摘要:

JOURNAL O F ANALYTICAL ATOMIC SPECTROMETRY MAY 1994 VOL. 9 65 1 Determination of Inorganic Arsenic in Seafood Products by Microwave-assisted Distillation and Atomic Absorption Spectrometry Jose C. Lopez Carmen Reija and Rosa Montoro* lnstituto de Agroquimica y Tecnologia de Alimentos (CSIC) Jaime Roig 1 1 4601 0 Valencia Spain Maria Luisa Cervera and Miguel de la Guardia Departamento de Quimica Analitica Universidad de Valencia Or. Moliner 50 46700 Burjasot Valencia Spain A precise and accurate procedure is proposed for the determination of inorganic As in seafood products by employing microwave-assisted distillation and atomic absorption spectrometry (AAS). The effect of the amount of HCI the prior reduction of As" to As"' and the different parameters that control the distillation of AsCI were studied.Organoarsenic compounds are not decomposed by the microwave-assisted distillation procedure. The inorganic As was determined by hydride generation AAS. The methodology developed has a detection limit of 0.068 pg g-' (dry mass) or 0.023 pg p-' (wet mass) of mussel product and a relative standard deviation of 9%. The percentage recovery for As" was 97 f 3% and for As" 100 _+ 3%. The proposed procedure has been applied to the analysis of real samples of seafood products in which the total As was also determined by inductively coupled plasma atomic emission spectrometry. Keywords Inorganic arsenic determination; microwave-assisted distillation; atomic absorption spec- trometry; inductively coupled plasma atomic emission spectrometry The various organic and inorganic species of As differ widely in their degree of toxicity inorganic forms being in general more toxic than organic ones.' For this reason the Food and Agriculture Organization-World Health Organization (FAO- WHO) Mixed Commission of the Codex Alimentarius rec- ommends establishing not only the total concentration of As present in foodstuffs but also the concentrations of those chemical species of which this element forms a part.2 Seafood generally presents a much higher As content than other foods so that the total amount of As ingested by humans depends on the amount of seafood included in their diet.In this context it is appropriate to mention that Australian legislation in its advanced recent Model Food Act while setting a maximum permitted level of As in fish crustaceans and molluscs applies this level to inorganic As only and does not take into account levels of total AS.^ This approach recognizes the toxicity of inorganic forms of As and could serve as an example for other nations.For determining inorganic As in biological materials in either oxidation state III or v or as the sum of both states three stages must be considered (i) a releasing stage; (ii) a separation of inorganic forms from organic As; and (iii) measurement of the As. In the literature several alternatives for each stage can be found which include a series of extractions and separations developed in different steps dramatically increasing the complexity of the process especially when As is determined in solid samples in which case a series of successive extraction and purification steps must be undertaken.'-* Another option for the determination of inorganic As by carrying out the determination of As"' and AsV species suggests the use of chromatographic alternatives linked to sensitive spectrometric methods.This involves the use of high cost instrumentation and also increases the analysis time.' The most promising method combining release and separ- ation of inorganic As in one stage at low cost and with ease of operation is that proposed in 1973 by Lunde." This procedure is based on conventional distillation of inorganic As as arsenic(rI1) chloride in the presence of concentrated hydrochloric acid. Subsequent determination of As was per- formed by neutron activation and X-ray fluorescence.Nevertheless this distillation procedure is time consuming and * To whom correspondence should be addressed. the non-degradability of organoarsenic compounds under the conditions proposed in the method was not stated. The above mentioned procedure has been applied by other workers to the determination of As by hydride generation atomic absorp- tion spectrometry (HG-AAS) having found that the distillation of AsCI is efficient and economic as compared with individual selective extraction procedures.6 In the present paper Lunde's method" has been modified by reducing the distillation process improving the reduction of AsV to AS"' confirming good recovery of As"' and AsV and proving the non-degradability of organoarsenic compounds. In order to update the conventional distillation method pro- posed by Lunde a microwave oven was then used to provide the appropriate temperature for the distillation of AsC1,.The effect of the amount of HC1 the prior reduction of AsV to As"' and the different parameters that control the distillation of AsC1 (microwave power supplied position of the reactor inside the oven cavity and time and experimental conditions required for the distillation and hydrolysis of AsCl,) were studied. Inorganic As was determined by HG-AAS and total As by inductively coupled plasma atomic emission spec- trometry (ICP-AES). The detection limit precision and accuracy for the micro- wave method were evaluated. Organoarsenic compounds are not decomposed and do not interfere in the determination of inorganic arsenic compounds.The method has been applied to the analysis of seafood products. Experimental Equipment A Perkin-Elmer 5000 atomic absorption spectrometer equipped with an MHS-10 hydride generation attachment and an 056 stripchart recorder was employed to carry out the determination of inorganic As by HG-AAS. A Perkin-Elmer 6000 ICP atomic emission spectrometer equipped with an ICP source (PlasmaTherm 2.5 kW 27.13 MHz) controlled by a Perkin-Elmer 7500 laboratory computer and a Perkin-Elmer PR-2 10 printer was employed to carry out the determination of total As A Heraeus muffle furnace Model 1100/3 controlled by a digital microprocessor (Jumo Model DPG-44/1) was652 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MAY 1994 VOL. employed for the dry ashing of seafood products in order to determine the total As content.An FTS systems lyophilyser equipped with a microprocessor controlled tray dryer chamber connected to an Epson Equity I + computer was employed. A distillation apparatus consisting of a 100ml round- bottomed distillation flask a 100 ml condenser and a 20 cm long refrigerant coil was used for the selective separation of inorganic As as AsCl,. Alternatively this procedure was carried out by following a microwave-assisted procedure using a Bauknecht domestic microwave oven Model MWT-732 pro- grammable for time and microwave power in various discrete steps with eight power settings (ranging from 120 to 750 W) and a maximum time of 99 min 59 s. The microwave oven was used without further modification and the generation and distillation of AsC1 was carried out in poly( tetrafluoro- ethylene) (PTFE) vessels of 120 ml internal volume and 10 mm wall thickness designed in this laboratory.The screw cap of these vessels includes a perforated PTFE disk located in the upper part of the vessel which serves to avoid sample ejection during distillation. One or two holes made in the cap of the vessel provide a means of introducing a carrier gas flow and for the evacuation of AsCl by using 1 mm i d . PTFE tubes. These tubes are introduced into the oven by means of the vent holes of the oven. The AsCl vapours are taken in one or two 250 ml gas collection tubes. Figs. 1 and 2 show the distillation manifolds employed for both thermal convection and microwave assisted procedures.Reagents Analytical-reagent grade water with a metered resistivity of 18 MQ cm was employed to prepare samples and standards. All reagents used were of the highest purity available and at least of analytical-reagent grade. A stock solution of As"' was prepared by dissolving 1.320 g i/ x; Fig. 1 Distillation apparatus employed for conventional distillation A round-bottomed distillation flask; B condenser; C long refrigerant coil; and D thermometer l+7i I / 9 Fig. 2 Distillation microwave assisted procedure A microwave oven; B additional load; C PTFE vessel; D collector gas tube; and E PTFE tube of arsenic(u1) oxide (As20,) (Riedel de Haen Hannover Germany) in 25 ml of 20% (m/v) KOH solution and neutraliz- ing with 20% (v/v) H2S04 and then diluting to 11 with 1% (v/v) H2S04.A stock solution of AsV was prepared by dilution of the corresponding Titrisol standard of 1000 ppm (Merck Darmstadt Germany) and arsenobetaine was supplied in an aqueous solution of 1000 ppm by the Service Central d'Analyse du CNRS (SCA) Vernaisson France. For the dry ashing of samples an ashing aid suspension was prepared by stirring 20 g of Mg(NO3),.6H2O (98%) and 2 g of MgO (90%) (Panreac Montplet & Esteban Montcada i Reixac Barcelona Spain) in 100 ml of water. Various solutions of KI (Panreac) were prepared freshly from the solid product and stored in a refrigerator. The HCl solutions were prepared from 12 moll-' concentrated acid (p= 1.19 g ml-l) (Pancreac) and diluted with H20. A 2.5% m/v aqueous solution of hydroxylamine hydro- chloride was prepared.Sodium tetrahydroborate (Probus Badalona Barcelona Spain) was prepared by dissolving 3 g of NaBH in 100 ml of 1% m/v NaOH solution filtered through Whatman No. 42 paper and stored in a refrigerator. Samples Several canned or frozen seafood products were purchased from local retail market outlets for the determination of As. The brine or sauce was removed by the method used for determining the drained mass in canned foods. The total contents of each package were emptied into a sieve with a 5 mm stainless-steel mesh made of 1 mm gauge wire the sieve being tilted slightly to facilitate drainage. To ensure total drainage the seafood products were allowed to drain for 5 min at ambient temperature afterwards the samples that had been processed in an oil sauce were pressed between two sheets of filter-paper.The oily residue can produce lyophilysed samples that tend to cake. The drained products were frozen at - 20 "C and then freeze-dried for 20h at a chamber pressure of 0.225 Torr ( 1 Torr = 133.322 Pa). Sublimation heat was sup- plied by conduction from heating plates at 20°C. The lyophi- lysed samples were crushed and homogenized to a fine powder in a Henkel mill which was refrigerated by water. The resulting powder was stored in previously decontaminated twist-off flasks and kept in the freezer until required for the analysis. General Procedures