5% K,S,O in 0.5moll-' of H2S04 and 0.0016mol1-' CuSO,) is pumped through the second channel at a flow rate of 0.4mlmin-'.Both flows merge at a T-piece and the resulting solution is mixed downstream in a second T-piece with a reducing solution (3% SnCl in 2.4 mol I-' NaOH) and pumped through a third channel of the peristaltic pump at a flow rate of 0.8 ml min-'. The generated volatile Hg is then continuously separated from the liquid in the gas-liquid separator and directed to the MIP. RESULTS AND DISCUSSION HPLC-CV-MIP-AES Coupling and its Application to Mercury Speciation Previous work in this lab~ratory'~ showed that the direct coupling of continuous CV methodologies to the low pressure MIP using tin chloride as reducing agent provides very good sensitivities for mercury determinations; therefore this reduct- ant and specific detector were chosen for on-line mercury detection.Since methylmercury is not reduced to atomic mercury with SnC12 an on-line step for oxidation of methyl- mercury (using potassium peroxidisulphate in H2S04 medium and copper sulphate as catalyst'' was introduced prior to the reduction to elemental mercury with SnC12 in basic medium. The CV interface between the HPLC column and the MIP detector can be seen in detail in Fig. 1. Chemical concentrations and flow rates were optimized for the HPLC-CV-MIP-AES system under study by following a univariant search.The analytical parameter selected to be maximized was the ratio between Hg net intensity (IN) and plasma background intensity (IB) IN I,. The observed opti- mum figures for the on-line CV generation are compiled in Table 1. Efect of organized media in the determination of mercury It has been shown that appropriate ordered-media can enhance the generation of volatile species and so improve the perform- ance of some analytical atomic methods." In this context it has been described how surfactants can further enhance the analytical sensitivity of atomic methods based on CV or HG for As Cd or Hg;21-26 therefore the coupling of surfactant- based HPLC separations to CV o r HG methods commonly used to increase the sensitivity of atomic detectors could give Journal of Analytical Atomic Spectrometry November 1995 Vol.10 1021rise to synergic combinations for the speciation of such elements. The beneficial effects of organized media on CV methods for the determination of Hg" by AAS and ICP-AES using NaBH as reducing agent,22325 prompted us to study the effect of different surfactants in the determination of mercury by CV-MIP-AES that is using SnCl in basic medium as the reducing agent. The organized media tested were the vesicle-forming surfac- tant didodecyldimethylammonium bromide (also used as mobile phase in the HPLC system) and the micelle-forming surfactants at concentrations above their critical micellar concentration (CMC); i.e. cationic surfactant hexadecyl- trimethyl ammonium bromide CTAB (with a CMC= 9.2 x lo- mol 1-') anionic surfactant sodium dodecyl sulfate SDS (CMC=8.1 x mol 1-') and non-ionic surfactant Triton X-100 (CMC =2 x lo- moll-').The experimental set- up for sample introduction into the MIP used for these experiments is shown in Fig. 2. The IN IB ratios obtained when the Hg cold vapour was generated from each of these four media at different surfactant concentrations were compared with those obtained without surfactants. In contrast to the results previously reported for Hg CV generation with NaBH (using AAS and ICP-AES as detectors) where only positive effects were observed when the DDAB vesicles were formed in solutions already containing the ana- lyte,22 it was observed in our experiments that increases of sensitivity were obtained both when DDAB solution (without previous sonication) was added to a Hg" solution followed by sonication and dilution to volume and also when the Hg" was added to previous DDAB-sonicated vesicles.The higher increase observed in the first case (see Fig. 3) indicates the advantage of Hg" being wrapped into the DDAB vesicles as they form under sonication.22 Significant enhancements in sensitivity were also observed when thick multiple bilayer aggregates were assayed i.e. neither the DDAB nor the DDAB + Hg' solution were sonicated (Fig. 3). The observed effect on emission of Hg of adding micelle- forming surfactants at concentrations over the CMC is plotted in Fig. 4(a). An important enhancement (about 100%) in the Hg sensitivity was observed in the presence of the cationic surfactant CTAB.As can be seen with greater detail in Fig. 4(b) this enhancement only appears to take place at concentrations near or above the CMC thus demonstrating that the observed positive effects for CTAB are produced only when micelles are present in the reaction medium. Mercury speciation There is no doubt that one major concern in any hyphenated detection system is the impact that the interface has upon 16 14 12 10 -m 2 a 2 0 :_.__._ 0 0.2 0.4 0.6 0.8 1 [DDAB]/mrnol I-' Pig. 3 Effect of DDAB concentration on the CV-MIP-AES signal for Hg A with sample sonication after Hg" has been spiked into the DDAB solution; B with sonication of the DDAB previously to the addition of the Hg"; and C without sonication (bilayer aggregates of DDAB) CMC = 9 .2 ~ 1 0 ~ mot r' 10 8 6 CMC = 2 ~ 1 0 ~ mol r1 I CMC = 8.1~1 O4 m ~ l I-' 0 *0 2 4 6 8 10 9 ,z Concentratiodx 2x1 O4 mol I-' for Triion X-lo0 x l ~ l O - ~ mol I-' for CTAB and SDS 4 -N 0 0.25 0.5 0.75 1 [CTABlhnmol r' Fig.4 Effect of different organized media on the IN:IB of Hg at 253.6 nm (2 ng ml-1 of Hg A cationic tensoactive cetyltrimetylam- monium bromide (CTAB); B non-ionic tensoactive Triton X- 100 and C anionic tensoactive sodium dodecyl sulfate (SDS) (a) Effect observed with concentrations of tensoactives over the critical micellar concentration (CMC). (b) Effect observed with concentrations of CTAB below the CMC resolution peak shape and efficiency. In this sense two gas- liquid separator designs were assayed in the vesicular HPLC-CV-MIP-AES system; a U-type GLS27 was filled with 3 mm od glass beads in order to allow a smooth separation of gases from liquids and the same GLS design without the glass beads was tested.Improvements of around 40% in the detec- tion limits (DLs) both for inorganic Hg and methylmercury were observed by removing the glass beads from the column of the GLS. This behaviour could be explained if it is considered that mercury