|
11. |
Cross-sections of spectrochromatograms for the resolution of folpet, procymidone and triazophos pesticides in high-performance liquid chromatography with diode-array detection |
|
Analyst,
Volume 121,
Issue 10,
1996,
Page 1367-1372
J. L. Martínez Vidal,
Preview
|
PDF (806KB)
|
|
摘要:
Antrlyst, October 1996. Vol. 121 (1367-1372) 1367 Cross-sections of Spectrochromatograms for the Resolution of Folpet, Procymidone and Triazophos Pesticides in High-performance Liquid Chromatography With Diode-array Detect ion J. L. Martinez Vidal, P. Parrilla, M. Martinez Galera and A. Garrido Frenich Depar-tnici?t of' Arwlytical Chemisti-y, 7 J ~ i i ~ r s i t y of' Alnr~iY'u, 04120 AImci-fa, Spa in The rapid-scanning photodiode array detector generates a considerable amount of data in HPLC, such as the three-dimensional (A, h, t ) matrix, but requires improvements in data analysis methodology to utilize all the available information. In this paper, a graphical technique is used for improving the selectivity of HPLC analysis, using the available spectrochromatographic information in both the time and wavelength domains.The technique consists in performing cross-sections through the data matrix to obtain selective analytical information for each of the analytes. In order to demonstrate the validity and simplicity of the method it has been applied in the simultaneous determination in mixtures of the three pesticides folpet, procymidone and triazophos. The procedure was applied with satisfactory results in the determination of these pesticides in groundwater at ppb levels after solid-phase extraction with C18 cartridges. Keywords: Pestic.idcs; f d p e t , procymidoiie nix/ ti.iiizophos; M'atei-; h ig Ii -peifoi-nmiic*e 1 iqii id chi-onza tog rap h y ; diode - ai*ra~ detection; c,r.o.ss-set.tions Introduction The monitoring of pesticides i n different environmental niatri- ces is an analytical problem of growing importance.Ideally, the deployment of few. inexpensive multi-residue (MR) methods would facilitate the rapid identi f'ication and quantification of a wide range of pesticides at the required sensitivity limit, in response to legislation in many countries. Different techniques have been applied in the determination of pesticides, mainly employing CCI-I and HPLC4-7 with a variety of detectors. However, the properties of relatively new classes of pesticides, such as phenylureas, phenoxy acids, carbainates and quaternary amines, make these more suitable to HPLC than to GC. There are several detection methods for HPLC analysis, such as UV/VIS, refractive index, electrochemical, fluorescence and chemiluminescence.UV/VIS-based diode-array detection methods (DAD) is one of the most commonly used multi-wave length detection methods in HPLC to gain more analytical inform at i on about t lie in i x t ures of i n t crcs t .8- I 0 The use of HPLC-DAD has several advantages beyond simple component identification. In the temporal domain, the analysis time can be shortened because the wavelength dimension allows the analyst to observe all UV/VTS-absorbing components during a single elution. In the spectral domain, an improvement in the detection level is gained owing to the availability of the total UV/VIS spectrum. In multicomponent mixtures, where the analytes are not resolved by the column but where spectral overlap is minimal, the analytes can be determined simultaneously by monitoring each component at a wavelength that is free of interference. Hence the analyst can obtain rnultiwavelength chromatograms from a single analysis of one sample in order to resolve overlapping signals.On the other hand, in the case of complex mixtures, where all the analytes comprising a single elution profile and where spectral overlap is severe," the technique of obtaining chromatograms at different wavelengths is not adequate t o resolve overlapping peaks. Typical examples of overlapped peaks can occur if new pesticides have to be checked with an established MR method or if interferents from complex sample matrices are co-eluted. In this situation, where the overlapping signals do not permit the analysis of all analytes in a single chromatographic run, it is possible to modify the MR method or to apply chemometric techniques in order to extract useful information from the overlapped region.The first solution is not the most adequate because of the great cost involved in developing a new method. Therefore, the second solution is usually chosen, but a single compromise detector wavelength has to be selected to apply the majority of chemometric methods. Moreover, one intrinsic advantage of multiwavelength detection in HPLC is that data are directly available in digitized form for storage and software manipulation, by a number of experimental procedures for the characterization of unresolved peaks. Also, many of the algorithms developed for analytical spectroscopy can be used for data analysis in HPLC-UV/VIS.l4-I9 The objective of this work was to apply a methodological approach to extract selective analytical information from the data generated by HPLC-DAD. The method consists in the generation of cross-sections through the three-dimensional ( A , h, t ) matrix providing the optimum signal relative to possible interfering analytes to gain selectivity. The proposed method allows the combined use of spectral and chromatographic information for the deconvolution of overlapping peaks, opening up new prospects for DAD in HPLC. This method has been applied in the determination of the pesticides folpet, procymidone and triazophos simultaneously present in synthetic mixtures. Generally, these compounds are determined by chromatographic methods, either GC or HPLC.Methods for the GC determination of folpet, procymidone and triazophos have been reported with electron-capture,2" ther- nioionic N-P," tlame ionization2* and mass detection.2? HPLC methods have been used with UV24325 and mass detection.26 The procedure was applied to the determination of these pesticides in groundwater at ppb levels after solid-phase extraction (SPE) with CIS cartridges.I368 Aiiirlyst, Octohci. 1996, Vol. 12 I Experimental R eage it ts HPLC-grade solvents were used. The pesticide standards (pestanal quality), summari& i n Table I , were obtained from Riedel-de HaEn (Seelze, Gerrnany). Solid \tandads were dissolved in acetonitrile (ACN) and \tored at 4 "C in the dark. where they were stable for \everal months. Working solutions were prepared daily by appropriate dilution with ACN.Mobile phases were de-gassed with heliuni prior and during u\e. Distilled water was obtained I.rom a Millipore (Bedford, MA, USA) Milli-Q water purification system. All solvents and samples were filiered through Millipore rneinbrane filters before iiijection into the column. Prepacked Sep-Pak C I x cartridges containing 360 mg of' C I chemically bonded silica (Waters, Milford, MA, USA) were used. Apparutus A Watery Model 990 liquid chromatographic system was used, equipped with ;I Model 600E constant-flow pump. a Rheodyne six-port injection valve with a 20 pl sample loop and a Model 990 photodiode-array detector. The spectral resolution used was 1.4 nm per diode in the range 200-280 nm. HPLC separations were carried out using ;t Eiypersil Shandon Green Env. I SO X 3 nirn id ( 5 pin particle sirre) C l x column.Sojtware A coinpati ble personal computer provided with a 486 micro- processor and mathematical coprocessor was used for acquisi- tion ancl treatment of the data. 