Dry ashing of samples for the determination of total As As described in an earlier report,'' the procedure employed was the following. Weigh 10.00&0.01 g of wet sample in a 250 ml beaker.Add 10 ml of the ashing aid suspension contain- ing 20% m/v Mg(N03)2 2% m/v MgO and 5 ml of 40% v/v HNO and mix well. After evaporation on a sand-bath until total dryness has been achieved ash in a muffle furnace at a temperature of not more than 450 "C 150 "C for 1 h; 200 "C for 30 min; 250 "C for 1 h; 300 "C for 3 h; 350°C for 30 min; and 450 "C for 12-14 h. In general it is necessary to wet the ash with HN03 (10% v/v) evaporate on a sand-bath and thenJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MAY 1994 VOL. 9 653 perform a second shorter ashing process (150 "C for 1 h; 300 "C for 30 min; and 450°C for 12-14 h) repeated once or twice until the ash is completely white. The 100°C temperature increases are carried out over a period of 15 min and increases of 150°C over 30 min.Dissolve the white ash in 5 ml of 6 moll-' HC1 and dilute to 25 ml with water. Prepare reagent blanks by applying the dry ashing procedure to 10ml of the ashing aid suspension and 5 ml of HNO (40% v/v). Conventional distillation of total inorganic As as AsCl Weigh 2.00f0.01 g of the lyophilysed and homogenized dry product and introduce it into a distillation flask. Add 25 ml of 9.2 moll-' HCl and 10 ml of KI (30% m/v) to the sample and leave to stand for 5 min. Connect the distillation apparatus shown in Fig. 1 and heat until a vapour temperature of between 100 and 130°C is reached. Distil a volume of approximately 15 ml and then store in a 100 ml calibrated flask. Repeat the distillation process once after addition of another 25 ml of 9.2 moll-' HC1 and 10 ml of KI (30% m/v) to the remaining solution.Take 15 ml from the second distillation and mix with the previous distilled fraction. After cooling the condenser and receiver were washed with distilled water and the washings added to the distillate and made up to volume after filtration through a Whatman No. 1 paper. This filtration step is carried out in order to avoid the presence of residues of I which have frequently been obtained in several experiments. Microwave-assisted distillation Weigh l.OOfO.O1 g of dry sample and introduce it into a PTFE vessel. Add 25ml of 6.6mol1-' HCl and 5 ml of KI (30% m/v) to the sample close the reaction vessel and place it inside the microwave oven. Irradiate the sample at the maximum power (750 W) for 4 min using an additional load in the oven of 25 ml of water; then allow the sample to stand replace the load in the oven by another 25 ml volume of cool distilled water and irradiate for an additional 3 rnin at 750 W.Under the above mentioned conditions the AsCl is distilled and hydrolysed in a flask containing 45 ml of water and 5 ml of a 25% m/v hydroxylamine hydrochloride. This solution is then filtered through a Whatman No. 1 paper and diluted to 100 ml in order to avoid the presence of I2 residues. HG-AAS determination of inorganic As The instrumental parameters are indicated in Table 1 and have been described in an earlier paper.' Pipette triplicate volumes Table 1 Instrumental parameters employed for the determination of As by HG-AAS and ICP-AES of 1 ml of the previously distilled sample solutions containing < 60 ng of As into hydride reaction flasks and add 4 ml of 62.5% v/v HC1.Then add 1 ml of 2% m/v KI solution and allow the mixture to stand for 5min. Purge the system with argon. Depress the plunger in order to deliver NaBH solution into the reaction flask. Release the plunger after the peak maximum on the recorder has been reached (from 10 to 15 s). Measure the peak height of the signal. Construct a calibration graph obtained from 0 10 20 40 and 60 pl of a 1 pg ml-' working standard solution of As"' mixed with 5 ml of 6 moll-' HC1. Add 1 ml of 2% KI and measure in the same way as the samples. ICP-AES determination of total As Measure the As using the selected analytical parameters indicated in Table 1.Calibrate the spectrometer with a 10 pg ml-' As standard solution in HC1 (10Y0 v/v) using a solution of HCl (10% v/v) as a blank. Check the calibration with standard solutions of 0.5 2 and 5 pg ml-' of As. Results and Discussion Distillation of AsCI Based on the distillation procedure proposed by Lunde,lo a series of modifications were carried out in order to improve the reduction of AsV to AS''' using KI and to accelerate the distillation of AsC1,. A dry sample mass of 2 g and a volume of HCl of 25ml were selected. The distillation was carried out in three sequen- tial steps in order to ensure the total recovery of inorganic As. To reduce AsV to As"' a series of reductants has been proposed in the literature such as metallic iron ascorbic acid sodium thiosulfate sodium or potassium iodide or a combi- nation of two of In the present study 10 ml of a KI solution (30% m/v) was employed for the reduction of AsV by using a preliminary reduction step of 5 min at ambient temperature before distillation.The temperature at which AsCl distils varies from 108 to 130°C and under these conditions it was confirmed that a distillation volume of 15 ml in each of the distillation steps produces quantitative recovery of inorganic As. Hence under the above conditions the time required for the total distillation of AsC1 in each step is reduced to 12-18 min as compared with the 40 min required by Lunde's procedure. The determination of As by HG-AAS in each of the three successive distillations demonstrated that the instrumental reading obtained for the third distillation is of similar order to that of the blank obtained for the reagents used and so it is confirmed that the distillation of AsCl is quantitative in only two distillation steps. ~ HG-AAS- Lamp power/W Wavelength/nm Slit-width (low)/nm Acetylene flow rate/l min-' Air flow rate/l rnin - Argon flow rate/ml min-' Recorder voltage/mV Chart speed/mm min- ICP-AES- Wavelength/nm Viewing height/mm Outer Ar flow rate/l min-' Inner Ar pressure/psi* Inner Ar flow rate/l min-' R.f.power/kW Sample solution flow rate/ml min-' Integration time/s 8.5 193.7 0.7 2 15.5 450 10 5 193.7 10 18 20 0.4 1.1 1.1 5 * 1 psi = 6894.76 Pa. Recovery of inorganic As by distillation of AsC1 In order to demonstrate the applicability of the developed procedure a series of experiments to study the recovery of added amounts of As"' and AsV was carried out.Con- centrations of As in both forms considered at a level of 1 pg g-' were added to previously analysed mussel samples containing of the order of 0.33 pg g-' wet mass of inorganic As. The percentage recovery of As"' was 97 & 3% and that of AsV 100&3% employing 10 ml of 30% m/v KI solution as the reducing agent. These results are better than those obtained by the same procedure 96&2% for As"' and 73+3% for AsV but using 10 ml of a 20% m/v solution of KI. Additional experiments carried out by adding arsenobetaine concentrations of the order of 10 pg g-' indicated that under the chosen conditions this compound remains undestroyed during the distillation of AsCl and so confirms that the inorganic As is quantitatively and selectively isolated and can be determined by means of the proposed procedure.654 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MAY 1994 VOL.9 Microwave-assisted Distillation of AsCl In order to simplify the distillation procedure for the determi- nation of inorganic As compounds and to reduce the analysis time required and the amount of sample handling the distil- lation of AsCl was studied using a microwave oven as a thermal source. In this way the effect of the microwave-oven parameters and that of the chemical parameters on the recovery of inorganic As was investigated. Microwave-oven parameters To accelerate the distillation of AsCl it is necessary to provide a temperature of the order of 130°C in the shortest possible time and hence the maximum power of the microwave oven (750 W) was employed.Taking into account the fact that in the oven cavity the distribution of the microwave radiation is not homogeneous the position at which the sample absorbs the highest amount of microwave radiation was selected in order to achieve a fast distillation procedure. An additional oven load consisting of a small volume of water was placed into the oven for safety reasons in order to avoid the possibility of the magnetron working at maximum power in the absence of any absorbent material. In this study the additional oven load was water in a beaker located inside the oven in the least hot position. Several experiments were carried out using different volumes of water.The best results were found with 25 ml of cool distilled water which was changed after each of the different irradiation steps had been employed. In all cases a multi-step programme was selected for irradiation of the samples including non-irradiation intermedi- ate steps of 1 min. These intermediate steps were employed in order to change the additional oven load and as they are also required to allow the remaining solution to condense and to be in a liquid form for the following irradiation step. The experiments carried out using different irradiation time programmes and different additional water load volumes are shown in Table 2. As can be seen the '4+3' programme including two irradiation steps of 4 and 3 min and an intermedi- ate non-irradiation step of 1 min with a low oven load yields the smallest residual volume of sample after distillation.This ensures that all inorganic As has distilled. To carry out the microwave-assisted distillation of AsCl two different types of reaction vessels were employed. Both are modified designs of those previously developed by the present workers for the pressurized wet ashing of mussel samples.20 However in this instance one or two holes were added to facilitate the distillation of AsCl and alternatively the use of a carrier gas stream as has been commented on under Experimental. Chemical parameters From experience with earlier experiments on the conventional distillation of AsCl the sample size and the concentrations of reagents were selected. Table 2 Effect of irradiation time and additional load of the micro- wave oven on the volume of sample remaining after the distillation of AlC1,.The irradiation programme includes several irradiation steps (time given in min) followed by non-irradiation steps of 1 min; all experiments were carried out by using an air carrier gas flow ~~ Irradiation programme/min 3+3+2 3+3+2 5+3 5+2 3 f 3 4 4+3 Load/ml 150 100 100 50 50 50 50 Residue/ml 20 9 13 14 14 13 3 Experiments were carried