has a high tendency to be retained in the glass beads (which should be coated with the DDAB) thus giving rise to broader transient signals in the plasma. Consequently a poorer resolution of the chromatographic peaks is also obtained in the first case. Table 1 summarizes optimum conditions selected for the CV interface operation for mercury optimum speciation.The ana- lytical performance characteristics obtained are given in Table 2. The detection limits estimated as three times the standard deviation of the baseline signal were in tenths Table 2 Figures of merit of the HPLC-CV-MIP-AES method Retention time/ Mercury species min DL/ng ml-' RSD (Yo)* MeHg 6.22 0.35 6.8 Inorganic Hg 10.00 0.15 6.7 * Relative Standard Deviation of five injections calculated for 20 ng ml- of inorganic Hg and 50 ng ml-I of MeHg. 1022 Journal of Analytical Atomic Spectrometry November 1995 Vol. 10Table 3 Detection limits (ng ml-') for mercury speciation using HPLC separation methods MeHg Detection technique* CV-MIP-AES PN-ICP-MS CV-ICP-MS USN-ICP-MS CV-ICP-AES CV-AAS CV-AAS Sample loop/ Pl 100 100 100 200 200 100 100 Methyl Hg 0.35 7 0.6 0.7 37 10 50 Inorganic Hg 0.15 16 1.2 0.4 35 16 10 Mode of DL calculation Ref.3 times the standard deviation of the baseline signal 2 1 signal to noise ratio 2 1 signal to noise ratio 3 times the standard deviation of the baseline signal Minimum concentration that gives a signal twice the peak-to-peak noise 3 times the baseline noise 3 times the blank value This work 28 28 29 30 19 31 * PN = pneumatic nebulization; MS =mass spectrometry; USN = ultrasonic nebulization. of ng ml-' for both mercury species. As can be seen in Table 3 the obtained DLs compare favourably with those reported with other atomic spectrometric methods for mercury speci- ation including a much more expensive detector such as the ICP-MS. The precision of the determinations was investigated by analysing five replicates (five injections of a mixed solution of both mercury compounds at known concentrations). The relative standard deviations (RSD%) of the peak heights were calculated for both forms of mercury and as can be seen from Table 2 similar precision levels (around 7%) were obtained for both species.Fig. 5 shows a typical chromatogram obtained with the proposed method. The applicability of this hybrid HPLC-CV-MIP-AES method to real samples for mercury speciation was investigated in sea-water and human urine. The corresponding real samples were filtered and prepared as indicated under Experimental before the final injection in the continuous flow system. The values obtained for these recovery tests (mean of five injections) are given in Table 4 both for inorganic mercury and methylmer- cury and are in good agreement with expected values with slightly poorer results in urine matrix than in sea-water.HPLC-HG-MIP-AES Coupling for Arsenic Speciation The feasibility of using vesicular mobile phases in HPLC with plasma detection for the speciation of toxic As species was t x .= C c - Timdmin Fig. 5 Typical chromatogram of a standard mixture containing 50 ng ml-' of MeHg and 20 ng ml-I of inorganic mercury. Experimental conditions as described in Table 1 Table 4 Recovery study of inorganic mercury and methylmercury added to sea-water and human urine Mercury Spiked/ Recovery in species ng ml-' sea-water (%) Inorganic Hg 20 98 10 101 5 105 Methyl Hg 50 91 25 103 10 96 Recovery in human urine (YO) 98 102 112 96 91 91 first demonstrated using ICP-AES detection." We have evalu- ated here the coupling of such vesicular HPLC separation to a low power MIP-AES at reduced pressure as detector and using on-line HG as interface.It has been already described that the presence of DDAB vesicles in the reaction medium enhances the detection limits attainable for arsenic by HG-AAS and HG-ICP-AES.22 Several designs of gas-liquid separators (Fig. 6) were assayed for on-line operation (see the separator position in Fig. 1). The common problem of all gas-liquid separators tested except Model 1 was that the gas stream was not as smooth as when the Browner-type separator filled with glass beads was used,27 and although separator Models 2 and 432 provided higher signals owing to their small dead volumes the noise produced by the bubbles of H which form themselves when NaBH merges with the acid solution was so high that detection limits were worse than those obtained with Model 1.Separator 3 exhibited the worst analytical performance characteristics. sample + +-I to detector sample + Ar f 4-+ 1 crn SEPARATOR MODEL 1 SEPARATOR MODEL 2 to detector 4 sample+Ar 15 mm sample to detector W SEPARATOR MODEL 4 SEPARATOR MODEL 3 Fig. 6 Laboratory-built gas-liquid separators made of glass Journal of Analytical Atomic Spectrometry November 1995 VoE. 10 1023Therefore the separator Model 1 was chosen for further experiments. It is also important to note that concentrated H2S04 had to be discarded as the drying agent (it was the common desiccant employed in previous workI4) as it was observed that the peak of DMAs was strongly diminished probably because dimethylarsine is absorbed by the sulfuric acid.The use of concentrated H3P04 did not produce such effect and was used as desiccant in further experiments for arsenic speciation. The results of optimization experiments for the four arsenic compounds speciation are summarized in Table 1. Analytical performance characteristics for As speciation obtained under the conditions given in Table 1 are shown in Table 5. As can be seen detection limits were in the range 1-6ngml-' for the four arsenic species under study and compare well with those obtained with other more powerful atomic sources such as the 1CP.l' Fig. 7 shows a typical chromatogram obtained with MIP-AES detection.The arsenic speciation method was verified for the determi- nation of the four species in tap water and human urine samples. The values for the real sample analyses correspond to the mean of three injections (Table6). As can be seen the recoveries are satisfactory for this type of analysis and are around loo+_ 10%. In order to examine the atomization power of the low power low pressure MIP used in this work and its efficiency to atomize/excite the formed organometallic hydrides the same HPLC-HG system was connected on-line to an atmospheric pressure ICP at 1 kW rf forward power. The experimental conditions for the ICP are similar to