'The liquid chromatographic system allow\ the acqui5ition of' ;t series of chromatograms at different wavelengths. The Waters 99 I software controlling the instrument generates a three-dimensional file ( A , A. f) in binary format. Then, the three-dimen~ional tile is converted into a series of / I individual spectra. each corresponding to an absorption spectrum, acquired tit a different time, with the ASCII converter included in the Waters 99 1 program. The resolution used i n the time doinain is 1.4 c .A converter program i n BASIC was used to transform the two-dimensional files into ASCII format tor the software packages SURFER and GRAPHER.27 The three-dimensional spectrochromatc,grams are obtained and presented as isometric plots (A, h, t). Alternatively. the data are presented as a contour plot in both the time and wavelength dimensions, by linking points of equal intensity to form the contour map. The SURFER program permits the generation of cross-sections and shows the trajectory followed in the contour or isometric plot. Using GRAPHER software, cross-section data are plotted to produce a profile from the two-dimensional data projection [A-f(h, t ) ] . When the data are plotted, the absorbance value is plotted a s the y coordinate. HPLC Operating Conditions The following conditions were used: flow rate, I ml min-1; chart speed, 0.5 cm min-1; detector sensitivity, 0.02 a.u.f.s.; and column at room temperature.The solvent programme was as follow\: initially 2 min isocratic with water-ACN-MeOH (56 + 27 + 17). followed by a 20 min linear gradient water-ACN- MeOH ( 5 + 90 + 5 ) ; an additional period of 1 0 min of gradient programme was sufficient to return the \ystem to the initial conditions for subsequent analysis runs. The solvents were filtered daily through a 0.45 p i cellulose acetate (water) or PTFE (ACN) membrane filter before use, and degassed with helium during and before use. Results and Discussion Fig. 1 ( a ) shows a chromatogram corresponding to 2 I pesticides selected for their agricultural interest.The mixture contains organochlorines, triazines, organophosphoru\ compounds. car- barnates and ureic and imidic derivatives with very different polarities. The composition of the mobile phase was optimized by an automated sequential procedure.'"-'x However, over- lyping of peaks occurs if the number of analytes increase\. Fig. l(h) shows a chromatogram containing a new analyte, 19rocyniidone (peak 1 I ) , and overlapping among the peaks of folpet, procymidone and triazophos can be observed. Taking into account the absorption data of the mixture in question, 2 10 nm was first selected as the monitoring wavelength for the detection of the three compounds compromise value. The Table 1 Retenlion time\ of pe4cides i n the multi-residue method Peak ho. Pemctde Re tent i o n ti me/ni i t i 1 2 4 3 0 7 X 9 I 0 I I I' I3 13 1s 16 17 18 I9 2 0 31 22 3 M e t om y I Dimethoate A 1 d i c arb Diclorvos C a r bofu ran Atrnzi tie D i u ron D ich lo ran Mcthiocarb Folpet Procy midone Trinzophos Iproclionc Vinclnzolin Ch lorf'env i nphos Chlorpyrifos methyl Endosulfan sulfate Tetrad ifon I',-Endosul fan oc-Endosulf'm Chlorpyrifos cthyl C'arbophenothion 3.3 3 .1 4.4 5.6 5.2 7.3 8.6 9.9 11.2 13.1 13.4 13.7 13.9 14.7 13.9 16.4 16.7 17.8 18.0 i 8.4 18.7 19.4 0.03 0 0.02 e 2 J r; 0.01 0.00 -11 771 1 T l T-m 5 10 15 T i m h i i n -0.0 I ? Fig. 1 ( a ) C'hromatogram obtained by iri.jection of 20 p1 of pesticide standard solution with a 20 min gradient (2 yg r i l l - - ' of each pesticide at 2 I0 nm). Numbers above the peaks correspond with those given in Table I .( h ) Chromatograni w,ith a new analyte. procymidone (peak number 1 I ) . is observed with 20 min gradient (9 1.18 ml- I of folpet, 4 pg m - k ' of procymidone and 6 png ml I of triazophos).Anolvst, Ortohcv- 1996. Vol. 12 I 1369 R, values are 0.9 for folpet and procymidone and 0.7 for procymidone and triazophos. Table 1 summarizes the retention times of each pesticide. Three-dimensional Spectrochroniatograms The diode-array detector allows the collection of full spectral data at rates of up to several scans per second. With the data it is possible to construct three-dimensional plots of absorbance. wavelength and time. Moreover, these plots can be manipulated to allow the data to be viewed from different angles, including from the end of the chromatogram towards the beginning.The corresponding absorption maxima are located at 226 nm for folpet. at 206 nm for procymidone and at 200 and 245 nm for triazophos (Fig. 2). From the observation of the corresponding absorption \pectra, it i\ evident that folpet and procyniidone present their absorption maxima at close wavelengths, whereas triazophos presents the second maximum absorption at a longer wavelength, but its absorption spectrum overlaps in part with that of procymidone. A potentially more informative way of presenting the chromatograms is to use the cartographic technique of a contour plot, a map of signal intensity in the wavelengtli-time domain (Fig. 3). From this plot it is easier to 4ec the incomplete resolution of folpet, procymidone and triazophos. Because of the highly overlapping peaks, conventional measures of the dillerent analytical signal\ (area or height of chromatographic peaks) cannot be realized. With the aim of resolving the ternary rn i x t u re, a c h e in o m c t r i c a p p ro ac h was e v a I u a t ed .Cross-section Optimization Through the Three-dimensional Data Matrix The contour plots are especially useful i n making cross-sections through the data matrix. in order to pass a s close as possible to the wavelength maxiinurn of each analyte avoiding absorption regions of the others in order to optimize both resolution and sensitivity. Trajectories can be defined, through the contour plot, by the initial and final coordinate (A, t ) pairs. In this work, two trajectories were performed to establish the corresponding cross-sections (Fig.3). In order to select the first linear path, four cross-sections were tested. Their initial coordinates (h, t ) are 200 nm, in the wavelength domain, and in the time domain they are 700,725,750 and 760 s (lines a, b, c and d, respectively, in Fig. 3); the final coordinates (A. t ) are (240. 900) in all cases. 0.60 r\ Wavelengthhm Fig. 2 of folpet and (3) 6 pg nil Absorption spectra of ( I ) 5 pg nil I ofprocymidone, (2) 3 pg inl- I ot triazophos. In the second path the initial and final coordinates (A, t ) are (240, The two-dimensional projections on the wavelength domain, generated by the selected cross-sections through the data matrix, are represented in Fig. 4. The analytical signals obtained 900)-(280, 500).900 800 C 700 d k 600 (I) --- Em i= 400 300 200 100 0 2 -_I-- ---~-- _--.--T . __ - 0 210 220 230 240 250 260 270 : 10 Wavelengthhm Fig. 3 Contour plot ot ( I ) folpcl. ( 2 ) procymidone and ( 3 ) triaiopho\ 'it conceiitiation\ ot 9, 4 and 6 lig nil- 1 , ie\pectively. where the foul trajectories wlected (a, b, c and d) in the f m t linear path optimization of the cro\a-section die plotted. 0.1 2 7- I 1 1 ,- 1 - : . I I T - 2 T 1 I I , ---- r - , ~---i ,__. 1 L - 7---- r T- , ~- 7- 4 200 220 240 260 280 200 220 240 260 280 Waveleng thhm Fig. 4 Two-dimensional projections of the cross-sections produced from the three-dimensional data by plotting absorbance iscr,\us wavelength: ( a ) trajectory a in Fig. 3, ( h ) trajectory b, ( c ) trajectory c and (0) trajectory d.Numbers above the peaks correspond to (1) folpet, ( 2 ) procymidone and (3) triazophos.1370 Analyst, Octohci- 1996, Vol. I2 I after this process are very different from those of the original chromatograms. The four cross-sections tested were selected in order to obtain two-dimensional projections with the best analytical character- istics (resolution and/or sensitivity). It is evident that the four trajectories selected are not the only possibilities. More complicated trajectories, with more than two linear paths, or even non-linear paths. niay be selected for the analysis. In the optimization of the first linear path, triazophos is separated from the other analytes whilst the resolutions between folpet and procymidone are for trajectory a 0.9, b 0.9, c 1.2 and d 1.1.We decided to use the two-dimensional projection obtained from trajectory c as it gave good resolution and sensitivity for the three compounds. In the selection of the second linear path. the initial coordinates (A, t ) are (240, 900) and to select the final coordinates (A, r ) the wavelength values used are 250, 260 and 280 nm, while the time is 500 s in all cases. In Fig. 5 are shown the- trajectories of the selected cross-sections in order to optimize the sensitivity of triazophos. The trajectories tested have little influence in the sensitivity of triazophos (Fig. 6), but in the two-dimensional projection corresponding to Fig. 6(c) the interference due to the peak that appears close to the triazophos peak is avoided. We selected the trajectories defined for the coordinates (A.t ) (200, 750)-(240, 900) for the first path and (240, 900)-(280, SOO) for the second path. In Fig. 7 is presented the isometric projection of the complete spectrochromatogram of the mixture analysed, in which the trajectory of the optimiLed cross-section is marked. In this way, the resolution of the mixture is accomplished, allowing the quantification of each of the arialytes through the adequate calibration lines. 