out by taking 1 g of sample adding 25 ml of 6.6 mol I-' HCl and 5 ml of KI solution as a reductant at two concentrations 20 and 30% m/v. To optimize the different parameters that control the distillation of AsC1 (microwave power supplied position of the reaction vessel in the oven cavity additional load in the oven and time and experimental conditions required for the distillation and hydrolysis of AsCl,) 20% KI solution was employed.To optimize the recovery of AsV 30% KI solution was used. To collect the AsCl three different solutions were assayed in order to provide a fast hydrolysis of AsC1 and to avoid losses of the gaseous products (i) pure water (ii) 3.3 mol IF1 NaOH solution which neutralizes the HCl coming from the reaction vessel and (iii) 2.5% m/v hydroxylamine hydro- chloride which can reduce the volatile organic matter that is distilled. From the three systems assayed the use of NaOH solution was shown to be unnecessary because a neutral pH solution is sufficient to avoid losses of AsC1,. However the presence of a reducing system such as hydroxylamine hydrochloride is necessary to reduce the volatile organic matter which can pass into the distillate and causes problems in the hydride gener- ation step.This produces badly defined peaks of H,As as can be seen in Fig. 3 where well defined peaks obtained in the presence of hydroxylamine hydrochloride are also shown. Development of the microwave-assisted procedure The results obtained for the recovery of As"' and AsV for different microwave-assisted distillations are shown in Table 3. The main differences in the distillations consists in the use of an air carrier flow differences in the KI concentration the use of 50 or 25 ml of water as additional load and the use of different condenser systems. From experiments carried out using an air carrier flow it can be concluded that only one condenser flask is necessary for the total hydrolysis of AsCl and that the use of NaOH solution is unnecessary.On the other hand the use of two condenser flasks hinders the distillation of AsCl and strongly reduces the recovery of As"'. The carrier gas reduces the temperature in the reaction vessel and thus provides a high residual volume and poor recovery of both As"' and AsV. To improve the recovery of AsV 5ml of 30% m/v KI solution must be added to samples instead of using a 20% m/v concentration of the reductant. The use of hydroxylamine hydrochloride favours the reduction of the organic matter which could distil with the A s C ~ ~ ' ~ ~ and could also act as a reductant for any possible AsV in the condenser flask formed from the oxidation of As"'. So the use of this reagent provides good results in the hydride generation step.By using the recommended procedure for the microwave- assisted distillation of AsCl it has been confirmed that the additional load in the oven boils after 1 min 10 s of irradiation Time - Fig.3 Effect of the addition of hydroxylamine hydrochloride on typical recorder tracings for As. A Sample without hydroxylamine hydrochloride; B the same sample after the addition of hydroxylamine hydrochlorideJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MAY 1994 VOL. 9 655 Table 3 Results obtained by microwave assisted distillation of AsCl,. In all cases the irradiation programme was 4 min of irradiation at 750 W 1 rnin without irradiation and 3 rnin of irradiation (750 W); 25 ml of 6.6 moll-' HCl were added to 1 g of sample and 5 ml of KI solution with different concentrations of KJ ~ ~~~ Distillation conditions Recovery (%) Carrier Load/ Residue/ Experiment gas KI ml Condenser system As" As' ml 1 Air 5 ml 20% m/v 50 2 Flasks with 50 ml water 46 53 8-12 3 None 5 ml20Y0 m/v 25 1 Flask with 50ml water 95 81 0.6-2.8 0.4-3.6 4 None 5 ml 30% m/v 25 1 Flask with 45 ml water 94-100 97-103 2 Air 5 ml 20% m/v 50 1 Flask with 50 ml NaOH 3.3 mol I-' 70 65 8-10 + 5 ml 25% m/v hydroxylamine hydrochloride and the distillation begins at 1 min 30 s.After the intermediate non-irradiation step the distillation takes place continuously when the sample is irradiated. The microwave-assisted distillation can be carried out in just one distillation step leading to a large reduction in the sample preparation time (from the more than 1 h that is required for conventional distillation using two distillation steps to less than 10min) and also reduces the amount of sample handling necessary.It was verified that breaking of organo-arsenic bonds was not achieved during the microwave-assisted distillation step. Additional experiments carried out by adding arsenobetaine concentrations of the order of 1Opgg-' gave evidence that this compound is not destroyed during the distillation of AsCl and the As content in the distillate was similar to that obtained in unspiked samples. Analytical Features of the Developed Procedure The method developed for the determination of inorganic As by HG-AAS is based on a specific sample preparation pro- cedure which ensures the quantitative recovery of both As"' and AsV and that avoids the destruction of organic arsenic compounds such as arsenobetaine which is not recovered at all when added to samples before distillation.The analytical characteristics such as detection limit precision and accuracy were evaluated in a mussel sample prepared as described under Experimental. The analytical features of the method are those correspond- ing to the determination of As by HG-AAS and a typical calibration curve y = 0.065 + O.O094x with a regression coefficient of 0.999 was obtained for an As concentration range of from 10 to 60 ng. The detection limit (0.068 pg g-' dry mass or 0.023 pg 8-l wet mass) was established as the concentration of As in a mussel product which provides an absorbance reading that is statistically different from that of the blank. This was calculated by dividing three times the standard deviation of the absorbance readings of six reagent blanks by the sensitivity and taking into account the sample mass and dilution used.The precision of the method (9"/0) was evaluated by analys- ing six sub-samples of the same sample of mussel and is expressed as the relative standard deviation (RSD). The per- centage recovery was evaluated by spiking six samples with 1 pg 8-l of As"' and 1 pg 8-l of AsV. As previously indicated in Table 3 under the recommended conditions the percentage recovery of the method for As"' is 97f3% and that for AsV is 10023%. Determination of Inorganic As in Real Seafood Samples Six samples of canned seafood and three different types of frozen fish were analysed using the recommended procedure.The total As content was determined by ICP-AES after dry ashing and the inorganic As was determined after microwave- assisted distillation by HG-AAS. The results found are summa- Table 4 Determination of inorganic and total arsenic in real samples Arsenic content/pg g- ' Sample Anchovies Squid Tuna Cockles Sardines Prawns Hake Mussels Sole Dry mass Inorganic Total 0.30 6.51 0.10 5.24 < L.D. 1.93 1.34 16.84 0.15 4.10 < L.D. 2.3 1 < L.D. 6.98 0.95 8.53 <L.D. 38.10 As* Ast Wet mass Inorganic As* 0.10 0.03 < L.D. 0.32 0.07 < L.D. < L.D. 0.33 < L.D. Total As? 2.12 1.75 0.82 4.01 1.77 0.34 1.61 2.95 7.76 Inorganic AS (Yo)$ 5 2 8 4 - - - 11 - ~ ~~ * Inorganic As was determined by HGAAS after microwave-assisted t Total As was determined by ICP-AES after dry ashing.$ Inorganic arsenic expressed as a percentage of the total arsenic. distillation. rized in Table 4. The levels of total inorganic As found did not exceed 0.3 pg 8-l wet mass. Expressed as a percentage of the total As this content varies from less than 2 to 11Y0. These results are of the same order as those reported by Brooke and Evans who found that inorganic As in fish varies from 1 to 5% of the total As content. Conclusions The proposed method enables an accurate determination of inorganic As in seafood to be made using microwave-assisted distillation and AAS. Under the conditions established the organoarsenic compounds are not decomposed. Microwave- assisted distillation provides a significant reduction in the sample preparation time compared with that required for conventional distillation and also reduces the amount of sample handling.The results obtained for the analysis of real samples of seafood products show the suitability of the proposed method for this type of determination. Funds to carry out this work were provided by the Comision Interministerial de Ciencia y Tecnologia (CICyT) Project ALI89-0521 for which we are very grateful. We also thank the Service Central d'Analyse du CNRS (SCA) of Vernaisson (France) for providing arsenobetaine sample. References 1 Concon J. M. Food Toxicology Part B Contaminants and Additives Marcel Dekker New York 1988 p. 1083. 2 FAO/WHO Lista de Dosis Mhximas de Contaminantes Recomendadas por la Comisibn Mixta FAOIOMS Programa conjunto FAO/OMS sobre normas alimentarias CAC/FAL-2-1973 FAO Rome 1973.656 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MAY 1994 VOL.9 3 4 5 6 7 8 9 10 11 12 13 14 Ministry of Agriculture Fisheries and Food Food Surveillance Paper No. 8 HM Stationery Office London 1982. Reilly C. Metal Contamination of Food Elsevier London 2nd edn. 1991 p. 79. Yasui A. Tsutsumi C. and Toda S. Agric. Biol. Chem. 1978 42 2139. Brooke P. J. and Evans W. H. Analyst 1981 106 514. Munz H. and Lorenzen W. Fresenius’ 2. Anal. Chem. 1984 319 395. Holak W. and Specchio J. J. At. Spectrosc. 1991 12 105. Tye C. T. Haswell S. J. ONeill P. and Bancroft K. Anal. Chim. Acta 1985 169 165. Lunde G. J. Sci. Food Agric. 1973 24 1021. Cervera M. L. Navarro A. Montoro R. and Gbmez J. Fresenius’ J. Anal. Chem. 1993 347 58. Cervera M. L. Navarro A. Montoro R. and Catala R. J. Assoc. Off. Anal. Chem. 1989 72 282. Fiorino J. A. Jones J. W. and Capar S . G. Anal. Chem. 1976 48 120. Hu C. T. Yu S. H. and Wei Y. T. in Proceedings of the Fifth International Congress of Food Science & Technology Kyoto 15 16 17 18 19 20 21 22 Japan 1978 eds. Chiba H. Fujimaki M. Iwai K. Mitsuda H. and Morita Y. Elsevier Amsterdam 1979. Inhat M. and Miller H. J. J. Assoc. Off. Anal. Chem. 1977 60 813. Rees D. I. J . Assoc. Public Anal. 1978 16 71. Holak W. J. Assoc. Off. Anal. Chem. 1980 63 485. Hon P. K. Lau 0. W. Cheung W. C. and Wong M. C. Anal. Chim. Acta 1980 115 355. Tam G. K. M. and Lacroix G. J. Assoc. Off. Anal. Chem. 1982 65 647. Ybaiiez N. Cervera M. L. Montoro R. and de la Guardia M. J. Anal. At. Spectrom. 1991 6 379. Carrillo F. Bonilla M. and Camara C. Microchem. J. 1986 33 2. Cervera M. L. Lopez J. C. and Montoro R. J. Dairy Res. 1994 61 83. Paper 31039306 Received July 6 1993 Accepted November 20 1993