those reported elsewhere." As shown in Fig.8 the obtained chromatogram with ICP Table 5 Figures of merit of the HPLC-HG-MIP-AES method Retention time/ DL*/ RSD (Yo) Arsenic species min ng m1-l (n=5) As"' 2.3 1 2.8 DMAs 4.1 6 3.1 MMAs 9.5 1.2 2.1 AsV 12 5 2.5 *Calculated as three times the standard deviation of the baseline signal.I As"' t x .t C LI - Fig. 7 Typical chromatogram of a standard mixture containing 250 ng ml-I of As" and MMAs and 500 ng ml-' of As' and DMAs. Experimental conditions as described in Table 1 Table6 Recovery study of arsenic species added to tap water and human urine Recovery Recovery in in human Arsenic species Spiked/ng ml-' tap water (%) urine (%) As"' 300 101 101 DMAs 400 99 104 MMAs 200 107 107 As' 400 97 111 I As"' t x .f C c - Time/min Fig. 8 Typical chromatogram of a standard mixture of 1 mg ml-l of inorganic As"' and MMAs and 2 mg ml-I of As' and DMAs obtained with an atmospheric Ar-ICP-AES as detection system.ICP experimen- tal conditions as in ref. 10 detection had the same shape as that obtained using MIP for detection (Fig. 7). These results suggest that the atomization of the inorganic arsine and the organometallic hydride compounds formed from MMAs and DMAs with NaBH is similar for both plasmas. Therefore the lower sensitivity for such compounds compared with arsine formed from As"' is not an artifact produced by the low pressure MIP detector but it is probably due to different hydride formation/volatilization efficiencies of the corresponding hydrides formed. CONCLUSIONS The present study shows a successful coupling of HPLC separation on-line with a MIP-AES detector and represents the first application of HPLC-MIP-AES coupling to tackle the modern challenge of trace metal speciation. Although most MIP detections (particularly for the analysis of non-metals) is carried out using the plasmas Ar was preferred since improve- ments in DLs of > 200% can be obtained for the determination of As by HG13 and a four times improvement in DLs for the determination of mercury by CV14 using in both cases the low pressure discharge.The low cost of the detection/excitation source and its maintenance (total flow rates of 120 ml min-' of Ar are used as the plasma gas) the low detection limits found for arsenic species (which are comparable to those obtained with other more powerful sources such as the ICP) and the very good detection limits obtained in the case of mercury reveal the reduced-pressure low power MIP as a significant atomic emission source to address present arsenic and mercury speciation problems.The experiments carried out here also show the importance of the selection of the gas-liquid separator which depends upon the particular analyte and upon the possibility of pro- duction of important amounts of gaseous byproducts such as H2 generated on-line during the chemical reduction step. The use of CTAB micelles or DDAB vesicles improved the determination of mercury by CV-MIP-AES using SnClz in basic media as the reducing agent complementing previous findings using NaBH4,22 and giving a new example of the beneficial use of surfactants to enhance some atomic spectro- metric determinations.21 Financial support from DGICYT (Spain) through project PB91-0669 is gratefully acknowledged.The authors wish to thank Jesus Zaton for technical assistance in the construction of some of the equipment. REFERENCES 1 McCormack A. J. Tong S. C. and Cooke W. D. Anal. Chem. 1965,37 1470. 1024 Journal of Analytical Atomic Spectrometry November 1995 Vol. 102 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Quimby B. D. Uden P. C. and Barnes R. M. Anal. Chem. 1978,50,2112. Long G. L. Ducatte G. R. and Lancaster E. D. Spectrochim. Acta Part B 1994 49 75. Billiet H. A. H. van Dalen J. P. J. Schoenmakers P. J. and De Galan L. Anal. Chem. 1983 55 847. Huf F. A. and Jansen G. W. Spectrochim. Acta Part B 1983 38 1061. Kollotzek D. Oeschsle D. Kaiser G. Tschopel P. Tolg G. Fresenius Z Anal. Chem.1984 318 485. Jansen G. W. Huf F. A. and de Jong H. J. Spectrochirn. Acta Part B 1985 40 307. Zhang L. Carnahan J. W. Winans R. E. and Neill P. H. Anal. Chem. 1989,61 895. Michlewicz K. G. and Carnahan J. W. Anal. Lett. 1987,20,1193. Liu Y. M. Fernandez Sanchez M. L. Blanco Gonzalez E. and Sanz-Medel A. J. Anal. At. Spectrom. 1993 8 815. Le X.-C. Cullen W. R. and Reimer K. J. Talanta 1994 41,495. Rivibre B. Mermet J. M. and Deruaz D. J. Anal. At. Spectrom. 1988 3 551. Lunzer F. Pereiro-Garcia R. Bordel-Garcia N. and Sanz- Medel A. J. Anal. At. Spectrom. 1995 10 311. Costa-Fernandez J. M. Pereiro-Garcia R. Bordel-Garcia N. and Sanz-Medel A. J. Anal. At. Spectrom. in the press. Armstrong D. W. Am. Lab. 1981 13 14. Hinze W. L. and Armstrong D. W. Ordered Media in Chemical Separations ACS Symposium Series No 342 American Chemical Society Washington DC 1987.Vela N. P. and Caruso J. A. J. Anal. At. Spectrom. 1993 8 787. Sanz-Medel A. Aizpun B. Marchante J. M. Segovia E. Fernandez M. L. and Blanco E. J. Chromatogr. A 1994,683,233. 19 20 21 22 23 24 25 26 27 28 29 30 31 32 Aizpun B. Fernandez M. L. Blanco E. and Sanz-Medel A. J. Anal. At. Spectrom. 1994 9 1279. Munaf E. Haraguchi H. Ishii D. Takeuchi T. and Goto M. Anal. Chim. Acta 1990 235 399. Sanz-Medel A. Fernandez de la Campa M. R. Valdes-Hevia y Temprano M. C. Aizpun Fernandez B. Liu Y. M. Talanta 1993 40 1759. Aizpun Fernandez B. ValdCs-Hevia y Temprano C. Fernandez de la Campa M. R. Sanz-Medel A. and Neil P. Talanta 1992 39 1992. Aizpun Fernandez B. Fernandez de la Campa M. R. and Sanz- Medel A. J. Anal. At. Spectrom. 1993 8 1097. Sanz Medel A. Valdes-Hevia M. C. Bordel-Garcia N. and Fernandez de la Campa M. R. Anal. Chem. in the press. Gutierrez J. M. Madrid Y. and Camara C. Spectrochim. Acta Part B 1993 48 1551. Madrid Y. Gutierrez J. M. and Camara C. Spectrochim. Acta Part B 1994 49 163. Pyen G. S. Long S. and Browner R. F. Appl. Spectrosc. 1986 40 246. Bushee D. S. Analyst 1988 113 1167. Huang C. W. and Jiang S. J. J. Anal. At. Spectrom. 1993 8 681. Krull I. S. Bushee D. S. Schleicher R. G. and Smith S. B. Jr. Analyst 1986 111 345. Sarzanini C. Sacchero G. Aceto M. Abollino O. and Mentasti E. Anal. Chim. Acta 1994 284 661. Michaelis M. R. A. Ph.D. dissertation Graz University of Technology Graz Austria 1990. Paper5/03096 J Received May 16 1995 Accepted July 18 1995 Journal of Analytical Atomic Spectrometry November 1995 Vol. 10 1025