900 800 700 600 ! 5oa i= 4oa 3OC 20c 1 oc c L 10 2 i o 220 230 2 i o 2jo 260 50 : Wavelengthhm 3 Calibration graphs Calibration graphs were obtained from peak heights of two- dimensimal projections for samples of mixture\ of the three compounds, containing different concentrations of folpet, procymidone and triazophos.Good linearity was obtained for all pesticides in the 1.0-10.0 pg ml-I range. Table 2 lists the straight-line equations obtained for the concentration intervals tested and the corresponding statistical parameter values obtained without replicating the experimental points. In order to study the repetitivity of the method, a series of six solutions were prepared, containing 2.0 pg ml- I of folpet, 2.0 pg ml- 1 of procymidone and 2.0 pg ml-l of triazophos, with results of 3.8, 4.5 and 3.l%, respectively, for the RSDs. The values obtained show the high repetitivity of the method. Resolution of Synthetic Ternary Mixtures To validate the method, mixtures of folpet, procymidone and triazophos, in the concentration range I .O-10.0 pg ml- for each pesticide, were prepared, and chromatograms were recorded according to the described procedure.Tablc 3 presents the results of the analysis of different mixtures. Satisfactory 0.08 a, 0.06 0 a 0.04 s: a 0.02 D 0.00 0.08 7- 0 0.00 - -7- - -7- 7 3 220 240 2 3 1 I!), I 0 7 1 + ,.--, - - , - ~ , ~ - - 200 220 240 2 0 280 Wavelengthhm Fig. 6 Two-dimensional projection of the cross-sections produced from the three-diitiensional data by plotting absorbance i'oi-xu.s wavelength: ( ( 1 ) trajectory a i n Fig. 5. (h) trajectory b and ( c ) trajectory c. Numbers above the peaks correspond to ( 1 ) folpel, (2) procymidone and (3) triazophos. Fig. 5 The three tni-jectories selected (a, b and c) in the second linear path optimization of the cross-section, plotted across the contour plot, of the spectrochromatogram of a mixture containing 9 pg nil-' of folpet ( l ) , 4 pg ml- I of procymidone (2) and 6 pg m-1 of triazophos ( 3 ) .Fig. 7 Isometric projection of the spectrochrornatoFrani of the mixture analysed, in which the trajectory of the selected cross-section c is marked.Analyst, October 1996, Vol. 121 1371 Table 2 Calibration graphs for the determination of folpet, procymidone and triazophos by measuring the peak heights at the selected projection of the cross- section Standard Standard error Pesticide Equation* r2 deviation of estimate Folpet y = 1.6154 x 10-3 + 6.6993 X 10V x 0.9973 0.02235 0.009 12 Procymidone y = 1.6105 X lop3+ 1.0103 X lop2 x 0.9973 0.03370 0.01376 Triazophos 4’ = -1.44.52 X 10-’+3.2468 X 10-3 x 0.9993 0.01081 0.00441 * x = Concentration of pesticide in pg ml-1; y = peak height.The calibration graphs were obtained from six experimental points. Table 3 Mean recoveries and RSDs ( n = 5 ) of folpet, procymidone and triazophos in synthetic ternary mixtures Folpet Procymidone Triazophos Recovery RSD Recovery RSD Recovery RSD c* (%) (%) c* (%) (%) c* (%) (% 1 3 91:O 6.2 I 95.8 4.9 4 90.1 6.4 4 105.3 4.4 1 106.3 5.2 7 98.3 5.4 10 106.0 4.8 10 104.9 5.0 10 103.7 5.6 8 96.5 4.5 8 108.4 4.5 8 107.2 3.9 8 90.7 5.0 6 93.5 3.8 7 108.5 5.4 * C = Concentration of pesticide added (pg ml-1). results were obtained, with recoveries ranging from 90.7 to 106.0% for folpet, from 93.5 to 108.4% for procymidone and from 90.1 to 108.5% for triazophos. The results indicate that the complete resolution of the mixture has been accomplished by the proposed approach, showing the high resolving power of the technique.Preconcentration of Pesticides in Water by Solid-phase Extraction (SPE) The proposed method was applied in the determination of pesticides in environmental water samples. A trace enrichment step is necessary to obtain detection limits as low as ppb levels. To evaluate the potential of trace enrichment of the pesticides, samples of ultra-pure water, spiked with 3 pg 1-1 of pesticides, were analysed. The 360 mg Sep-Pak C18 cartridges were conditioned with 5 ml of ACN followed by 5 ml of ultra-pure water without allowing the cartridges to dry out. Water samples of 400 ml were passed through a 0.4 pm filter, connected by PTFE tubes to the conditioned cartridges, at a rate of 8-10 ml min-1; the cartridges were then sucked dry for 5 min. ACN was chosen as solvent for the elution of analytes owing to its suitability for the RP-HPLC system and, finally, 20 p1 were injected.Good linearity was obtained for all substances in the ranges studied (3-11 pg 1-1). The regression coefficients are higher than 0.991 in all cases (M = 7). The detection limits,29 calculated statistically, are 0.3 1,0.29 and 0.43 pg 1- for folpet, proc y midone and tri azophos, respectively . The mean recoveries of the pesticides were 101,98 and 85% for folpet, procymidone and triazophos, respectively. The repeatability in terms of peak height at various concentrations was studied using the conditions described above. The data obtained for 3 pg 1-l indicate that the RSD ranged from 5.5% (triazophos) to 8.4% (procymidone).With groundwaters spiked at a level of 3 pg 1-1, the recoveries were 91, 85 and 83% for folpet, procymidone and triazophos, respectively, and the RSDs were 8.3. 9.1 and 8.7%, respectively. A blank of water without fortification was also analysed in each experiment. The proposed method was applied to the determination of pesticide levels in ground waters from Almeria (Spain) and the chromatograms obtained showed no peaks of the studied pesticides. Conclusions The determination of folpet, procymidone and triazophos mixtures was performed by means of the proposed technique with good repetitivity and sensitivity. The technique is particularly useful for analysing mixtures of analytes in complex samples, as is the case in MR pesticide analysis.The usefulness of the proposed methodology is the resolution of overlapping chromatographic peaks maintaining, at the same time, as much sensitivity in the determination as possible. In addition, the approach would allow a decrease in the time of analysis in certain cases. This can be the case in the separation of several analytes with similar polarities from one with a very different polarity. The method has been applied to the determination of folpet, procymidone and triazophos in water samples at ppb levels with good results. In conclusion, the combination of advanced computational capability with the DAD technology applied in HPLC offers a powerful approach for the resolution of highly overlapping peaks.The authors are grateful to DGICYT (Project PB95-1226) for their financial support. References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Edgell, K. W., Erb, E. J., Wesselman, R. J., and Longbottom, J. E., J . AOAC Inf., 1993, 76, 1098. Hernindez, F., Morell, I., Beltran, J., and Lhpez, F. J., Chronzatogra- phia, 1993, 37, 303. Hong, J., Eo, Y., Rhee, J., Kim, T., and Kim. K., J . Chromatogr., 1993,639,261. Edgell, K. W., Erb, E. J., Longbottom, J. E., and L6pez-Avila, V., J . AOAC Int., 1992, 75, 