ISSN:0267-9477    

DOI:10.1039/JA9940900651

出版商:RSC    

年代:1994

数据来源: RSC

摘要:

JOURNAL O F ANALYTICAL ATOMIC SPECTROMETRY MAY 1994 VOL. 9 657 Determination of Selenium in Fruit Juices by Flow Injection Electrothermal Atomization Atomic Absorption Spectrometry Marco A. 2. Arruda Mercedes Gallego and Miguel Valcarcel* Department of Analytical Chemistry Faculty of Sciences University of Cordoba E- 74004 Cordoba Spain An automatic flow injection method for the determination of Se in fruit juices by electrothermal atomization atomic absorption spectrometry is proposed. The flow manifold used which is coupled to the autosampler cup enables dilution of the slurry addition of a chemical modifier and the determination of Se in the filtered liquid phase. The effects of various chemical modifiers furnace temperatures and major foreign ions present in fruits were investigated.Selenium was successfully determined in fruit and tomato juice by using the standard additions method (juice samples were diluted 10-fold) with ,a relative standard deviation of 4.5-12.0%. The limit of detection (3sJ of the proposed method is 5 pg I- . Keywords Selenium determination; fruit juice analysis; electrothermal atomization; atomic absorption spectrometry; slurry atomization; flow injection; continuous filtration Selenium is both essential and toxic to man and animals depending on the concentration at which it is supplied.' Selenium deficiencies are recognized as being responsible for cardiomyopathy muscular dystrophy and reproductive dis- orders in a variety of animal species.2 Concentrations of Se below 30 ng g-' in feed lead to severe deficiency in livestock whereas concentrations between 0.1 and 0.4 pg g-' are con- sidered to be optimal3 and can be achieved in most instances by the addition of Se compounds to the feed.Selenium requirements in man are unknown. The concentrations of Se in foods are dependent on the soil conditions the type of food concerned and the preparation method used. Suitable sources of Se include kidney liver other meat products seafood and whole grains. Fruits and vegetables are poor sources of Se since much of the Se content in the latter can be lost in the cooking water. The effect of processing techniques on the final Se content in prepared meat and fish is poorly documen ted.4 The significance of Se is apparent from the host of available methods for its determination.'s6 Various analytical techniques including neutron activation analy~is,~ gas cromatography- mass spectrometry atomic fluorescence spectrometry' and differential pulse voltammetry," have been used for this pur- pose.One other widely used technique for the determination of Se is hydride generation atomic absorption spectrometry," which however has some drawbacks such as the need for sample digestion prior to analysis the adverse effect of compet- ing species that can mask or suppress the evolution of SeH and physical problems (e.g. leakage) associated with the use of pressurized gas streams. Recently several workers have shown the usefulness of electrothermal atomic absorption spectrometry (ETAAS) for the determination of Se.6912-15 However some problems arising from matrix interferences and volatilization of Se during ashing and atomization have been reported.'2*1c18 Se veral chemical modifiers such as pal- l a d i ~ m ~ ' ~ ~ palladium plus magnesium," n i ~ k e l ~ ~ .' ~ platinum plus magnesium,13 and palladium plus ascorbic acid2' have been used to circumvent analyte volatilization problems and vapour-phase interferences. Gaseous monohalides from the analytes are the most frequently detected chemical interferences in this context while phosphorus molecules cause the most severe spectral interference^.'^*'"'^,^^ Although flow injection (FI) has rarely been used in conjunc- tion with ETAAS it enables automatic implementation of such operations as sample dilution reagent addition and homogen- ization in a closed system thereby minimizing contamination.In this context several worker^^,-^^ have developed on-line FI ~~ ~~ * To whom correspondence should be addressed. preconcentration systems that are directly fitted to the capillary of the arm of the autosampler. These systems avoid the need to collect samples in exposed vessels thereby diminishing contamination problems and no chemical modifier is required since only the analyte once isolated from the matrix is injected into the furnace. Sample preparation methods based on acid decompo~ition~~ and the use of ~ l u r r i e s ~ ~ . ~ ~ have been employed for the determi- nation of Se in environmental and biological materials by ETAAS. Direct introduction of solids or slurries in food analysis is a good choice that avoids losses of Se during digestion.By using this procedure Se has been determined in bovine and calf liver,27 corn bran and egg powder,,* milk and milk powder,29 rice flour3' and However no reference to the determination of Se in fruit juices or fruits by this procedure has been reported so far. The purpose of this work was to minimize sample handling reduce the analysis time and particularly automate slurry homogenization by using a mixing chamber. The FI system used to this end enables sample dilution addition of a chemical modifier and acquisition of information on the analyte distri- bution in the solid and liquid phase of the slurry. Experiment a1 Apparatus A Perkin-Elmer 1100-B atomic absorption spectrometer (Uberlingen Germany) equipped with a HGA-700 graphite furnace and an AS-70 autosampler was used. A Perkin-Elmer selenium hollow cathode lamp operated at 16 mA and pyrolytic graphite tubes furnished with L'vov platforms (Perkin-Elmer Madrid Spain) were also used.Selenium atomic absorption was measured at 196.0 nm by using a 2.0 nm spectral band- width; atomization signals (in the peak area mode) were printed on a EPSON FX-850 printer (Wembley UK). The optimum instrumental conditions are given in Table 1. Argon as the inert gas and deuterium-arc background correction were employed throughout. The FI system comprised a Gilson- Table 1 10 pl of sample + 10 pl of modifier for conventional determination Furnace programme for determination of Se; injected volume ~ Ar flow rate/ Step TemperaturerC Ramp/s Hold/s ml min-' Read 1 100 15 30 300 2 140 15 30 300 3 900 5 20 300 4 1900 0 4 0 - 1 5 2650 1 1 300658 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MAY 1994 VOL.9 minipuls-2 peristaltic pump (Villiers-Le-Bel France) fitted with poly(viny1 chloride) tubing; a laboratory-made three-piece injector commutator32 with built-in T-shaped perspex connec- tors and poly( tetrafluoroethylene) (PTFE) transmission lines of 0.7 mm id.; an Omnifit 3303 PTFE filter (modified by using circular channels on both ends in order to increase the chamber inner volume up to 100 p1 and the filtration area to ~3 cm') furnished with a paper disk (Watman No. 1); and a customized PTFE mixing chamber of 1 ml in volume that included a PTFE covered magnetic stirring bar for efficient mixing and sample dilution.A Heidolph magnetic stirrer and an ultrasonic bath (Bandelin Tk 52 Berlin Germany) for homogenizing samples were also used. Reagents All reagents used were of analytical-reagent grade. High-purity water (Milli-Q Water System Millipore Madrid Spain) was employed throughout. A lOOOmgl-' Se stock solution was prepared by dissolving 1.000 g of Se metal in a minimum volume of concentrated HNO 14.4 moll-' and evaporating to dryness then 2 ml of H20 were added and the mixture was again evaporated to dryness finally the residue was dissolved in 10% v/v HC1 and diluted to 1 1 with 10% v/v HC1. Standard solutions containing 50.0-500.0 pg 1-l of Se were prepared from this stock solution by serial dilution with 0.2% v/v HN03 prior to use. These standard solutions were diluted 10-fold in the FI manifolds.Solutions in 0.2% v/v HNO containing 6000mgl-' K+ as KNO 270mgl-' Na+ as NaCl 420 mg 1-' Ca2+ as Ca(N03)2 40 mg 1-' Fe3+ as Fe(NO,) 1020 mg 1-' P as K2HP04 450 mg 1-' S as CuSO4.5H20 and 500 mg I-' C1- as NaC1 were also prepared to check for potential interferences. Several chemical modifiers including Ni(N03)2 Mg(N03)2 and Pd(N03)2 dissolved in H20 were used at concentrations between and lO-'mol 1-'. All reagents were purchased from Merck (Darmstad Germany). Cleaning and Storage Material Vessels of PTFE were cleaned by soaking in 10% v/v HNO for 48 h rinsing five times with Milli-Q water and filling with ultrapure water until use. The juice samples held in Tetrabrik packages were refrigerated (4 "C) after opening. Juice Sample Preparation As a preparatory step juice samples (both as received or spiked with Se) were homogenized in an ultrasonic bath for 10 min.Two different sample preparation procedures were used. (i) Wet ashing in which a volume of 5 ml of fruit juice in a platinum crucible was mixed with 6ml of concentrated HNO and evaporated to dryness in a sand-bath. After evaporation a fresh 6ml portion of HNO was added followed by evapor- ation to almost dryness two more times. Finally the resulting residue was dissolved to 10 ml with 0.2% v/v HNO,. A reagent blank was also prepared in parallel. (ii) Direct slurry measure- ments on juice samples were made by simply diluting 1 ml of juice to 10 ml with 0.2% v/v HNO after spiking with Se. The diluted samples were placed in autosampler cups for the measurement of Se.Conventional Determination of Selenium The calibration curve was run by sequentially introducing 10 p1 of standard solutions containing 5 10 20 30 40 and 50 pgl-' of Se plus 10 pl of 1 x 10-2mo11-' Pd(N03)2 as chemical modifier into the graphite furnace. Juice samples were analysed by using the standard addition methods; for this purpose each diluted juice (1 +9) was spiked with 5 10 15 20 or 3Opgl-' of Se. Blanks containing lop1 of 0.2% v/v HNO and lop1 of chemical modifier were used and a standard containing 30 pg 1-' of Se (for re-sloping) were run every five samples. The temperature programme used is shown in Table 1. All measurements were made in triplicate. Flow Injection Manifold The manifold and operational sequence used are depicted in Fig.1. In the first step 200 pl (Lsl) of juice fruit slurry (spiked with 5-30 pg 1-' of Se) were driven by the carrier solution (0.2% v/v HNO,) and merged with the chemical modifier CO.01 moll-' Pd(NO3),] at the confluence point (X); the mixed solution was continuously filtered and the filtrate trans- ferred to the mixing chamber for dilution and homogenization of the slurry. After mixing the filtrate solution was loaded into the autosampler cup for 120 s (2 ml). The flow rate ratio of carrier to chemical modifier (1 +9) allowed for the extent of dilution required for the determination of selenium. During t IC UB Fig. 1 FI manifold and operational sequence used for the determination of Se in juice liquid phase. L and L sample loops (200 pl); M chemical modifier; IC injector commutator; MC mixing chamber; UB ultrasonic bath; P peristaltic pump; W waste.Flow rates of carrier matrix modifier and washing solution (0.2% HNO,) 0.5 0.5 and 1.0 ml min-' respectivelyJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MAY 1994 VOL. 9 659 the filtration step the carrier washed the residue retained on the filter and the washings were collected in the autosampler cup; the fruit juice was aspirated through sampling loop Ls2 for further measurments. In this step only the Se content in the liquid phase of the slurry was determined. By switching the commutator the 200 pl slurry sample plug ( Ls2) was driven by the carrier solution and mixed thoroughly with the chemical modifier at confluence point (X). The diluted sample was then passed through the mixing chamber and collected in the autosampler cup for 120 s (2 ml) for the determination of total Se.During this step a stream of 0.2% v/v HN03 was circulated upstream in order to wash the filter while another loop (Lsl) was filled with fruit juice. Results and Discussion Conventional Determination of Selenium According to the l i t e r a t ~ r e ' ~ * ~ ~ ~ ~ vaporization and atomiz- ation of Se from a chloride matrix can lead to the formation of SeCl prior to the appearance of Se(g). This matrix strongly depresses the Se signal as a result of gaseous selenium chlorides diffusing from the optical path prior to atomization. The influence of HCl and HNO at concentrations between 0.1 and 0.001 moll-' was therefore investigated and HCl was found to result in low absorbances.On the other hand 0.2% HNO (= 0.03 moll-') introduced no negative effects and resulted in better peak shapes and was therefore chosen as the matrix for sample and standard preparations. The furnace programme used (Table 1) includes two tem- perature ramps in the dry step intended to prevent the sample from 'splattering' owing to the high boiling-point of the matrix. Maximum sensitivity was achieved by stopping the internal argon flow through the furnace during the atomization step. In order to establish the best conditions for the determi- nation of Se the optimum extent of dilution of the slurry samples was determined. Orange juice was used to study the effect of variables potentially influencing the determination of Se in fruit juices.Direct measurements of Se in the juice provided very high concentrations that suggested the need for dilution. The results obtained at a 1 +4 dilution with 0.2% v/v HN03 were not proportional to the extent of dilution (the concentrations were more than five times lower) as a result of interferences from the sample matrix so direct measurements were ruled out. For this purpose commercially available orange juice spiked with 20 pg 1-' of Se was diluted to various extents with 0.2% v/v HNO,. The chemical modifier CO.01 moll-' Pd(N03),] and diluted samples were placed in the autosampler and analysed under the conditions given in Table 1. The precision [repeatability as relative standard devi- ation (YO RSD)] for five sequential measurements of the same orange juice sample diluted 1 + 4 1 + 9 1 + 19 and 1 +29 was 8.4 6.1 7.5 and 11.3% respectively; the corresponding back- ground signals were 0.1,0.040 0.015 and 0.009 (as peak areas).A 1 + 9 dilution was chosen because it resulted in an acceptable background signal and the highest possible precision. These results are consistent with previous reports25 that the precision is degraded by highly diluted slurries because only a small number of particles remain in the slurry. The high volatility of Se poses some problems in its determi- nation by ETAAS. Element losses can occur during conven- tional ashing and charring also some interferences in the vapour-phase may be encountered. Chemical modifiers are an excellent choice for circumventing these problems. To this end various chemical modifiers including Mg( N03)2 Pd(N03)2,Ni(N03)2 and a mixture of