ISSN:0267-9477    

DOI:10.1039/JA9951001019

出版商:RSC    

年代:1995

数据来源: RSC

摘要:

COMMUNICATION Assessment of Direct Solid Sample Analysis by Graphite Pellet Electrothermal Vaporization Inductively Coupled Plasma Mass Spectrometry J. M. REN R. RATTRAY AND ERIC D. SALIN Department of Chemistry McGiEl University Montreal Quebec Canada H3A 2K6 D. CONRAD GREGOIRE Geological Survey of Canada 601 Booth Street Ottawa Ontario Canada K1 A OE8 Graphite pellet vaporization was evaluated as a method for solid sample introduction for ICP-MS. A detection limit of 0.1 ppb of Cd was obtained indicating that the technique might be very useful for screening. The accuracy was poor using the technique of external standards and four reference materials suggesting that the technique of standard additions might be required if the technique was to be used for anything other than semi-quantitative measurements.Keywords ICP-MS; solid sample analysis; sample introduction; pellet; graphite Since its commercial introduction inductively coupled plasma mass spectrometry (ICP-MS) has found widespread use in environmental geological and biochemical research. As with inductively coupled plasma atomic emission spectrometry (ICP-AES) the most commonly used sample introduction technique for ICP-MS is pneumatic nebulization. The merits of this technique are its simplicity high sample throughput good stability and low cost whereas low analyte transport efficiency and difficulties in handling high salt samples are among the frequently mentioned shortcomings. Arc nebulization,' direct sample in~ertion,~.~ laser ablation4-' and electrothermal vaporization ( ETV)8-12 are alternative sample introduction techniques for ICP-MS.The advantages of using ETV for sample introduction are lower detection limits due to improvement in analyte transport efficiency; reduced spectroscopic interferences from oxide and polyatomic ion formation as a result of solvent removal; the ability to analyse small samples; and the ability to analyse samples high in organics solids and acids. In a previous study,13 we described graphite pellet- electrothermal vaporization (ETV)-ICP-AES for solid sample analysis and obtained detection limits in the lower ppb range in the solid. This communication assesses the use of the same graphite pellet-ETV system for solid sample analysis with ICP-MS detection. EXPERIMENTAL The pellet-ETV system used in this study has been described previ~usly'~ and a schematic of the experimental setup is shown in Fig.1. Briefly solid sample powder (pure oxide or reference material) is mixed with graphite powder (Bay Carbon Bay City USA) and an accurately weighed portion of the resulting mixture is pressed manually into a cylindrical pellet (9 mm long with 4 mm in diameter). The pellet is then placed between two brass contact rings in the ETV device and is electrothermally heated in three steps drying pyrolysis and high temperature vaporization. A manually operated three- Journal of Analytical Atomic Spectrometry ' Ar Fig. 1 Pellet-ETV-ICP-MS system way two-position valve is used to vent the furnace during the drying and pyrolysis steps and to direct analyte vapour to the plasma during the vaporization step.The vapor is swept into the plasma by an Ar carrier gas stream. A Perkin-Elmer Sciex Elan 5000 ICP Mass Spectrometer was used. The experimental conditions for both the Elan 5000 and the ETV system are given in Table 1. RESULTS AND DISCUSSION Graphite Purification The commercially available spectroscopically pure graphite powder used to make the pellets was found to contain signifi- Table 1 Instrumental operating conditions and data acquisition parameters ICP mass spectrometer Coolant Ar flow Auxiliary Ar flow Carrier Ar flow E TV heating programme Drying step Rf power 1000 w 15.0 1 min-' 0.85 1 min-' 0.90 1 min-' TemperaturerC Ramp time/s Hold time/s oxide standards 100 1 2 reference materials 200 1 60 oxide standards 200 1 1 reference materials 250 1 120 oxide standards 1250 1 5 reference materials 2250 1 5 Pyrolysis step Vaporization step Data acquisition Dwell time 10 ms Scan mode Peak-hopping Analyte isotope "'Cd Signal measurement Area Journal of Analytical Atomic Spectrometry November 1995 Vol.10 1027cant Cd present as a containment. A purification method using inductive heating was devised as described below. A graphite cup (20 mm o.d. 20 mm height) was made out of porous graphite (ERG Oakland California USA). This cup was filled with graphite powder and placed on top of a glass rod inside a normal ICP torch. When 900 W of rf power was applied to the load coil the graphite cup and the powder inside was heated to incandescence. Two 2min cycles of radiofrequency (rf ) power application completed the purifi- cation of 0.4 g of graphite.By using this technique the blank Cd signal was reduced from 7700 to below 300 counts. Calibration Curve and Limit of Detection for Cd Pellets containing a range of concentrations of Cd present as CdO were prepared. Fig. 2 is a calibration curve (peak area uersus concentration) for Cd from 1 to 1000ppb. Curve a (Fig. 2) is the uncorrected calibration curve which exhibits some deviation from