858. de la Colina, C., Baez, M. E., Peiia, A., Romero, E., Dios, G., and Sinchez-Rasero, F., Sci. Total Environ., 1994, 153, 1. Huen, J. M., Gillard, R., Mayer, A. G., Baltensperger, B., and Kern, H., Fresenius’ J . Anal. Chem., 1994, 348, 606. Slobodnik, J., Groenewegen, M.G. M., Brouwer, E. R., Lingeman, H., and Brinkman, U. A. Th., J . Chronzatogr., 1993, 642, 359. Fell, A. F., Clark, B. J., and Scott, H. P., J . Chromatogr., 1984, 316, 423. Gluckman, J. C., Shelly, D. C., and Novotny, M. V., Anal. Chern., 1985, 57, 1546. Wegrzyn, J., Patonay, G., Ford, M., and Warner, I., Anal. Chern., 1990,62, 1754. Garrido Frenich, A., Martinez Galera, M., Gil Garcia, M. D., and Martinez Vidal, J. L., J . Chmmatogr., 1996, 727, 27. Martinez Galera, M., Martinez Vidal, J. L., Garrido Frenich, A., and Gil Garcia, M. D., J . Chronzatogr , 1996, 727, 39. Garrido Frenich, A., Martinez Galera, M., Gil Garcia, M. D., Martinez Vidal, J. L., Mufioz de la Peiia, A., and Salinas, F., J . Chromatogr., submitted for publication. Clark, B. J., Fell, A. F., Scott, H. P., and Westerlund, D. J., J Chromatogr., 1984, 286, 261. Clark, B. J., and Fell, A. F., Cheni. Br., 1987, 23, 1069. Fasanmade, A. A., Fell, A. F., and Scott, H. P., Anal. Chim. Acta, 1986, 187, 233. Fasanmade, A. A., and Fell, A. F., Anal. Chern., 1989, 61, 720. Muiioz de la Peiia, A., Salinas, F., Galeano, T., and Guiberteau, A., Anal. Chim. Acta, 1990, 234, 263. Parrilla, P., Martinez Galera, M., Martinez Vidal, J. L., and Garrido Frenich, A., Analyst, 1994, 119, 2231.1372 Analyst, October 1996, Vol. I21 20 21 22 23 24 25 Dimuccio, A., Girolimetti, S., Ausili, A., Ventriglia, M., Generali, T., and Vergori, L., J . Chromatogr., 1993, 643, 363. Holland. T. P., Naughton, E. D., and Malcolm, P. C., J. AOAC Int., 1994, 77, 79. Ogawa, M., Ohtsubo, T., Tsuda, S., and Tsuji, K., J . AOAC Int., 1993, 76, 83. Liao, W., Joe, T., and Cusick, G., J . AOAC Int., 1991, 74, 554. Parrilla, P., Martinez Vidal, J. L., and Fernandez Alba, A. R., J . Liq. Chromutogr., 1993, 16, 4019. Parrilla, P., Martinez Vidal, J. L., Martinez Galera, M., and Frenich, A. G., Fwseniub’ J . Anal. Chern., 1994, 350, 633. 26 27 28 29 Bellar, T. A., and Budde, W. L., Anal. Chem., 1988, 60, 2076. GRAPHER and SURFER for Windows Software Package Version 5.0, Golden Software, CO, 1994. Martinez Vidal, J. L., Parrilla, P., Fernandez Alba, A. R., Carreiio, R., and Herrera, F., J . Liq. Chromatogr., 1995, 18, 2969. Long, G. L., and Winefordner, J. D., Anal. Chem., 1983, 55, 713. Paper 6102345B Received April 3, I996 Accepted June 7, 1996
ISSN:0003-2654
DOI:10.1039/AN9962101367
出版商:RSC
年代:1996
数据来源: RSC
|
12. |
Study of the interaction of a soil fulvic acid with UO22+by self-modelling mixture analysis of synchronous molecular fluorescence spectra |
|
Analyst,
Volume 121,
Issue 10,
1996,
Page 1373-1379
Joaquim C. G. Esteves da Silva,
Preview
|
PDF (993KB)
|
|
摘要:
Analyst, October 1996, Vol. 121 (1373-1379) 1373 Study of the Interaction of a Soil Fulvic Acid With U022+ by Self-modelling Mixture Analysis of Synchronous Molecular Fluorescence Spectra Joaquim C. G. Esteves da Silva, Adelio A. S. C. Machado* and Cesar J. S. Oliveira LAQUIPAI, Fuculdude de CiCncias, Alegre 687, P41.50 Poi-to, Portugal The interactions between U022+ and fulvic acids (FUA) were studied by a methodology that involves synchronous molecular fluorescence spectroscopy, to monitor the quenching of the intrinsic fluorescence of FUA by UOZ2+, and a self-modelling mixture analysis method (SIMPLISMA), to treat spectroscopic data. This methodology was applied to the analysis of the interaction of U022+ with salicylic acid at pH 3.5 and with a soil FUA at pH 3.5 and 7.0, in this case in the presence of various concentrations of carbonate ion ( l O - 5 , l O - 4 and 10-3 mol 1-I).From the calculated quenching fluorescence intensity profiles, using either the Stern-Volmer relationship or a non-linear least-squares method, mean conditional stability constants (log values with standard deviations in parentheses) were estimated for salicylic acid [respectively 2.72(4) and 2.77(6)], for FUA at pH 3.5 [3.93(2) and 4.4(1)] and pH 7.0 [4.06 and 4.1 (average values for the various concentrations of carbonate ion)]. The non-linear least-squares method also allowed the estimation of the number of binding sites that exist in FUA (0.11 and 0.24 mol g-’ at pH 3.5 and 7.0, respectively). Keywords: Fulvic acids; uranyl ion; soils; synchronous molecular fluorescence; self-modelling mixture analysis Introduction The effect of humic substances (HUS) on the environmental speciation of uranium and other actinides has been studied for assessing the safety of radioactive waste repositories and deactivated or open mines.’-lO Owing to their natural ubiquity and to the existence in their molecules of relatively large amounts of carboxylic and hydroxylic structures, which are hard oxygen donors for metal ion coordination, HUS may play an important role in the environmental migration of actinides. In work 1 1-15 to obtain quantitative information about the coordination of metal ions by humic substances using spectro- scopic techniques, it was observed that the U0Z2+ion induces significant quenching of the molecular fluorescence of the most soluble fraction of HUS, fulvic acids (FUAj.This effect is similar to that observed for other paramagnetic ions, “ 3 , l4,I6- 2o which allowed the calculation of quantitative information about the complexation reactions for some, particularly Cu”. The quenching of the fluorescence of ligands by metal ions as a consequence of complex formation (static quenching) is the basis of a recently developed methodology for the analysis of equilibria between FUA and metal ions.” This methodology is based on the use of (i) the synchronous fluorescent mode for collecting spectra of FUA solutions during titrations with the metal ion, (ii) SIMPLISMA,21>22 a self-modelling mixture analysis method, for analysis of spectral data and (iii) a non- linear least-squares procedure, for the estimation of equilibrium parameters.The use of a self-modelling mixture analysis method as a pre-processing step before the analysis of the quenching profiles is important in the case of complex mixtures such as FUA, because different spectral variations may exist in the raw data and need to be resolved before adjusting the experimental data to a simple model. The objective of this work was to investigate the usefulness of the above methodology for the study of the interactions between a FUA extracted from the lower soil horizon of a pinewood soil and U022+ at pH 3.5, where the formation of U022+hydrolysis products is greatly reduced, and at pH 7, which is a representative pH for natural waters, at several concentrations of carbonate ion (from the environmental point of view this is the most important inorganic ligand for U022+ apart from hydroxide).2.23 To assess the experimental and data analysis methodology, a preliminary study of the interaction of with salicylic acid at pH 3.5 was also carried out.This ligand was chosen because it forms stable complexes with U022+ and is usually considered as a model structure for the binding of metal ions to FUA. uo22+ - Theory FUA are complex mixtures of macromolecules with different constitutions. The spectrum of these substances will always be that of a mixture with superimposition of bands, even if the most selective analytical technique is used. Any improvement on the purity of the analytical signal, i.e., an experimental signal proportional to the amount of a single species, will result in an over-all improvement of the detection methodology. In the context of a set of species with similar chemical properties such as FUA, binding sites with similar complexation properties may be detected.This increase in purity is particularly useful when relatively simple models, such as those described below, are used. The analytical methodology used in this work was intended to