Pd(N03) + Mg(N03) at concentrations from 1 x lOP4-l x lo-' mol I-' were tested.The effect of chemical modification on the charring temperature at a constant atomization temperature (1900 "C) in the determi- nation of Se was studied in parallel on an orange juice sample diluted 1 +9 and spiked with 20 pg 1-' of Se and a standard P m 300 700 1100 1500 Charring temperaturerc Fig. 2 Influence of the charring temperature of (a) orange juice diluted 10-fold (spiked with 20pg 1-1 Se) and (h) 20pg1-' Se standard showing peak areas in the absence of chemical modifier A and in the presence of chemical modifier B 0.01 moll-' Pd(N03)z; C 0.01 mol I-' Mg(NO,),; and D 0.1 mol I-' Ni(N03)2. Injected volumes lop1 for the sample and chemical modifier. Furnace con- ditions as shown in Table 1 solution containing 20pg1-' of Se. The results are shown in Fig.2 from which tested concentrations that provided poor results have been omitted for simplicity. In the absence of a chemical modifier the analytical signal for the juice and standard decreased above 600 and 700 "C respectively thus reflecting the high volatility of Se. On the other hand more stable Se compounds were formed if a chemical modifier was used; also the charring temperature could be raised to 900°C [Fig. 2(u)] and 1300°C [Fig. 2(b)] for the juice and standard respectively. Although Ni( N03)2 is recommended for the deter- mination of Se'6.17 it was ruled out in this method owing to the poor sensitivity.Palladium nitrate Mg( N03)2 and Pd( NO3) + Mg( N03)2 performed similarly as modifiers for the juice and standards. Palladium nitrate was finally chosen because it gave rise to slightly higher sensitivity for juice samples than did Mg(N03)2. According to some workers palladium modifies the vaporization of Se by inhibiting the formation of its hydroxide dimer carbide and dioxide.,' A charring temperature of 900°C was selected as this was the highest that the juice samples could withstand and was also acceptable for the standards. An atomization temperature of 1900°C was high enough for the atomization of Se and was therefore used in subsequent experiments. By using an orange juice sample diluted 1 +9 and spiked with 5-30 pg 1-' of Se recoveries of between 92.5 and 111.5% were obtained.Hence the proposed method is accurate enough for the determination of Se in this type of matrix. Flow Injection Manifold The manifold used depicted in Fig. 1 permits dilution of the juice samples and addition of the chemical modifier both being mixed prior to injection into the graphite furnace which increased sample homogenization. The manifold included a filter for the on-line determination of total Se in the slurry and the liquid phase of the juice sample which provided infor-660 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MAY 1994 VOL. 9 mation on its distribution in fruit juices. An ordinary paper disk (Whatman No. 1) was used to filter the sample and was discarded daily. Millipore membranes were tested for this purpose but the filter became clogged after five samples which resulted in a high irreproducibility ( ~ 2 0 % ) ; for these and for economic reasons a paper filter sandwiched into the filter of the manifold was finally chosen.In order to confirm the optimum dilution factor (1+9) obtained for orange juice with the manual procedure 70 100 200 and 400 pl (final volume in the autosampler cup 2 ml) of sample corresponding to 1 + 29,1+ 19,1+ 9 and 1 + 4 dilutions respectively were assayed in the FI system. The Se content in this juice was increased in order to make it adequate for quantitative determination. Thus four juice samples of 25 ml were used spiked with Se to 50 200 400 or 600 pg 1-' from which after homogenization 400 200 100 and 70 pl respect- ively (final concentration in the autosampler cup ~ 2 0 pg I-') were injected by using the FI system.The volume injected could not be increased at will in fact a volume of 400 pl resulted in the highest irreproducibility and background signal. The results obtained for a commercially available orange juice at several dilution factors by manual and automatic dilution are listed in Table2. As can be seen the results obtained by automatic dilution were similar to those provided by manual dilution; also the best precision was obtained at a dilution factor of 1+9. Automatic dilution was more precise than manual dilution because of the diminished human partici- pation. The background signal was also lower on account of the more efficient sample homogenization; thus the signal was 0.030 s with a dilution factor of 1 +9.An injected volume of 200 pl was chosen for all subsequent measurements. The flow rates were adapted accordingly in order to obtain a 1+9 dilution. Thus the flow rate of the carrier solution and chemical modifier used were both 0.5 ml min-' (total volume in the autosampler cup 2 ml in 120 s). Figures of Merit of the Proposed Method The linear portion of the calibration curve obtained by using the flow manifold depicted in Fig. 1 (r>0.997; n = 6 ) ranged from 50.0 to 500.0 pg 1-1 of Se (from 5.0 to 50.0 pg 1-' of Se after 10-fold dilution). The detection limit was calculated as the concentration corresponding to a measurement level of 3sb above the value found in the absence of the analyte with the modifier,36 and was found to be 5 pgl-'.The precision for a commercially available orange juice spiked with 20 pg I-' of Se was determined in duplicate both within runs (n=20 4.0% RSD as repeatability) and between days (11.1% RSD as reproducibility ) . Potential interferences in the determination of Se by ETAAS were investigated by analysing for most of the major ions usually present in fruit juices. Several ions were determined at concentrations 3-fold or higher than those usually found in the mineral composition of the orange juice.37 A given species was considered to interfere significantly if its presence resulted in a change of z & 10% in the signal for a standard containing 15.0 pg I-' of Se. The selectivity results are listed in Table 3 and were fairly good. Also differences in the analytical response (-8.3 to + 11.5%) were quite small so the amounts of these Table 3 Tolerated levels of interferents in the determination of 15 pg 1-' of Se Mineral composition of orange juice/mg I-' K+ 2000 Na+ 90 Ca2+ 140 Fe3 + 3 P 340 S 40 c1- 35 Tolerated level/ mg I-' 2000 6000 2707 4207 10 30 40 340 3 60 1020 120 200 400 450 105 400 500 ~ ~~ Signal difference* (%I - 2.8 + 72.2 - 8.3 + 2.8 + 5.5 - 4.2 - 47.6 + 1.0 + 10.5 + 206.0 + 4.8 + 11.5 + 33.3 - 1.6 - 8.3 - 76.2 - 6.0 * Percent difference between the signals obtained in the presence 7 Maximum concentrations assayed.and absence of interferent. ions usually present in fruit juices should cause no interference especially after samples are diluted 1 +9. The greatest inter- ference was caused by phosphorus and iron.Interference by phosphorus can be ascribed to the formation of molecular bonds (e.g. phosphorus containing molecules) absorbing near the analytical lines for Se.38 At phosphorus concentrations below 360 pg I-' palladium effectively stabilized the formation of phosphorus and various Pd,P phases. However the amounts of these ions usually present in fruit juice should cause no interference. Analysis of Real Samples Prior to the application of the proposed method to real samples the time during which the sample was immersed in the ultrasonic bath for homogenization was varied between 5 and 20 min which resulted in slight differences (Z 5%) in the results relative to those obtained with no ultrasonic agitation. Therefore 10 min of ultrasonic agitation was chosen because the slurry was then most readily transported through the flow system thanks to a decreased particle size and clogging of the filter and transmission lines was avoided.Various juices purchased at local markets in their original Tetrabrik containers were analysed by using the proposed FI system. For this purpose the juices were spiked with Se at concentrations from 50 to 300 pg 1-' prior to introduction into the FI system. The Se content of the juices after 10-fold dilution was too low for quatitative determinations so the standard additions methods was used. Fig. 3 shows the cali- bration curves obtained by standard additions to four types of juice (tomato orange apple and pineapple); the slope of the lowest curve obtained in the analysis of aqueous Se standards was similar to that obtained in the presence of the slurry Table 2 Comparison of the results obtained by using manual and autorriatic dilution in the determination of Se in a commercial orange juice; results given as integrated absorbance f SD Dilution Volume Absorbance RSD Absorbance RSD injected*/pl (manual dilution)/s (%I (automatic dilution)/s W) 1 +4 400 0.035 f 0.00295 8.4 0.023 k 0.00095 4.1 1+9 200 0.029 f 0.00178 6.1 0.030 f 0.001 10 3.7 1+19 100 0.027 & 0.00202 7.5 0.026 0.00148 5.7 1 +29 70 0.024 0.00270 11.3 0.025 & 0,00228 9.1 * In a final volume of 2 ml using the FI system.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MAY 1994 VOL.9 66 1 Table 4 Se contents (pg l-’)* in the slurry and liquid phase (filtrate) of the juices as determined by FI-ETAAS directly in mineralized samples; results given f SD (n = 3) FI-ETAAS Juice Pineapple (A)? Orange( B) Orange(A) Natural Orange Tomato( B) Peach and grape(B) Peach and apple(A) Pineapple and grape(B) Apple( B) Slurry/pg 1 - ’ 15.0f 1.8 55.0 i- 3.2 70.0f3.1 30.0 f 2.1 60.0 f 5.3 150.0 f 10.3 30.0 f 1.9 50.0 i- 4.7 30.0 f 2.8 Filtrate/pg 1-’ 14.5 22.0 34.6 i- 3.8 42.7 & 2.1 20.4 f 3.1 52.6 f 4.2 102.0f 15.3 16.22 1.5 27.0 i- 2.7 27.3 & 3.3 Conventional ETAAS Mineralized sample/pg/l- 14.6i-0.9 - - 30.6 i- 1.7 57.0 f 4.6 157.9f 11.4 47.4 i- 3.6 - - * Dilution factor 1 +4.t (A) and (B) Zumosol and Don Simh Spanish trade names respectively. 0.05 rn \ 8 0.04 C m + 0.03 ^Q -0 a 0.02 2 + WJ g 0.01 - -20 -10 0 10 20 30 Concentration of Se added/pg I-’ Fig.3 Calibration curves for various fruit juices obtained by using the standard additions method A tomato; B orange; C apple; and D pineapple juice; and E calibration curve for an Se standard.All juices were diluted 10-fold except pineapple juice which was diluted 5-fold. Integrated absorbance measured at 196.0 nm matrix. The results obtained for the juice slurries and filtrates are given in Table 4. Identical samples were pre-treated by wet ashing and analysed by direct ETAAS in order to check the accuracy of the proposed method. The good agreement between the results obtained by using the standard additions methods and direct measurements on digested samples reveals that the proposed method is fairly accurate. A comparison of the two sets of results allows one to draw the following conclusions.(i) The proposed method can be used for the direct determi- nation of Se in juices by using the slurry technique with results that are similar to those achieved by digestion of the juice. (ii) The Se content in natural orange is slightly lower than that in commercially available orange juice which indicates that industrial processing/packaging of orange for drinking probably increases the Se content. (iii) Selenium is largely contained (z60-90%) in the liquid phase of foodstuff slurries; the fruit juices with highest solid content (i.e. orange tomato peach) gave the lowest Se concentrations in the slurry filtrates. (iv) The estimated safe and adequate daily dietary intake of Se is 10-200pg,39 so juices can be considered good Se supplements.The Spanish Comision Interministerial de Ciencia y Tecnologia is acknowledged for financial support awarded in the form of Grant No. PB94-0000. M.A.Z.A. is grateful to the Brazilian Government [Conselho Nacional de Desenvolvimento Cientifico e Tecnologico (CNPq) Brazil] and the Spanish Government (Programa de Cooperacion Cientifica con Iberoamerica Spain) for additional financial support. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 References Gennaro A. R. Remington- Farmacia Medica Panamericana Buenos Aires 17th edn. 1987 vol. 2. Arvy P. and Lamand M. Acad. Agric. F. C.R. Seances 1983 69 489. Lamand M. Ann. Nutr. Aliment. 1971 25 B379. Berman E. Toxic Metals and Their Analysis Heyden London 1980. Verlinden M.Deelstra H. and Adriaenssens E. Talanta 1981 28 637. Ebdon L. and Parry H. G. M. J. Anal. At. Spectrorn. 1988,3,131. Mannan A. Waheed S. Ahmad S. and Qureshi I. H. J. Radioanal. Nucl. Chem. 1992 162 111. Clark S. and Craig P. J. Mikrochim. Acta 1992 109 141. Corns W. T. Stockwell P. B. Ebdon L. and Hill S . J. J. Anal. At. Spectrom. 1993 8 71. Smyth W. F. and Jan M. R. Fresenius’ J. Anal. Chem. 1993 346,947. Ma Y. P. Fenxi Shiyanshi 1993 12 87. Hocquellet P. and Candillier M. P. Analyst 1991 116 505. Kumpulainen J. and Saarela K.-E. J. Anal. At. Spectrom. 1992 7 165. Welz B. Bozsai G. Sperling M. and Radziuk B. J. Anal. At. Spectrom. 1992 7 505. Welz B. Schlemmer G. and Mudakavi J. R. J. Anal. At. Spectrom. 1992 7 1257. Cedergren A. Lindberg I. Lundberg E. Baxter D.C. and Frech W. Anal. Chim. Acta 1986 180 373. Radziuk B. and Thomassen Y. J. Anal. At. Spectrom. 1992 7 397. Aller A. J. and Garcia-Olalla C. J. Anal. At. Spectrom. 1992 7 753. Shan X.-q. and Hu K.-j. Talanta 1985 32 23. Styris D. L. Prell L. J. Redfield D. A. Halcombe J. A. Bass D. A. and Majidi V. Anal. Chem. 1991 63 508. Kumar S. J. and Gangadharan S. J. Anal. At. Spectrom. 1993 8 127. Fang Z. Sperling M. and Welz B. J. Anal. At. Spectrom. 1990 5 639. Sperling M. Yin X. and Welz B. J. Anal. At. Spectrom. 1991 6 295. Sperling M. Yin X. and Welz B. J. Anal. At. Spectrorn. 1991 6 615. Bendicho C. and de Loos-Vollebregt T. C. J. Anal. At. Spectrom. 1991 6 353. Jackson K. W. and Qiao H. Anal. Chem. 1992,64 57R. Lindberg I. Lundberg E. Arkhammar P. and Berggren P.-O. J. Anal. At. Spectrom. 1988 3 497. Ihnat M. and Stoeppler M. Fresenius’ J. Anal. Chern. 1990 338 455. Wagley D. Schmiedel G. Mainka E. and Anche H. J. At. Spectrosc. 1989 10 106. Mochizuki T. Sakashita A. Iwata H. Ishibashi Y. and Gunji N. Fresenius’ J. Anal. Chem. 1991 339 889. Qian S. and Yang P. Fenxi Huaxue 1990 18 1064. Bergamin F. H. Reis B. F. Jacintho A. O. and Zagatto E. A. G. Anal. Chim. Acta 1980 117 81.662 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MAY 1994 VOL. 9 33 Wilhelm M. and Onesorge F . K. J. Anal. Toxicol. 1990,14,206. 34 Shekiro J. M. Skogerboe R. K. and Taylor H. E. Anal. Chem. 1988 60 2578. 35 Slavin W. Carnrick G. R. and Manning D. C. Anal. Chem. 1984 56 163. 36 Analytical Methods Committee Royal Society of Chemistry Analyst 1987 112 199. 37 Belitz H. D. and Crosh W. Food Chemistry Springer Verlag New York 1987. 38 Saeed K. and Thomassen Y. Anal. Chim. Acta 1981,130 281. 39 National Academy of Science National Research Council Food Chem. News 1979 21 23. Paper 3/07011 E Received November 25 1993 Accepted February 8 1994