linearity. It has been suggested by Gregoire and Sturgeon14 that the argon dimer may be used as a diagnostic tool to indicate plasma loading effects. When the ratio of the Cd peak area to the argon dimer ATz+ signal intensity is plotted versus Cd concentration the linearity is improved as shown in curve b indicating that in this instance the argon dimer can be used as an internal standard.The instrumental detection limit for Cd was estimated from signals obtained from six blanks and the calibration curve shown in Fig. 2b. An instrumental detection limit (defined as the Cd concentration equivalent to 3 times the standard deviation of Cd signal obtained from the blanks) of 0.1 ppb was obtained. This value is one order of magnitude better than that found using ICP-AES dete~ti0n.l~ Analysis of Reference Materials Pellets containing sample graphite in a 1 10 ratio were pre- pared using two botanical reference materials (Ontario Ministry of Environment and Energy) and two marine sediment reference materials (National Research Council of Canada). Fig. 3(a) shows that for botanical samples two Cd peaks of varying intensity were obtained during the vaporization step suggesting two sources of Cd from the pellet.During heating of graphite pellets at temperatures above incandescence it was observed that pellets were heated unevenly with the central part of the pellet being much hotter that the two ends which were in contact with the water-cooled brass sample holder. During the pyrolysis step it is likely that Cd present in an I 1 Fig. 2 Calibration curves for Cd. PKA is peak area. a Cd peak area versus Cd concentration; b (Cd peak area/argon dimer peak area) uersus Cd concentration I 5 10 15 Seconds Fig. 3 Typical vaporization profile of Cd in reference materials (a) without pre-pyrolysis; (b) with pre-pyrolysis in a muffle furnace at 500°C for 1 h organic matrix and located at the centre of the pellet is converted from an organic to an inorganic form whereas material located near the cool ends of the pellet is not effectively pyrolysed and Cd remains in its original form.When vaporized at higher temperatures Cd originally from the central portion of the pellet is released and accounts for the first Cd peak. Cadmium bound in its original organic (unpyrolyzed) form is vaporized much later from the ends of the pellet and accounts for the second peak. The broad and irregular nature of both peaks indicates that pyrolysis may not be complete at either the centre or ends of the graphite pellet. No second peak was observed for CdO standard pellets or those containing sediment . In order to eliminate the second peak and to improve the shape of the analyte signal the mixture of graphite and botanical reference material was pre-pyrolysed in a muffle furnace at 500°C for 1 h prior to making the pellets.The resulting pellets contained uniformly pyrolysed sample material and produced a more intense and smooth single-peak Cd signal as is shown in Fig. 3(b). Table 2 shows the results for the analysis of reference mate- rials for Cd. Clearly from a standpoint of accuracy the technique provided disappointing results. Except for MESS-1 the Cd concentration found was at least twice the reference Table 2 Analysis of reference materials Sample Cd found (ppb) Cd certified (ppb) White Birch 75 36 Norway Maple 40 8 BCSS-1 22 10 MESS-1 35 23.6 1028 Journal of Analytical Atomic Spectrometry November 1995 Vol.10value. The vaporization of other sample components found in the biological materials may have increased the Cd transport efficiency13 relative to Cd originating from the oxide reference making it impossible to use external standards for calibration purposes. The detection limits are highly encouraging however and are well below the trace element levels of many natural materials. The poor accuracy but excellent detection limit suggests that the technique may be useful now as a rapid screening technique for toxic metals. However matrix-matched standards may be necessary to achieve the accuracy required for more demanding analyses. 4 5 6 7 8 9 10 11 12 13 REFERENCES 1 Jiang S.-J. and Houk R. S. Spectrochim. Acta Part B 1987 42 93.2 Boomer D. W. Powell M. Sing R. L. A. and Salin E. D. Anal. Chem. 1986,58 976. 3 Karanassios V. and Horlick G. Spectrochim. Acta Part B 1989 44 1361. 14 Pang H. Wiederin D. R. Houk R. S. and Yeung E. S. Anal. Chem. 1991,63 390. Moenke-Blankenburg L. Schumann T. Giinther D. Kuss H.-H. and Paul M. J. Anal. At. Spectrom. 1992 7 251. Yasuhara H. Okana T. and Matsumara T. Analyst 1992 117 395. Gray A. L. Analyst 1985 110 551. Gregorie D. C. J. Anal. At. Spectrom. 1988,3 309. Gray A. L. and Date A. R. Analyst 1983 108 1033. Park C. J. and Hall G. E. M. J. Anal. At. Spectrom. 1987,2,473. Ediger R. D. and Beres S. A. Spectrochim. Actu Part B 1992 47 907. Shibata N. Fudagawa N. and Kubota M. Spectrochim. Acta Part B 1993,48 1127. Karanassios V. Ren J. M. and Salin E. D. J. Anal. At. Spectrom. 1991 6 527. Gregoire D. C. and Sturgeon R. E. Spectrochim. Acta Part B 1993 48 1347. Paper 5/00827A Received February 10 1995 Accepted September 7 1995 Journal of Analytical Atomic Spectrometry November 1995 Vol. 10 1029