achieve the highest possible purity at two different levels: (ij the experimental methodology is the synchronous mode of molecular fluorescence, which is a very selective technique,24.25 and also, as this technique measures only the fluorescent fractions of FUA which contain the most reactive structures, other structures, i.e., part of the structures of FUA, do not interfere in the measured signal; and (ii) the spectroscopic signals are subjected to treatment by a self-modelling mixture analysis method to resolve, as far as possible, the raw data into the pure components.Self-modelling Mixture Analysis Method (SIMPLISMA) The first step of the SIMPLISMA procedure is the determination of the number of components of the system under1374 Analyst, October 1996, Vol. 121 analysis (represented by a data matrix D , size nv X ns, where FZV is the number of variables (wavelengths) and ns is the number of spectra). This is done by a ‘user-friendly’ graphic interface based on three spectra: mean spectrum, n?; standard deviation (SD) spectrum, s; and, purity spectrum, p . The elements of the mean spectrum are defined by ( i = 1 .. . nv) (1) mi = (l/ns) x(j = I . . . ri.7)di.l those of the SD spectrum by si = [(l/m) C,j= I _ _ .,,,,) and those of the purity spectrum by pI = si/rni (i = 1 . . .mi) (3) The visualization of vectors s and p in the form of a spectrum allows the detection of variables with the highest pure character (pure variables). The variations associated with the selected first pure variable are then removed from the SD and purity spectra and the procedure is repeated to determine the second pure variable, and so on. This process is repeated until the mean and SD spectra show only noise, i.e., when all the useful information has been removed from the data. Two error functions, R, (relative total intensity of the SD spectra) and R, (ratio of the relative total intensities of thei SD and thej + 1 SD spectra), can be used as a rough indication of the number of components: (4) ( 5 ) The second step of the SIMPLTSMA procedure is the resolution of the raw data into spectra (matrix S, size np X nv) and concentration profiles (matrix C , size ns X np) for the detected species.Assuming that all the components have the same fluorescence efficiency, the concentration profiles are equal to the fluorescence intensities of the correspondent pure variables after normalization. If the data matrix is expressed as D T = c s (6) the spectra and concentration profiles can be calculated by a least-squares procedure: (7) s = (ci‘ c)-i CI D r c = D1’ s.1’ (S ST)-’ (8) In eqn. (7), the intensities of the pure variables in the D spectra matrix are used in the C matrix.- mi)2]I/’ (i = 1 . . .in,) (2) R,, 100 x(i = I . . . n i ’ ) si.j/z(i = I . . . I l l . ) si.1 Rrj = RsjIRsci + 1 ) Analysis of Quenching Profiles Stei-n-Volniei- unalysis When a ligand (L, fluorophore) shows fluorescence, the following processes occur: L + hv + L* (9a) (9b) L* + L (9c) L* + L + hv where eqn. (9a) represents the absorption of radiation, eqn. (9b) the emission of radiation and eqn. (9c) a non-radiative decay. If U022+ is present and forms no fluorescent complexes with L (static quenching), the following reaction occurs (charges of the ligand and complex are not shown for simplicity), with conditional stability constant K’: ( 9 4 The quenching of the ligand fluorescence may also be due to a dynamic process: L* + UO*” ---$ L + U 0 2 2 + (9e) L + UO’” e [(UO*)L] In this work, as the effect of FUA on the U022+ environmental speciation is due to complexation and both FUA and salicylic acid form stable complexes with U022+4,6,26 [eqn.(9d)], our interest is focused on static quenching, but dynamic quenching must be considered to evaluate its influence on the data. Tn the absence of dynamic quenching, if the fluorescence intensity of the ligand solution is measured (lo) together with the fluorescence intensity of mixtures of the ligand with increasing amounts of UOZ2+ (f), the Stern-Volmer relationship for static quenching is ~ b t a i n e d : ~ ~ - ~ ~ (10) If only static quenching is observed, the plot of Zo/Z as function of the total metal concentration (CLo2) allows the calculation of K‘.If both static and dynamic quenching are present, a plot with upward curvature will result.27-29 An estimation of K’ and of the Stern-Volmer constant for dynamic quenching (Ksv) can be obtained from the following rearranged Stern-Volmer equa- ( 1 1) If [(Ioll - 1)/Cuo2 versus Cuo2 is linear, K,, and K’ can be calculated from K,, + K’ = intercept and K,,K’ = slope. On the other hand, a plot of Zo/Z versus CUo2 [eqn. (lo)] that deviates from linearity towards the x axis indicates the existence of two or more binding structures (fluorophores), not equally acces- sible to the complexation of U022+.2030 loll = 1 + K’ CUO’ t ion: 27-29 [(W) - 1 IlCcro’ = K” f K’) + K,, K’ Cuo2 Non-linear least-squares analysis For poorly defined ligands such as FUA, the method used for the calculation of the conditional stability constants from the observed quenching of fluorescence due to a static mechanism should also provide an estimation of the concentration of the ligands.The fraction of the total FUA bound ( = [UO2L]/C,) is given by 16-I* (12) where CL is the total ligand concentration and fUozL is the fluorescence intensity due to the bound ligand. After setting Zo = 100, the following relationship between the synchronous fluorescence SyF intensity profiles Z and K’, Zuo2L, CL and CM, is 0btained13.1~7’6-18 [U02LI/CL = (10 - M I 0 - IU02L) I = [I”fJ2L - 100/(2K’CL)] { (K’CL + K’CU02 + 1) - [(K’CL + K’CUO2 + 1)’ - 4K’2CLCU02]1’2) + 100 (13) For the cases where experimental evidence suggests that the complex formed between L and U022+is not fluorescent (IUozL = 0), this equation simplifies to I = 100 - [lOO/(ZK’CL)] {(K’CL + K’CUQ + 1 ) - [(K’CL + K’CUo2 + 1)2 - 4K’2CLCuo,] ”* } ( 14) Eqns.(1 3) and (14) can be solved for K’, CL and Zuo2L by non- linear regression analysis, using the concentration profiles calculated with STMPLTSMA, using a procedure described previously.13,14 The quality of the non-linear adjustment of the quenching fluorescence intensity data for the calculation of K and the concentration of the binding site is assessed by the analysis of two error functions, the sum of the squares of the residual (SSR) and the average deviations (AD); SSR = c ( L x p - lcalc)2 (1 5a) (1 5b) where the summations are over the total number of points used in the calculations ( N J , Zexp is the experimental fluorescence intensity and Zcalc is the calculated fluorescence intensity.AD = (Z I L x p - L l ‘ I)/&Analyst, October 1996, Vol. 121 1375 Experimental Reagents Analytical-reagent grade reagents were used. The extraction of FUA and its characteristics were described previously. 