ISSN:0267-9477    

DOI:10.1039/JA9940900657

出版商:RSC    

年代:1994

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MAY 1994 VOL. 9 663 Determination of Nickel in Rocks by Use of a Continuous Precipitation-Preconcentration System Coupled On-line to a Flame Atomic Absorption Spectrometer* Paulo C. S. Mendest and Ricardo E. Santelli Department of Geochemistry University Federal Fluminense Niteroi-24020-007 Brazil Mercedes Gallego and Miguel Valcarcel Department of Analytical Chemistry Faculty of Sciences University of Cordoba E- 74004-Cordoba Spain A rapid accurate and precise method is described for the determination of Ni at the microgram per gram level in silicate rocks. The method is based on the continuous precipitation of Ni with 1 -nitroso-2-naphthol and dissolution of the complex formed in ethanol. The sensitivity of Ni determination is greatly enhanced by the presence of hydrogen peroxide.The metal can be preconcentrated 50-fold using a 10 ml volume of sample with a sampling frequency of 15-20 h-' and a relative standard deviation of 3.2% at 50 pg I-' of Ni. Tiron was used to increase the tolerated amount of Fe up to an Fe:Ni ratio of 7500:l. Six reference silicate rocks were analysed for Ni to test the accuracy of the proposed method; the results were all consistent with the compiled values. Keywords Flame atomic absorption spectrometry; continuous nickel preconcentration; l-nitroso-2- naphthalato-nickel complex; cobalt and iron interference; silicate rocks Nickel is a trace element distributed in igneous rocks as a siderophile element. It occurs at relatively high concentrations in olivine-rich and mafic rocks.On the other hand Ni concen- trations are usually much lower in granitic rocks.' During weathering the element is readily mobilized by coprecipitation with iron and manganese oxides in sediments.2 Nickel in rock sample solutions can be analysed directly by a variety of techniques including X-ray fluorescence (XRF) spectrometry atomic absorption spectrometry (AAS) and inductively coupled plasma atomic emission spectrometry (ICP-AES). For the determination of Ni at concentrations of f 3 pg g- ' the use of a preconcentration technique is r e q ~ i r e d . ~ ~ ~ ~ Even though continuous precipitation was devised in 1987 precipitation has been the least frequently used separation and preconcentration technique in flow injection (FI) system^.^ Flame atomic absorption spectrometry (FAAS) is a technique commonly used in conjunction with FI for both direct and indirect determinations of inorganic anions metal ions' and pharmaceuticals.' Several trace metal determinations using FAAS have been reported.Thus trace levels of Pb in tap water were preconcentrated by using a continuous precipita- tion-dissolution system," the method is based on Pb precipi- tation with ammonia solution and dissolution of the precipitate retained on a filter with a stream of nitric acid. Preconcentration factors of up to 700 were achieved for a 250 ml volume of solution. Also Cu (Ref. 11 ) and Co (Ref. 12) in rocks have been determined after collection of their precipi- tates with rubeanic acid (dithiooxamide) and 1 -nitroso-2-naphthol.The Cu precipitate was dissolved in potassium dichromate and that of Co in ethanol. Preconcentration factors of up to 500 for Cu and 400 for Co were thus achieved. Calcium was preconcentrated after separ- ation from synthetic rock solutions by precipitation with ammonium oxalate; a preconcentration factor of up to 650 for 0.1-1 ngml-I of Ca was obtained.13 More recently a pro- cedure for the coprecipitation of Pb with the iron@-hexahyd- roazepinium hexahydroazepin-1-ylformate complex and collection in a knotted reactor without the use of a filter was reported.14 An enhancement factor of 66 was obtained by * Presented at the Second Rio Symposium on Atomic Absorption t Permanent address USIMINAS Research Center Rod. BR 381 Spectrometry Rio de Janeito Brazil June 21-28 1992.Km 210 Zpatinga/MG 35160-900 Brazil. dissolving the precipitate in isobutyl methyl ketone. Up to 250mg1-1 of Fe in the sample solution could be tolerated which made the method applicable to the determination of Pb in biological materials. The aim of this work was to develop a continuous precipi- tation method for the preconcentration of trace Ni in silicate rocks with a view to establishing a procedure which would take advantage of the high sample throughput of FI systems rather than achieving high enrichment factors at the expense of long analysis times. Experimental Apparatus An atomic absorption spectrometer (Baird Atomic 3400) equipped with a nickel hollow cathode lamp was used. The instrument was set at 232.5 nm and an air-acetylene flame was used according to the manufacturer's recommendations.A single-pen recorder (Instrumentos Cientificos CG) was used to record absorbance signals. A peristaltic pump (Desaga PLG) furnished with silicone rubber tubes was used to pump the solutions. Two Rheodyne 5041 four-way valves connected to two channels were used to switch between the preconcentra- tion and elution solutions. The manifold tubing was 0.5mm i.d. poly( tetrafluoroethylene) (PTFE). A stainless-steel filter (Scientific System 0.5-105 pore size 0.5 pm chamber inner volume 580 pl filtration area 3 cm2) was used to retain the precipitate. Reagents and Solutions All the chemicals used were of analytical-reagent grade and all the solutions were prepared in distilled water. An Ni stock solution 1000 p g ml-' was prepared by dissolving 1.000 g of Ni metal in a minimum volume of nitric acid (1 + 1 v/v) and diluting to 11 thereby keeping the nitric acid concentration at about 1 % v/v.Working strength solutions were prepared daily by appropriate dilution from the stock solution. A 1 mol 1-1 acetic acid-acetate buffer was prepared by dissolving 136.1 g of sodium acetate trihydrate and 57.5 ml of concentrated acetic acid in about 900 ml of water. The pH was adjusted to 5 with sodium hydroxide and the volume was made up to 1000ml. A 0.1% m/v solution of 1 -nitroso-2-naphthol was prepared by dissolving 0.1 g of664 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MAY 1994 VOL. 9 reagent in 50 ml of 95% ethanol and diluting to 100 ml with water. This solution was stable for at least 2 weeks.Anhydrous ethanol hydrogen peroxide triethanolamine and Tiron (4,5-dihydroxy-1,3-benzene-disulfonic acid disodium salt) were used as received. Sample Preparation The samples used in this work were international rock reference materials. Silicate rock materials were decomposed by using standard methodology4 as follows; an accurately weighed amount of finely powdered rock (between 0.25 and 5 g) was placed in a PTFE beaker and moistened with water followed by the addition of 10 ml of 48% hydrofluoric acid and 1 ml of 70% perchloric acid. After evaporation to perchloric acid fumes in a sand-bath the residue obtained was cooled a further 5ml of 48% hydrofluoric acid were added and the contents of the beaker were again evaporated to near dryness. The cooled residue was dissolved in the required volume of buffer and the solution was adjusted to pH 5 with sodium hydroxide.Next Tiron was added to the solution followed by dilution with water (to 50 or 250m1 depending on the Ni content) to obtain a concentration of 0.2 moll-' in the buffer and 10% m/v in Tiron. Any solution turbidity (e.g. from aluminium or silica) was eliminated by filtering the solution through Whatman No. 1 paper. The sample solutions thus obtained were stored in polyethylene bottles 0.1 ml of hydro- gen peroxide (30% v/v) per 10 ml of sample was added before processing by the preconcentration flow system. Procedure The flow manifold used is depicted in Fig. 1. A volume of 1Oml of sample containing 0.1-1.2pg of Ni" in 0.2moll-' acetate buffer at pH 5.0 (plus 0.3% hydrogen peroxide and 10% m/v Tiron) was continuously mixed with the precipitating reagent solution (0.1 YO 1-nitroso-2-naphthol in an etha- nol-water mixture).The 1-nitroso-2-naphthalato-nickel com- plex produced was retained on the filter and the sample matrix was directly driven to waste. Simultaneously a water stream was aspirated to flush the AAS nebulizer and the dissolving agent was re-circulated through the system. After the whole sample volume was introduced both of the switching valves (SV1 and SV2) were actuated simultaneously; thus the ethanol stream was passed through the precipitate to dissolve it and sweep the Ni to the instrument. A transient signal was then recorded the peak height of which was used as the analytical signal.No blank or precipitate washing was required. Results and Discussion Optimization of Chemical Variables By using the manifold shown in Fig. 1 and a sample volume of 10m1 the influence of chemical variables was studied to I ml min-' I 1 Fig. 1 Flow injection manifold used for the preconcentration and determination of nickel in rocks D ethanol; S sample; R reagent (0.1 % m/v 1-nitroso-2-naphthol); W waste; SV switching valve; and AAS atomic absorption spectrometer establish the best chemical conditions for the precipitation reaction and dissolution of the precipitate. The univariate method was used for this purpose. Firstly the effect of sample pH values in the range 2-6 was studied by using standard solutions containing 100 pg 1-' of Ni. The maximum precipitation efficiency was obtained between pH 4.5 and 5.5; below pH 4.0 no signal was obtained because Ni was not precipitated.A pH of 5 adjusted with acetic acid-acetate buffer was selected. The effect of the buffer concentration in the sample solution was studied between 0.1 and 0.8 moll-'; this variable did not affect the analytical signal so a 0.2moll-' acetic acid-acetate buffer was chosen for subsequent experiments. The influence of reagent concen- tration was tested over the range 0.001-0.1% m/v. The best results were obtained at the highest concentration assayed. Because of insolubility of the reagent in the ethanol-water solution concentrations above 0.1% m/v were not assayed as the ethanol content would have to be increased thereby increasing the solubility of the precipitate. The effect of the reaction coil/filter temperature was studied over the range 15-60 "C.This variable did not affect the atomic signal up to 30 "C above which the signal decreased probably due to increased solubility of the precipitate. The dissolving agent was selected from a test involving 95% ethanol anhydrous ethanol and methanol. All three solvents proved satisfactory; however