ISSN:0267-9477    

DOI:10.1039/JA9951001027

出版商:RSC    

年代:1995

数据来源: RSC

摘要:

CUMULATIVE AUTHOR INDEX JANUARY-NOVEMBER 1994 Abell I. D. 591 Aboal-somoza Manuel 227 Adams Freddy C. 11 1 Alexandrova Anka 799 Allen Lloyd A. 267 Alvarado JosC 483 Alvarez Nestor 487 Alvarez Walter Oliver L. 487 Amarasiriwardena Dulasiri 505 Arbore Philippe 381 Arpadjan Sonja 799 Arruda Marco A. Z. 55 501 Asami Naoto 999 Asensio Jesus Sanz 975 Axner Ove 539 Barciela-alonso M. C. 247 Barnes Barbara S. 177 Barnes Ramon M. 505 935 Baroni. U 555 Baxter Douglas C. 711 769 Beatriz de La Calle Guntifias Becker Johanna Sabine 637 Becker-Ross Helmut 61 127 Beissler Hermann 885 Belazi Abd Ulhafid 233 Bellini Alessandra 433 Bermejo-Barrera Adela 227 Bermejo-Barrera Pilar 227 247 Bernal J. Galban 975 Betti Maria 381 Bettinelli Maurizio 555 Bizzarri. G 555 Boonen Sylvie 81 Bordel-Garcia Nerea 31 1 649 Borodin Alexander V.703 Boulos Maher I. 935 941 Bratter Peter 487 Brockhoff Carol A. 443 Broekaert JosC A. C. 583 849 Brown Garrett N. 527 Brueggemeyer Thomas W. 177 Brunetto M. R. 343 479 Bryant M. F. 295 Bulska Ewa 49 Burden Trevor J. 259 Burguera J. L. 343 473,479 Burguera M. 343,473 479 Bye Ragnar 803 809 Cabon J. Y. 993 Cabrera Horacio P. 511 Caldwell Kathleen L. 367 Camara Carmen 321 815 871 Carrero P. 343 479 Caruso Joseph A. 7 601 853 Castle Laurence 303 Cavalli Paolo 885 Cernohorskp TomaS 155 Cervera M. L. 353 Chakraborty Ruma 353 Chan W. T. 295 Chekalin Nikolai 539 Chen Chih-jung 955 Chen Hengwu 533 Chen Xiaoshan 837 Cho Jung H. 335 Cho Kyu H. 335 94 1 M. 815 897 145 247 1011 1011 671 1019 Cielo P. 643 Cimadevilla Enrique Alvarez- Cabal 149 Cloud Jacques 287 Coedo Aurora G.449 859 Corns Warren T 287 Cossa Daniel 287 Costa-Fernandez Jose M. 649 Creed John T. 443 Crews Helen M. 303 625 Crighton J. 591 Cristiano Ana Rita 483 Crowe John B. 177 Curtius Adilson JosC 329 483 Dams Richard 81 569 575 Das Arabinda K. 353 Davidson Christine M 233 241 Deaker M. 423 de Bibvre Paul 395 de Gendt Stefan 681 689 de la Calle Guntinas M. de la Guardia Miguel 353 Diemiaszonek Robert C. 661 Dietze Hans-Joachim 637 897 Di Marco Marco 1003 Doleialova Katefina 763 Donard 0. F. X. 865 Dorado Teresa 449 859 D'Ulivo Alessandro 969 1003 Dunemann Lothar 655 Dyakov Alexey O. 703 Dybdahl Bjorn 769 Dyvik Geir 769 Ebdon Les 317 Ebihara Mitsuru 25 Efstathiou Constantinos E. 221 Ek Paul 121 Enger Jonas 539 Entwistle Andrew 395 Epifanie Arnaud 923 Evans R.Douglas 595 619 Fairman Ben 281 Fang Zhaolun 533 Fang Zheng 359 Fariiias Juan C. 51 1 Fell Gordon S. 215 Feng Liang 875 Fernandez B. 8;s Fernandez-Garcia Matilde 671 Ferron-novais M. 247 Fisher Andy 519 Florek Stefan 61 127 145 Fodor Peter 609 Fordham Peter J. 303 Fornari Roberto 433 Frech Wolfgang 711 769 Fukushima Masami 999 Furuta Naoki 25 Gaillat Ana 935 941 Gallego Mercedes 55 501 Gallignani M. 343 479 Garcia Alonso J. Ignacio 381 Gijbels Renaat 849 Golloch Alfred 161 Gomes Anne-Marie 923 G6mez Gomez M. M. 89 Goodall Phillip 3 17 Gramshaw John W. 303 Grazhulene Svetlana S. 161 Greenfield Stanley 183 Greenway Gillian M. 929 1019 Beatriz 111 321 GrCgoire D. Conrad 823 1027 Grotti Marco 325 