12715 Aqueous solutions of 100 mg 1 - I FUA and 0.3 mmol 1-1 salicylic acid were prepared in 0.1 mol 1-1 potassium nitrate solution. For adjusting the titrated solutions to a constant pH value, 0.04 mol 1-' decarbonated potassium hydroxide solu- tions were used. Aqueous solutions of U0l2+ were prepared by dissolving 0.050 g of U02(N0T)2-6H20 in 0.1 mol 1--' nitric acid (approximately 10 nil) and diluting with water to 100.0 ml (the pH of the final solution was 2).Procedures pH adjustments, titrations and SyF measurements (a 20 nm wavelength increment was used) were made at 25.0 "C as described previously. x13 The ranges of UO22+ concentration in the titration vessel were 0.008-1.1 mmol 1-1 for the salicylic acid experiments, 0.004-0.2 mmol 1- I for the FUA experiments at pH 3.5 and 0.001-0.2 mniol 1-1 for the FUA experiments at pH 7.0. At these concentration levels, and for the instrumental settings used (slits and synchronous wavelength increment), no U02*+ fluorescence was detected and consequently no inter- ference occurred with the measurements of the fluorescence of the ligands. Programs and Data Treatment All data analysis (SIMPLISMA and equilibrium calculations) and data simulation software were developed in this laboratory and written and compiled with Turbo Pascal 5.0 (Borland International, Scotts Valley, USA).Results and Discussion Salicylic Acid Data Fig. l(u) shows the effect of the presence of U0Z2+ on the SyF spectra of salicylic acid. The decrease in fluorescence intensity with increase in U02*+concentration at pM 3.5 shows that quenching occurs. The shape of the spectra in Fig. l ( a ) is identical with that of the spectrum of the hydrogensalicilate species (LH-), which is the predominant form at pH 3.S.31 SvF spccti-al dutn amlysis The SD spectra resulting from the SlMPLlSMA analysis of the salicylic acid data set are shown in Fig. 2(a)-(c). The first SD spectrum [Fig. 2(a)] is identical with the spectrum of the LH- species of salicylic acid, suggesting that the main spectral variation is due to the over-all quenching of this spectrum.The analysis of the second and third SD spectra only shows the existence of a relatively high noise band in the lower wavelength range. Therefore, the SIMPLISMA analysis of the salicylic acid spectral data shows that only one component is sufficient to account for all the observed spectral variation besides experimental noise. The values of the error functions in Table 1 support this conclusion. Indeed, the K, parameter of the second component is small whereas the R, parameter of the first component is relatively large. I ~ ( ~ 2 1 270 320 370 420 470 520 270 320 370 420 470 520 270 320 370 420 470 520 1.46E-02 h .2 0.00 L 3 270 320 370 420 470 520 270 320 370 420 470 520 t 270 320 370 420 470 520 270 320 370 420 470 520 + - 6.00E-05 3.00E-05 O.OOE+OO 270 320 370 420 470 520 270 320 370 420 470 520 1.30E-02 2.00E-09 1.00E-09 O.OOE + 00 0.WE+00 270 320 370 420 470 520 270 320 370 420 470 520 Wavelengthhm Fig. 2 Standard deviation spectra re\ulting from the SlMPLlSMA analysis: (u)--(c) Salicylic acid; (d)-(g) FUA (pH 3.5): and ( h - ( k ) FUA (pH 7.0). Wavelengthhm Fig. 1 Synchronous fluorescence spectra of (u) salicylic and (h) and ( c ) FUA [(h) pH 3.5; and ( 1 . ) pH 7.01 as a function of U02>+ concentration.1376 Analyst, Octohei- 1996, Vcd. 121 The result of the second step of SlMPLISMA, i.e., the calculation of the spectrum and fluorescence intensity (quench- ing) profile of the detected component, is shown in Figs.3(a) and 4(a), respectively. The spectrum in Fig. 3(a) (see also Table 2) is that of salicylic acid at pH 3.5.31 This result shows that the only spectral variation observed when UO22+ is added is due to a decrease of the amount of free salicylic acid. Stern-Volnzei- analysis o j the quenching The Stern-Volmer plot [eqn. (1 O)] for the quenching of salicylic acid by U022+ is linear. The parameters from the least-squares adjustment of /o// versus Cuo2 are given in Table 3. The correlation coefficients of the plots obtained in repeated experiments were always larger than 0.9990 and the intercepts were close to the theoretical value of 1.0 expected from the Stern-Volmer equation. The average log K’ value obtained from the slopes of five plots was = 2.72(4).The absence of upward curvature suggests that static quenching occurs between salicylic acid and U0l2+, which fonn stable complexes.26 The absence of downward curvature confirms that there is only one type of binding site. Using the data in Table 4, which describe the most important equilibria in aqueous acidic solutions of mixtures of salicylic acid and U022+-2632 a conditional stability constant of log K’ = 2.52 was calculated for pH 3.5 and typical concentrations of salicylic acid and U0Z2+ used in this work, 0.6 and 0.3 Table 1 Typical values of the SlMPLISMA error functions for the salicylic acid and FUA quenching experiments Salicylic acid FUA QH 3.5 pH 3.5 pH 7.0 N- R, R, R, R, R, Kr I 100.0 396 100.0 54 100.0 294 2 0.3 186 1.8 3337 0.3 1117 3 1.4 X 5.5 X 10-4 2538 3.0 X 10-4 2286 4 2.2 x 10-7 1.3 x 10-7 * Number of the pure variable.mmol 1-1, respectively. Owing to the limitations of these calculations, this value can be considered similar to the experimental value of 2.72(4). In conclusion, the present analysis shows that only static quenching occurs owing to the formation of a complex between salicylic acid and U022+. Analysis of the quenching by the non-linear. leasr-squares method To assess the performance of the non-linear least-squares method for the calculation of the conditional equilibrium parameters, the quenching profiles were also treated by this procedure. A first attempt to estimate the three unknown parameters of eqn. (13) ( K , CL and IuOzL), failed to provide a value for CL (concentration of the experimental salicylic acid solution) but gave good estimates for K and for /l,02L ( I ~ ~ 0 2 L ) .As the complex is not fluorescent, in a second attempt to obtain an estimate for C12 eqn. (14) was used. The results, presented in Table 5 , show that no reasonable value could be obtained for CL. Indeed, values for this parameter in the range 1 O F - 10-3 mol 1-1 were found. However, this failure did not affect the values of the other calculated parameters. The conditional stability constant, log K = 2.77(6), is similar to that obtained from eqns. (13) and to the value calculated from the Stern- Volmer equation. The analysis of the error functions, SSK and AD, shows that the adjustment of the experimental data to eqns. (13) and (14) is fairly good, as indicated by the low AD obtained [typical values in the 0.2 units range (Table 5 ) ] .These results show that, although the calculation of the concentration of the ligand is impossible, the method provides adequate values of the conditional stability constants. Indeed, reasonable estimates of CL can only be obtained by this method if the magnitude of the stability constant is higher than in the present situation.1”16-1* Fulvic Acid Data Fig. l(b) and ( c ) show the effect of the presence ofU022+ on the SyF spectra of FUA. Although the spectra at pH 3.5 and 7.0 differ, because their shape is pH dependent,33 quenching is detected at both values. 270 320 370 420 470 520 - cn w .- + -\ / \ 0.50 4 / 0.00 1 0.00 4-4- 300 350 400 450 500 550 300 350 400 450 500 550 Wave1 engthhm Fig.3 Calculated spectra of the ( a ) salicylic acid and (h) and (d) first and (c.) and (e) second FUA components; (h) and (c) pH 3.5; and (d) and ( e ) pH 7.0.Analyst, October 1996, Vol. I 2 I I377 SyF specti-ccl dutu analysis The SD spectra from the SIMPLTSMA analysis of the FUA spectral sets at the two pH values are shown in Fig. 2(d)-(k). Their analysis indicates that the number of components showing linear independent variations is two for both data sets. Indeed, for the pH 3.5 experiment, only the first two SD spectra [Fig. 2(d) and ( e ) ] contain signals, the others [Fig. 2 0 and ( g ) ] containing only noise. For the pH 7.0 experiment, the I I t I 0.0 -I I I 0 5 10 15 20 25 a, 20.0 0.0 ; I : ; : ; : 0 5 10 15 20 25 20-o 0.0 t 0 5 10 15 20 25 Spectrum No. Fig.4 Calculated tluorescence intensity profiles for the (a) first and (m) second components: ( a ) salicylic acid; (h) FUA (pH 3.5); and (c) FUA (pH 7.0). observation of the SD spectra [Fig. 2(h)-(k)] allows no clear attribution of the number of components. Indeed, there is a high noise band in the lower wavelength range [Fig. 2(i)] that masks the detection of the second component. However, the analysis of the error function, shown in Table 1, provides evidence for the existence of two components. Indeed, after selection of the second pure variable, there is a marked increase in the R, parameter. However, the identifica- tion of the second component in the pH 7.0 experiment is not as sound as that for pH 3.5. The calculated spectra and quenching profiles of the two components detected are shown in Figs.3(b)-(e) and 4(h) and (c), respectively. Table 2 shows the position of the main bands of the spectra. The comparative analysis of the calculated spectra of the first component