the signal obtained with methanol was approximately 20 and 40% higher relative to anhydrous ethanol and 95 % ethanol respectively. Anhydrous ethanol was eventually chosen on account of its lower toxicity and cost. Finally the effect of sonication of the reaction coil/filter in the preconcentration or elution step was studied; ultrasound had no enhancing effect on the signals.Optimization of Flow Variables Flow variables were also optimized by using the univariate method. The effect of sample flow rate was investigated between 1.2 and 4.4 ml min-l. The maximum signals were obtained at 3.0 ml min-'; lower and higher flow rates gave rise to smaller signals. Because of the longer residence times at lower flow rates the precipitate adhered to the tube walls which increased the dispersion. On the other hand residence times at higher sample flow rates were too short for quantitative formation of the precipitate. The influence of the reagent flow rate was studied over the range 0.2-0.8 ml min-' where the signal was found not to change. A reagent flow rate of 0.8 ml min-' was thus chosen because it resulted in a larger amount of reagent being intro- duced into the manifold.The effect of the flow rate of the dissolving agent was studied between 1.2 and 3.0 ml min-'. The best results were obtained at 3.0 ml min-' as for the AAS nebulizer aspiration rate. The influence of the reaction coil length was studied from 50 to 200cm. The best system performance was obtained with a 100 cm reaction coil. Longer and shorter reactors resulted in decreased signals as for the sample flow rate (by virtue of the differences in the residence times). Study of Interferences Under the optimum chemical conditions and using the mani- fold depicted in Fig. 1 an exhaustive study of interferences was carried out. In previous work 1-nitroso-2-naphthol was used for the quantitative precipitation of Co over the pH range 3.6-4.812 in the presence of several cations; however Ni" and Fe"' were found to interfere at concentrations ten times higher than that of Co because both precipitated with 1-nitroso-2-naphthol in acid media and the Co precipitate was partly occluded in the Ni or Fe precipitate.Both elements were therefore studied. Portions (10 ml) of Ni solutions (100 pg I-') containing Co at several concentrations wereJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MAY 1994 VOL. 9 665 assayed. Concentrations of Co ten times higher than that of Ni were found not to interfere. Although the Co Ni ratio is never that high in silicate rocks several masking reagents (hydrogen peroxide and triethanolamine) were investigated to overcome the interference. Hydrogen peroxide increased the atomic signal of Ni probably because the composition of the complex salt was similar to that for Co where the metal is trivalent.Cobalt(I1) is oxidized to Co"' by the reagent but some workers recommend using hydrogen peroxide to convert Co" into Co"' before reacting it with l-nitroso-2-naphthol. Thus addition of 0.1 ml of H202 (30% v/v) per 10ml of sample increased the analytical signal for 100 pg I-' of Ni by about 65% which lowered the detection limit of Ni; however hydrogen peroxide did not overcome the Co interference. Triethanolamine was also investigated to overcome the interference of Co; the results were higher than with hydrogen peroxide [between 1 and 2% v/v of triethanola- mine added to a sample containing 0.1 pg ml-' Ni and 10 pg ml-' Co eliminated the interference of the latter]; how- ever the analytical signal for Ni increased by only about 30%.The precipitate from Fe"' clogged the filter and rendered the determination of Ni impossible. Since Fe" has a lower affinity for this reagent than does Fe"' a reductant such as hydroxylam- ine hydrochloride was investigated. The Ni signal decreased substantially as a result so use of a reducing agent was rejected. Iron (11) ligands such as o-phenanthroline and bipyridine proved ineffective since they also reacted with Ni thereby completely suppressing the Ni signal. Only Tiron was found to complex Fe(II1) effectively with no disturbing effect on Ni. Tiron forms a highly stable purple complex with Fe(II1) in an acid medium. As can be seen in Table 1 10% Tiron suppressed the inter- ference of Fe(m) up to 750 pg ml-' [Fe"':Ni ratio of 7500:1].However the presence of Tiron and triethanolamine in the same solution was chemically incompatible; in fact the Ni signals decreased by about 70% as a result. Therefore as the Co" concentration in silicate rock solutions is never ten times greater than that of Ni" 0.3% hydrogen peroxide and 10% Tiron solutions were used with no triethanolamine. The effect of Tiron was studied by introducing it continuously into the system before the sample was mixed with l-nitroso-2-naphthol; the results were not as good as those obtained with manual addition because the sample was substantially diluted. Major and minor elements commonly found in silicate rocks some of which can precipitate in the weakly acidic medium used or with l-nitroso-2-naphthol and clog the filter were also investi- gated.No interference was caused by Ca2+ Mg2+ Na+ K+ Mn2+ or A13+ at concentrations of 1000 pg ml-' (the tolerated amounts 10000 times higher than that of Ni). Copper (11) Zn2+ and Cr6+ were only assayed at concentrations up to 100-fold that of Ni since they normally occur at concentrations only 5-20 times higher in silicate rocks; none of these elements was found to interfere (the tolerated ratio was 1OO:l). Determination of Nickel Two calibration graphs were run under the optimized chemical and flow conditions using a sample volume of 10ml and the Table 1 Effect of Tiron hydrogen peroxide and triethanolamine on the interference of Fe"' on the signal of 0.1 pg ml-' of Ni Fe"'/ pg ml- ' 0 0 0 0 0 500 500 750 1000 H202 (Yo v/v) 0 0.3 0.3 0 0.3 0.3 0.3 0.3 0.3 Triethanolamine 0 0 1 1 0 0 1 0 0 (Yo v/v) Tiron (YO m/v) 0 0 0 0 10 10 10 10 10 Absorbance 0.065 0.105 0.098 0.083 0.106 0.105 0.033 0.103 0.090 Table 2 Analysis of silicate rocks for nickel Nickel content/pg g-' Sample Found* Compiled'* JR-1 (GSJ) rhyolite 0.58 f 0.03 0.66 (proposed) QLO-1 (USGS) quartz latite 5.6 f0.8 5.8 (information) SDC-1 (USGS) mica schist 39f2 38 (proposed) MAG-1 (USGS) marine mud 56f 1 53 (recommended) BHVO-1 (USGS) basalt 125 f 10 121 (recommended) DNC-1 (USGS) diabase 242 +_ 16 247 (information) * Average of four individual determinations f standard deviation.manifold depicted in Fig. 1. In the absence of masking agents the calibration graph was linear for 20-200 pg 1-' of Ni.In the presence of masking agents (i.e. 0.3% hydrogen peroxide and 10% Tiron) the calibration graph was linear for 10-120 pg 1-' of Ni. The corresponding equations were absorbance =0.003 + 6.3 x 10-4x (x 20-200 pg I-' of Ni); and absorbance=0.002+ 1.04 x 10-3x (x 10-120 pg 1-l of Ni). The slope of the calibration graph for 10-120 pg 1-' of Ni was higher than that for 20-200 pg I-' of Ni owing to the presence of hydrogen peroxide (as noted above) hence the sensitivity of the method for the determination of Ni was increased by about 65%. The detection limit ( 5 pg 1-' of Ni) was calculated as three times the standard deviation of the absorbance for ten determinations of a sample containing 10 pg 1-' of Ni using the calibration graph for 10-120 pg 1-' of Ni.The precision obtained for 11 samples containing 50 pg I-' of Ni expressed as the relative standard deviation was 3.2%. The sample throughput (excluding sample dissolu- tion) was 15-20 samples h-'. A concentration factor of 50 calculated as the ratio between the slopes of the calibration graphs obtained by using the proposed method and direct aspiration [A =0.001+2.0 x (Ni2+)] was achieved. Analysis of silicate rocks The applicability of the proposed method was checked by analysing international silicate rocks reference samples from the United States Geological Survey (USGS) and the Geological Survey of Japan (GSJ). These materials were dissolved as described under Experimental and the Ni content of the resulting solutions was determined by using the rec- ommended procedure.The results obtained are listed in Table 2. As can be seen the proposed method is suitable for the determination of Ni in rocks with contents as low as 0.6 pg g-' with satisfactory accuracy and precision. By using volumes greater than 10 ml (e.g. 100 ml) the preconcentration factor can be increased 10-fold as shown in the experiments described above performed to determine Co in rocks by using a similar manifold.12 With slight alterations the proposed method could be used for sequential determinations of Co and Ni in silicate rocks. The authors gratefully acknowledge Conselho Nacional de Desenvolvimento Cientifico e Tecnolbgica Fundagiio de Amparo a Pesquisa do Estado do Rio de Janeiro and Pro- Reitoria de Pesquisa e Pos-Gradnaqiio Universidade Federal Fluminense for the award of financial support and a research fellows hip.References 1 Jeffery P. G. Methods of Rock Analysis Pergamon Press Oxford 2nd edn. 1975. 2 Handbook of Geochemistry ed. K. H. Wedephol Springer-Verlag Berlin 3rd edn. 1978 vol. 11. 3 Reeves R. D. and Brooks R. R. Truce Element Analysis of Geological Materials Wiley New York 1978.666 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MAY 1994 VOL. 9 4 5 6 7 8 9 10 11 12 Sulcek Z. and Povondra P. Decomposition in Inorganic Analysis 13 CRC Press Florida 1989. 14 van Loon J. C. Analytical Atomic Absorption Spectroscopy Selected Methods Academic Press New York 1980. 15 Martinez-Jimenez P. Gallego M. and Valcarcel M. Anal. Chem. 1987 59 69. 16 Kuban V. Fresenius’ J. Anal. Chem. 1993 346 873 Valcarcel M. and Gallego M. Trends Anal. Chem. TrAC 1988 17 8 34. Valcarcel M. Gallego M. and Montero R. J. Pharm. Biomed. 18 Anal. 1990 8 655. Martinez-Jimenez P. Gallego M. and Valcarcel M. Analyst 1987 112 1233. Santelli R. E. Gallego M. and Valcarcel M. Anal. Chem. 1989 61 1427. Santelli R. E. Gallego M. and Valcarcel M. J. Anal. At. Spectrom. 1989 4 547. Adeeyinwo C. E. and Tyson J. F. Anal. Proc. 1989 26 375. Fang Z. Sperling M. and Welz B. J. Anal. At. Spectrom. 1991 6 301. Callahan C. M. Fernelius W. C. and Block B. P. Anal. Chim. Acta 1957 16 101. Erdey L. Gravimetric Analysis. Part 2 Pergamon Press Oxford 1965. Sandell E. B. and Onishi H. Photometric Determination of Traces of Metals. General Aspects Wiley New York 1978. Govindaraju K. Geostand. Newsl. 1989 Special Issue 113. Paper 3/07098K Received December I 1993 Accepted January 24 1994