Guern Y. 993 Guo Gang-ping 753 Guo Xiao-wei 987 Guo Xu-ming 987 GutiCrrez Ana Maria 871 Hahn Lothar 777 Halls David J.169 Hang W. 689 Haraguchi Kensaku 999 Harnly James M. 187 197 Harrison Iain 215 Harrison W. W. 689 Harville Tina R. 671 Hattingh Cornelius J. 727 Hayashi Yasuhisa 37 439 He Bin 747 Heitkernper Douglas T. 177 Held Andrea 849 Hill Steve J. 317 409 519 Hinds Michael W. 527 Hintelmann Holger 619 Hochstrasser Chantal 947 Hoffmann Volker 677 Houk R. S. 267 837 Hu Bin 455 Huang Meng-fen 31 HuldCn Stig-Goran 121 Hutton R. C. 591 Hutton Robert C. 929 Hwang Chorng-jev 3 1 Imai Shoji 37 439 Imbert Jean-Louis 93 Ingelbrecht Chris 849 Inoue Yoshinori 363 Ito Saburo 999 Ito Tetsumasa 843 Ivaska Ari 121 Jackson Kenneth W. 43 Jacquiers-Roux Dimitri 777 Jakubowski Norbert 583 Jantzen Eckard 105 Jedral Wojciech 49 Jiang Shiuh-jen 31 963 Jiang Zucheng 455 Jin Qinhan 875 Jin Qun 875 Johansson Magnus 71 1 Jones Alice V.785 Jones Phil 281 Jurcek Petr 947 Kabil Mohamed A. 733 Kawabata Katsuhiko 363 Keating Gillian E 233 Kerrich Rob 99 Khvostikov Vladimir A. 161 Kim Ha S. 335 Kim Hyo J. 335 Kinard W. F. 295 Kirschner Stefan 161 Knutsen Einar 757 Koch Lothar 381 Kopajtic Zlatan 947 Kotrebai Mihaly 505 Kotrly Stanislav 155 763 Kozma Laszlo 631 Krushevska Antoaneta 505 Lajunen Lauri H. J. 117 Lam Joseph W. H. 551,981 Lampugnani Leonardo 969 Larrea Maria T. 511 1003 Lasztity Alexandra 505 Le Bihan A. 993 Le Cor Yann 721 Ledergerber Guido 947 Lee Gae H. 335 Lee Kee B. 335 Le Garrec H. 993 Lerat Yannick 137 Liang Yan-zhong 699 Littlejohn David 215 233 241 Lobinski Ryszard 11 1 Lund Walter 405 803 809 Lunzer Florian 311 1019 L'Vov Boris V.703 Lyon Ian C. 273 Madon Lydie 923 Madrid Yolanda 321 815 Maher W. 423 Mahmood Tariq M. 43 Manzoori Jamshid L. 881 Mao X. L. 295 Marawi Isam 7 Marcus R. Kenneth 671 Martin Theodore D. 443 Martinez-Soria M. Teresa 975 Martinsen Ivar 757 Marunkov Alexander 539 Masera Eric 137 Massart D. Luc 207 Matusiewicz Henryk 981 Matveev Oleg 885 Mauchien Patrick 137 Mazzucotelli Ambrogio 325 McCartney Martin 233 McCrindle Robert I. 399 McCurdy Ed 303 McLaren James W. 371 551 McLeod C. W. 89 Mester Zoltan 609 Methven Bradley A. J. 551,981 Mile Brynmor 785 Miyazaki Akira 1 Mizuno Seiichiro 415 Moenke-Blankenburg Moens Luc 81 569 575 Monteiro Maria In& C. 329 Montoro Rosa 459 Moreda-Piiieiro Antonio 227 Moreda-Piiieiro Jorge 101 1 Muller Hans 777 Muller Victor 681 Muller-Vogt German 777 Muiioz Olivas Riansares 865 Naka Hirohito 823 Nakagawa Kohichi 999 Nakamura Susumu 467 Negretti De Bratter Virginia E.Nelms Simon M. 929 Nemet Bela 631 Ni Zhe-ming 493 699 747 Niemax Kay 563 Nishiyama Yasuko 439 Nolte Joachim 655 Novichikhin Alexander V. 703 Ogata Toshio 999 Okuhara Kyoichi 37 Olson Lisa K. 7 Omenetto Nicol6 885 Outridge Peter M. 595 253 Lieselotte 655 1011 487 Journal of Analytical Atomic Spectrometry November 1995 Vol. 10 1031Owen Linda M. W. 625 Paama Lilli 117 Pang Ho-ming 267 Parent Magali 575 Park Yang S. 335 Parry Susan J. 303 Parsons Patrick J. 521 Paschal Daniel C. 367 Pasullean Benyamin 241 Pellegrini Giovanna 969 Penninckx Wim 207 Peramaki Paavo 117 Pereiro-Garcia Rosario 