at the two pH values shows the existence of similarities. Indeed, the main band of both spectra is located in the 470 nm range, suggesting that U022+ is inducing the quenching of fluorescence of similar structures at the two pH values. The calculated quenching profiles shown in Fig. 4(h) and (c) are similar to that of salicylic acid, have the expected shape of a decreasing curve and show that the quenching for the first component is larger than for the second. This result is expected because SIMPLISMA detects first the larger relative spectral variations.For the pH 7.0 experiments, the quenching profiles are similar. This situation is compatible with the above- discussed doubts regarding the selection of the number of components. Stern-Volmel- analysis o j the quenching The Stern-Volmer plots for the quenching profiles of the two components of the pH 3.5 experiments (shown in Fig. 5 ) and of the first component at pH 7.0 (the second component at this pH is almost identical with the first) have different characteristics. Table 2 Positions of the SyF bands o f the components detected in salicylic acid and FUA quenching experiments Position of main bands1 pH Component nin Sulic~lic ucili- 3.5 1 340 FUA- 3.5 1 465, 360 2 145,390 7.0 I 470, 455, 440 2 390 Table 3 Results of the linear Stern-Volmer plots for the quenching induced by U022+ on the components detected by SIMPLISMA for salicylic acid and FUA PH Nc ~ I C 0 ~ 2 - 1 N, Log K’ Intercept N,, I’ AIU02*+1 3.5 1 - 4 2.72(4) 1.01 8(7) 28 0.9993 0.008-1.1 3.5 2 - 3 3.93(2) 0.95( 1 ) 27 0.9992 0.002-0.3 7.0 I 5 3 4.06(1) I .02( 1) 26 0.9995 0.001-0.2 7.0 1 4 5 4.03(3) 0.99(2) 24 0.9992 0.001-0.2 7.0 1 3 3 4.09(2) 0.983 1) 25 0.9994 0.001-0.4 Scrlic ylic mid- FUA- ’ pH = pH al which the experiment was performed; N, = numbcr of the component; P / C O ~ ~ - I = antilogarithm of the concentration of the carbonate anion; N, = number of independent experiments used in calculations; K’ = conditional stability constant; N, = number of points used in calculations; I’ = correlation coefficient; A I U022+ I = U022+ concentration range used in calculations (mmol 1-1).Average and standard deviations in brackets for log K and intercept, and typical values for the other parameters.1378 Aiwlyst, October- 1496. Vol. 121 Indeed, whereas the plot for the first component at pH 3.5 shows marked upward curvature, the other two plots are linear with a quality of the adjustment similar to that for salicylic acid. These results show that the two components detected by SIMPLISMA for the experiments at pH 3.5 have different quenching mechanisms. Indeed, the upward curvature of the Stern-Volmer plot of the first Component suggests that both static and dynamic quenching is being observed,27-29 whereas the linear plot for the second component shows the existence of static quenching only, since stable non-fluorescent complexes are formed.To obtain further information about the quenching observed for the first component at pH 3.5, the quenching profiles were analysed using eqn. ( 1 1). A linear plot was obtained although the quality of the adjustment was worse than for the previous cases (a lower correlation coefficient, 0.997, and a narrower U022+ concentration range of linear variation were observed between 0.02-0.20 mmol I-’). This suggests that the static and dynamic quenching are probably occurring at pH 3.5 and that the first quenching profile calculated by SIMPLISMA, corre- sponds to the cumulative effect of both. However, no physical meaningful values could be obtained from these plots. This is probably a consequence of the complex photochemistry of both U022+ and FUA and to the oversimplified model implicit in eqn.( 1 1). As no downward curvature was observed in the Stern- Volmer plots at pH 3.5 and 7.0, only one type of binding site structure in the molecules of FUA participates in the complexa- tion reaction in each case.20 In the pH 7.0 experiments, no marked variation of the estimated conditional stability constant is observed for the three levels of carbonate concentration (Table 3) and an average of log K’ = 4.06 is obtained. At pH 7.0 the carbonate does not compete with the hydroxide and FUA for the complexation of U022+, because it is almost completely converted into hydro- gencarbonate, and the stabilities of the complexes formed with the other two ligands are large. The estimated log K’ at pH 3.5 Table 4 Major chemical species in mixtures of salicylate (IL2-) and UOz2+ in water and corresponding stability constants ( K ) at 25 “C and ionic strength 0.I mol 1- 2h.i2 Reaction UOJ’+ + H20 g [(UOz)(OH)]+ + H+ 2UO2” + 2F-120 g [(UO2)2(OH)l]*+ + 2H’ 3UOl’+ + SH’O e [(UO,),(OH),j+ + 5H+ L2- + H+ ;t (HL)- L’- + 2H+ $ (HlL) U02?+ + L’- - - I(UO2)I~I UO?’ + L’p + H+ e [(UO?) (HL)J+ Log K -S.95 -5.79 -16.1s 13.0 15.72 12.04 1 14.6 [3.93(2)1 is smaller than at pH 7.0 (about 4.06) and both are larger than that for salicylic acid [2.72(4)]. Analysis o j the queiwlzing by the non-liiieai- least-squares method Table 5 shows the results from the nun-linear adjustment to eqn. (12) to the FUA quenching profiles of the components that originated linear Stern-Volmer plots, namely the second component of the pH 3.5 and the first of the pH 7.0 experiments.The over-all analysis of the SSR and AD parameters shows that the quality of the adjustment was good, particularly because AD < 1. However, the number of experimental points. i.e., the range of U022+ concentration, used in this adjustment is smaller than for the Stern-Volmer analysis. The conditional stability constants calculated by this proce- dure are similar to those from the Stern-Volmer plots for the pH 7.0 experiments, similarly to the results found for salicylic acid, and higher for the pH 3.5 experiments. The estimated ZuOzL parameter has a value of Lero for the pH 7.0 and about 15 for the pH 3.5 experiments. These results suggest that the complex is not fluorescent.At pH 3.5, the lU02L parameter is not zero, probably owing to the existence of a background signal or to a more complex complexation scheme, involving competition of the proton for the binding sites. This was probably the cause of the observed differences in the log K’ estimated from the Stern-Volmer plots and from the present adjustment . As the stability of the FUA complexes is larger than for salicylic acid, values for the ligand concentration (C,) at the two pH values were obtained in the present case, as shown in Table 5 . Similarly to what was observed for the conditional stability constants, the calculated concentration of binding sites at pH 7.0 is independent of the carbonate ion concentration (average Cli =: 0.024 mmol 1-l). The number of binding sites available to complex UOz2+ at pH 7.0 is larger than at pH 3.5 (0.01 1 mmol 1-l).This increase is due to the deprotonation of weaker acid structures with binding properties as the pH is raised. As 100 mg 1 - I FUA solutions were analysed in this study, the number of binding sites in the FUA molecules is about 0. I I and 0.24 mmol g- I at pH 3.5 and 7.0, respectively. These values are of the same order of magnitude as calculated for the same sample of FUA when the complexation of Al”’ (at pH 4)34 and Cu” (at pH 6 ) 3 5 were studied by SyF spectroscopy, 0.35 and 0.48 mmol g-1, respectively, and when the complexation of Cu” (at pH 6) was studied by potentiometry (Chi’ ion-selective electrode)35, 0.3 1 mmol g-’. Moreover, the calculated values for K’ fall in the range reported for other humic substances, namely 4.0 (pH 7 and I = 0.1)7 and 5.1 1 (pH 4 and I = 0.1 j. I These results support the quality of the methodology for the study of the interactions of U022+ with HUS presented in this paper. Table 5 Equilibrium parameters for the complexation of UOzl+ by salicylic acid and by the stronger binding site of FUA obtained from non- linear least- squares adjustment” PH N , pICO3’ 1 N, Log K’ Conc. IU03_L Np SSR AD A I CTO22+ I Scrlic:\’lic, acsid- 3.5 1 - 5 2.77(6) - _- 14 1.6 0.2 0.2-1 3.5 2 - 3 4.4(3) 0.011(2) lS(3) 17 2.2 0.3 0.06-0.3 7.0 I S 2 4.18(5) 0.02 l(2) 0.0 20 3.2 0.3 0.01-0.2 7.0 1 4 2 4.06(1) 0.83(2) 0.0 12 3.2 0.4 0.03-0.2 7.0 1 3 3 4.13(4) 0.022(4) 0.0 16 2.6 0.3 0.01-0.1 FlIA- * See footnote to Table 3; Conc.