ISSN:0267-9477    

DOI:10.1039/JA9940900663

出版商:RSC    

年代:1994

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MAY 1994 VOL. 9 667 Aboal-Somoza Manuel 469 Absalan G. 45 Adams F. 151 Akman Siileyman 333 Alves Luis C. 399 Amarasiriwardena Dula 199 Anderson David R. 67 Anderson S. E. 263 Anghel Sorin D. 635 Anzano Jesus M. 125 Argentine Mark D. 199 Arnold J. T. 263 Arriagada Lorna 93 Arruda Marco A. Z. 657 Avila Akie K. 543 Back M. H. 45 Barciela-Alonso Carmen 469 Barnes Ramon M. 199 Barrios Carlos 535 Barshick Christopher M. 83 Baxter Douglas C. 297 Bayne Charles K. 83 Beauchemin Diane 509 Becerra Jose 535 Begley Ian S. 171 Belarra Miguel A. 125 BeneS Petr 303 Bermejo-Barrera Adela 469 483 Bermejo-Barrera Pilar 469 483 Bernasconi G. 151 Berndt Harald 39 193 Betti Maria 385 Biffi Claudio 443 Blanco Gonzalez E. 281 Boge Edward M. 369 Branch Simon 33 Briand Alain 17 Brown Nicole V.363 Bruhn Carlos G. 535 Bruno Sergio N. F. 341 BudiC Bojan 53 Burakov V. S. 307 Cabon J. Y. 477 Camara Carmen 291 Campbell Michael 187 Campos Reinaldo C. 341 Carrion Nereida 205 217 Caruso Joseph A. 145 Castillo Juan R. 125 311 Cervera Maria Luisa 651 Chakrabarti C. L. 45 Chartier Frederic 17 Cheam Venghout 3 15 Chirinos Jose 237 Cimadevilla Enrique Alvarez- Cabal 117 Cobo I. G. 223 Coedo A. G. 223 Cooper 111 C. B. 263 Cordos Emil A. 635 Cornejo Silva G. 93 Crews Helen M. 615 Cserfalvi Tamas 345 Cujes Ksenija 285 Curtius Adilson J. 341 543 Dadfarnia Shayessteh 7 Dahl Kari 1 Dams Richard 23 177 187 Dean John R. 615 de la Guardia Miguel 651 Deruaz D. 61 Desrosiers Roland 3 15 Doner Giileren 333 483 CUMULATIVE AUTHOR INDEX JANUARY-MAY 1994 Dorado M.T. 223 Du Xiaoguang 629 Duan Yixiang 629 Durrant Steven F. 199 Ebdon Les 33 611 615 Elgersma Jaap W. 619 Eljuri Elias 205 Elmahadi H. A. M. 547 Emteborg Hiikan 297 Epler Katherine S. 79 Fadda Sandro 519 Fell Gordon S. 457 Fernandez de la Campa M. R. Fernandez Alberto 205 217 Fischer Johann L. 623 Fischer W. 257 375 Fisher Andrew S. 611 Florian K. 257 Fonesca Rodney W. 167 Foster Robert D. 273 Foulkes Michael E. 615 Frentiu Tiberiu 635 Gallego Mercedes 657 663 Geertsen Christian 17 Gilmutdinov Albert Kh. 643 GomiSCek Sergej 285 Gonzalez Urcesino 535 Goode Scott R. 73 Goossens Jan 177 187 Gower Stephen A. 363 369 Gras Nuri T. 535 Greenfield S. 565 Greenway Gillian M. 547 Gregoire D. Conrad 393 605 Hadgu Negassi 297 Harnly James M. 419 Hatterer Andre 525 Hauptkorn Susanne 463 Heitmann U.437 Hernandez Cordoba Manuel Hese A. 437 Hiernaut Tania 385 Hinds Michael W. 451 HlavaCek I. 245 251 HlavaEkova I. 245 251 Holcombe James A. 167 415 Horlick Gary 593 Houk R. S. 399 Hoult Gavin 7 Howe Alan M. 273 Hu Yanping 213 Huang Zhuoer 11 Hudnik Vida 53 Hutton J. C. 45 Hutton Robert C. 385 Imai Shoji 493 Isaevich A. V. 307 Itriago Ana 205 Jackson Jason G. 167 Jakubowski Norbert 193 Janssens K. 15 1 Jaramillo Victor H. 535 Jepkens Brigitte 193 Jin Qinhan 629 Jones Delwyn G. 369 Katskov Dmitry A. 321 431 Kimber Graham M. 267 Kmetov Veselin 443 Koch Lothar 385 Kogan Valentina V. 451 Kolihova Dana 303 Kratzer Karel 303 Krieger Brian L. 267 Krivan Viliam 463 23 1 553 Krushevska Antoaneta 199 Kubova Jana 241 Kumamaru Takahiro 89 Kurfiirst U.531 Lacour Jean-Luc 17 Lazik C. 45 Le Bihan A. 477 Lechner Josef 3 15 Lee Julian 393 Lile E. S. 263 Lopez Garcia Ignacio 553 Lopez Jose C. 651 Lopez-Gonzalvez M. Angeles Lord 111 Charles J. 599 Luecke Werner 105 Manickum Colin K. 227 Manninen Pentti K. G. 209 Manzoori Jamshid L. 337 Marais Pieter J. J. G. 321 431 Marchante Gayon Juan Manuel 117 Marcus R.K. 45 Marin Sergio R. 93 Martinez-Garbayo Maria Paz Martinsen Ivar 1 Massey Robert C. 615 Mauchien Patrick 17 McAllister Trevor 427 McCrindle Robert I. 321 431 McLeod Cameron W. 67 Mendes Paulo C. S. 663 Mermet Jean-Michel 17 61 217 Mezei Pal 345 Michel Robert G. 501 Milagros Gbmez M. 291 Miller-Ihli Nancy J. 605 Milton Dafydd M.P. 385 Minnich Michael G. 399 Misakov P. Ya. 307 Moens Luc 177 187 Montoro Rosa 651 Moreda-Piiieiro A.483 Moreda-Piiieiro J. 483 Mori Toshio 159 Moulton Gary P. 419 Murillo Miguel 205 217 237 Nagulin K. Yu. 643 Nakahara Taketoshi 159 Naoumidis A. 375 Naumenkov P. A. 307 Nevoral Vladislav 241 Nickel H. 257 375 O’Haver Thomas C. 79,419 Okamoto Yasuaki 89 O’Neill Peter 33 Outred Michael 381 Palacios M. Antonia 291 Patriarca Marina 457 Pauwels J. 531 Payling Richard 363 369 Peachey Russell M. 267 Perez-Arantegui J. 311 Perez Parajbn Juan M. 11 1 Petrucci G. A. 131 Petty John D. 267 Popescu Adrian 635 Poussel E. 61 Prudnikov Evgeniy D. 619 Quentmeier Alfred 355 QuerrC G. 3 11 Rademeyer Cor J. 623 Radziuk Bernard 1 29 1 125 Raikov S. N. 307 Rasmussen Gert 385 Reija Carmen 651 Reyes Olga 535 Rivoldini Alessandro 5 19 Rodriguez Aldo A.535 Romon-Guesnier Sabine 199 RonEevik Sanda 99 Rubio J. 151 Rummeli Mark H. 381 Salbu Brit 1 Saleemi Abdollah 337 Salud Seremi 535 Santelli Ricardo E. 663 Sanz-Medel Alfredo 11 1 117 Schaldach Gerhard 39 Scheie Andrew J. 415 Schneider Germar 463 Schoknecht G. 437 Schwarzer Rudolph 43 1 Sekerka Ivan 3 15 Selby Mark 267 Sharp Barry L. 171 Sheppard Brenda S. 145 Shtepan Aleksander M. 321 Siroki Marija 99 Sjostrom Sten 17 Smit Henri C. 619 Smith B. W. 131 Smith Clare M. M. 419 Smith David H. 83 Smith Fraser O. 267 Smith Trevor A. 67 SpevaEkova Vera 303 Steers Edward B. M. 381 Stevenson C. L. 131 StreSko Vladimir 241 Stuewer Dietmar 193 Sturgeon Ralph E. 493 605 Su Evelyn G. 501 Sugawa Kazumitsu 89 Sy T. 437 Thomas Christopher L. 73 Thomassen Yngvar 1 Thompson K.Clive 7 Tittarelli Paolo 443 Tsalev Dimiter L. 405 Turak Elvan E. 267 Turk Gregory C. 79 Valcarcel Miguel 657 663 Valdes-Hevia y Temprano Vanhaecke Frank 187 Vanhoe Hans 23 177 187 Veber Marjan 285 Verbeek Alistair A. 227 Versieck Jacques 23 Vifias Pilar 553 Vincze L. 151 Wade Jeffery W. 83 Walter Serge 525 Webb C. 263 Weiss ZdenEk 351 Wiederin Daniel R. 399 Winefordner J. D. 131 Worsfold Paul J. 611 Wrobel Katarzyna 117 28 1 Xiao Grace 509 Zakharov Yu. A. 643 Zander A. T. 263 Zhang Zhanxia 213 Zheng Jianguo 213 Zilkova Jana 303 231 281 M. C. 231

ISSN:0267-9477    

DOI:10.1039/JA9940900667

出版商:RSC    

年代:1994

数据来源: RSC