31 1 Perera Indral K. 273 Perez Conde M.Concepcion Perez-corona M. Theresa 321 Perkins Charles V. 253 Petrucci Giuseppe A. 885 Petzold G. 371 Piiri Lindy 117 Pin Christian 93 Piperaki Efrosini A. 221 Pitts Les 409 519 Polzik Leonid K. 703 Powell J. J. 259 Praler Frank 677 Prange Andreas 105 Proulx Pierre 935 941 Qiao Huan-cheng 43 Qin Yong-chao 455 QuCtel C. R. 865 Quijano M. Angeles 871 Rademeyer Cornelius J. 399 Radziuk Bernard 127 197 415 Raith A. 591 Rattray Robin 829 1027 Reinicke Albrecht 487 Ren J. M. 1027 649 671 1019 871 739 739 Riondato Jorgen 569 Rivas C. 343 479 Robb Paul 625 Roberts David J. 721 Rodel Gunther 127 415 Rollin Stefan 947 Romanova Natalia 739 Rondh C. 343 Rowlands Christopher C. 785 Russo R. E. 295 Sabsabi M.643 Sadler Daran A. 253 Saito Kengo 37 Salin Eric D. 829 1027 Sanjuan Jane 287 Sanz-Medel Alfredo 149 281 Saprykin Anatolij I. 897 Sarrette Jean-Philippe 923 Saverwijns Steven 575 Schelles Wim 68 1 Schmecher Gisela R. 61 Schnurer-Patschan Christoph Schoppenthau Jorg 655 Schulze Gerhard 487 Schumann Joachim 677 Schumann Thomas 655 Sena Fabrizio 381 Seubert A. 371 Shan Xiao-Quan 791 Shemirani Farzaneh 881 Skaugset Nils Petter 739 Skogstad Asbjorn 739 Slavin Walter 521 Smeyers-Verbeke Johanna 207 Smichowski Patricia 815 Smith Clare M. M. 187 Smith Monica M. 349 Sniatecki K. 615 ,Spezia. S 555 Sramkova Jitka 763 311 649 671 1019 563 Stenz Herbert 127 Stockwell Peter B 287 317 Sturgeon Ralph E. 981 Stuwer Dietmar 583 Styris David L. 527 Sumi Masao 999 Sun Han-wen 753 Suzuki Yoshihito 363 Taddia Marco 433 Tanaka Toshiyuki 37 439 Tang Shida 521 Tanner Scott D.905 Tao Hiroaki 1 Taylor P. D. 259 Taylor Philip D. P. 395 849 Tejedor Wedleys 459 Telgheder Ursula 161 Telouk Philippe 93 Thomaidis Nikolaos S. 221 Thomas Christoph 583 Thomas P. 615 Thomassen Yngvar 739 Thompson Diana 303 Thompson K. Clive 317 Thompson R. P. H. 259 Ting Bill G. 367 Tischendorf Reinhard 61 Tomlinson Medha J. 601 853 Trassy Christian C. 661 Tsalev Dimiter L. 1003 Turner Andrew D. 721 Turner Grenville 273 Uchida Hiroshi 843 Uchino Tomonori 25 Uggerud Hilde 405 Valcarcel Miguel 55 501 Van Grieken Rene 681 689 Van Staden Jacobus F. 727 Van Straaten Mark 849 Vanhaecke Frank 81 569 Vanhoe Hans 575 Vankeerberghen Peter 207 Velez Dinoraz 459 Villeneuve Janice Y.619 W. Joseph 981 Walder Andrew J. 395 Wang Jian-shang 7 601 853 Wang Yun-zhou 359 Warren Arnold R. 267 Wei Wen-Ching 955 Weiss Zdengk 891 Wen Bei 791 Wendl Wolfgang 777 Wernli Beat 947 Wetzig Klaus 677 White Mark A. 349 Wibetoe Grethe 757 Wickstram Torild 803 Wickstram Torild 809 Willie Scott N. 981 Wilson H. Kerr 349 Woller Agnes 609 Wolnik Karen A. 177 Worsfold Paul 409 519 Wr6be1 Katarzyna 149 Xie Qian-li 99 Xu Dong-qun 753 Xu Shukun 533 Yamamoto Kouei 415 Yang Hueih-jen 963 Yang Mo-hsiung 955 Yang Peng-yuan 699 Yang Wei-min 493 Ybaiiez Nieves 459 Zamboni Roberto 969 1003 Zeiher Michael 4 15 Zeng Yun’e 455 Zhang Hanqi 875 Zhang Ke 359 Zhang Yuan-fu 359 Zoorob Grace 853 Zybin Aleksandr 563 1032 Journal of Analytical Atomic Spectrometry November 1995 Vot. 10

ISSN:0267-9477    

DOI:10.1039/JA9951001031

出版商:RSC    

年代:1995

数据来源: RSC