= concentration of the ligand (mmol I-I); IuozL, = fluorescence intensity due to the bound ligand; SSK = sum of squares of residuals; A D = average deviation of the estimates. Average and average deviations in parentheses for log K and Conc. and typical values for the other parameters. No reasonable values were obtained, but fixed concentration values between lo-” and 10-3 mol 1-1 did not affect the other calculated parameters.Analyst, October 1996, V d . 121 1379 0 1 ! 0 0.1 0.2 0.3 Total uranyl concentration/mmol I-' Stem-Volmer plots for the calculated FUA quenching profiles at Fig. 5 pH = 3.5: (a) first and (M) second components. Conclusions The methodology described for the analysis of the interactions of fluorescent ligands with U022+ at the micromolar concen- tration level, owing to its high sensitivity (which is dependent on the fluorescence equipment) and resolving power, which results from the coupling of SyF spectroscopy with a self- modelling mixture analysis method, shows great potential for the analysis of equilibria involving structurally complex ligands such as FUA.The linear Stern-Volmer plots of the quenching induced in the FUA fluorescence by U022+ at pH 3.5 and 7.0 suggests that there is one type of binding site structure that predominates. Both the Stern-Volmer and the non-linear least-squares analy- sis allowed the calculation of the conditional stability constants between FUA and U022+. The latter method also provided FUA binding site concentrations. From the environmental point of view, the stability of the complexes is relatively high, suggesting that FUA play an important role in the speciation of UO22+ in acidic soils, in the absence of other fxtors.An MSc grant (to C.J.S.O.) is acknowledged with respect to the Project PRAXIS XXI. A Perkin-Elmer LS-50 luminescence spectrometer and a Christ Alpha 1 -4/LCD- 1 Freeze Dryer were acquired through Project CIENCIA 27/M/90 awarded by JNICT (Lisbon). W. Windig (Eastman Kodak, Rochester, NY, USA), is thanked for providing a copy of the SIMPLlSMA package. References 1 2 3 Shanbhag, P. M., and Choppin, G. R., J I n o q NucI. Chem., 198 1,43, 3369. Allard, B., Olofsson, U., and lorstenfelt, B., Inorg. Clzim. Ac ta, 1984, 93, 205. Christopher, C. J., in Conipi.ehensive Coordination Chenzistry, The Synthesis, Reactions, Properties and Applications of Coordination 4 5 6 7 8 9 10 I 1 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 Compounds, ed.Wilkinson, G., Gillard, R. D.. and McCleverty, J. A., Pergamon Press, Oxford, 1987, ch. 65. Moulin, V., Tits, J., Moulin, C., Decambox. P., Mauchien, P., and Ruty. 0. de, Radiochim. Acta, 1992, 58/59, 128. Higgo, J. J. W., Kinniburgh. D., Smith, B., and Tipping, E.. Ratrliochinz. Acta, 1993, 61, 91. Kim, J. I., Rhee, D. S., Wimmer, H., Buckau, G., and Klenze, R., Radiochim. Acta. 1993, 62, 35. Tipping, E., Rudiochim. A d a , 1993, 62, 141. Marley, N. A., Gaffney, J. S., Orlandini, K. A., and Cunningham, M. M., Environ. Sci. Trchnol., 1993, 27, 2456. Dozol, M., and Hagemann, R., Pure Appl.Chrm., 1993, 65, 10x1. Johnson, R. O., Ground Water, 1994, 32, 293. Esteves da Silva, J. C. G., and Machado, A. A. S. C.. Chetnom. Intcll. Lab. Syst., 1993, 17, 155. Silva, C. S. P. C. O., Esteves da Silva, J. C. G., and Machado, A. A. S. C., Appl. Spectrosc., 1994, 48, 363. Machado, A. A. S. C., Esteves da Silva, J . C. G., and Maia, J. A. C., Anal. Chin?. Acta, 1994, 292, 121. Esteves da Silva, J. C. G., and Machado, A. A. S. C.. Chmom. Int~~ll. Lab. Syst., 1995, 27, 115. Esteves da Silva, J. C. G., Machado, A. A. S. C., and Garcia, T. M. O., Appl. Spectrosc., 1995, 49, 1500. Ryan, D. K., and Wcber, J. H., Anal. Chrm., 1982, 54, 986. Ryan, D. K., and Weber, J. H., Environ. Sci. Technol., 19x2, 16, 886. Ventry, L.-S., Ryan, D. K., and Gilbert, T. R., Miu-nchmm. .I., 199 1, 44, 201. Cabaniss, S. E., Environ. Sci. Technol., 1992, 26, 1133. Cook, R. L., and Langford, C. H., Anal. Chcni., 1995, 67, 174. Windig, W., and Guilment, J., Anal. Chem., 1991, 63, 1425. Windig, W., Heckler, C. E., Agblevor, F. A., and Evans, R. J., Chetnonz. Intcll. Lab. Syst.. 1992, 14, 195. Ahrland, S., in Environmental Irzoi-ganic~ Chen?i.str.y, ed. Irgolic. K. J., and Martell, A. E., VCH, Weinheim, 1985, ch. 11. Vo-Dinh, T., Anal. Chem., 1978, 50, 396. Lloyd, J. B. F., Analyst, 1980, 105, 97. Gonqalves. M. L. S., and Mota, A. M., Tulania, 1987, 34, 839. Lakowicz, J. R.. Principles of Fluoreswnw Sjwctt-oscop-y, Plenum Press, New York, 1983, ch. 9. Carraway, E. R., Dcmas, J. N., and DeGraff, B. A., Anal. Chenz., 199 I , 63, 332. Fraiji, L. K., Hayes, D. M., and Werner, T. C., J . Chem. Ed~tc,., 1992, 69, 424. Leher, S. S., Biochemistry, 1971, 10, 3254. Esteves da Silva, J. C. G., and Machado, A. A. S. C., Analyst, 1995, 120, 2553. Sylva, R. N., and Davidson, M. R., J . Chrm. Soc., Dalton Trans., 1979,465. Esteves da Silva, J. C. G.. Machado, A. A. S . C., and Silva, C. S. P. C. O., Anal. Chim. Acta, 1996, 318, 365. Esteves da Silva, J. C. G., and Machado, A. A. S. C., Appl. Spectrosc., 1996, 50, 436. Esteves da Silva, J. C. G., PhD Thesis, Faculdade de Ciencias do Porto. 1994. Paper 6/01 086E Received February 14, I996 Accepted June 28, I996
ISSN:0003-2654
DOI:10.1039/AN9962101373
出版商:RSC
年代:1996
数据来源: RSC
|
13. |
Pesticides by solid-phase microextraction. Results of a round robin test |
|
Analyst,
Volume 121,
Issue 10,
1996,
Page 1381-1386
Tadeusz Górecki,
Preview
|
PDF (822KB)
|
|
摘要:
Analyst, O(*tober 1996, Vol. 121 (1381-1386) 1381 Pesticides by Sol id-p hase Microext raction. Results of a Round Robin Test* Tadeusz G6reckiu,t Raymond Mindrupb and Janusz Paw1iszynu.i' N2L 3G1, Camrdu i7S~/pelc.o, I n c . , Supelcw Parh, Bellefiinte, PA 16823, USA Department of' Chemistry, Univtwity of Waterloo, Waterloo, Ontario, The applicability of solid-phase microextraction (SPME) for the analysis of semi-volatile compounds in water was verified by an interlaboratory study on pesticide analysis by SPME. Eleven laboratories in Europe and North America took part in the test. No previous experience with SPME was required. The test sample contained 12 pesticides representing all main groups (organochlorine, organonitrogen and organophosphate) at low ppb levels. The results of the test proved that SPME is an accurate and fast method of sample preparation and analysis.It can be an excellent alternative to currently used methods. Keywords: Solid-phase microextrac*tion; pesticide aimlysis; round robit2 test Introduction Solid-phase microextraction (SPME) is being applied increas- ingly often in many applications. The method utilizes a fused- silica fibre coated with a polymeric stationary phase for analyte extraction from the matrix. The fibre is mounted for protection in a syringe-like device. The analytes partition into the stationary phase until an equilibrium is reached in the system. The amount extracted under these conditions is dependent on the partition coefficient between the sample and the coating. Initially, SPME has been used mainly for the analysis of volatile organic compounds, including substituted benzenes,'-3 volatile organic compounds in watee 6 and chlorinated hydrocarbons.7 The reported applications for semi-volatile compounds include among others caffeine in beverages,x polycyclic aromatic hydrocarbons and polychlorinated biphenyls,9 phenolslo-12 and pesticides.1 7-20 In order for SPME to gain further acceptance, it is necessary to demonstrate that this method can perform reliably for a variety of semi-volatile compounds at trace concentration levels. Pesticides were chosen for this purpose, as this broad group includes many classes of compounds with very different chemical characteristics. The test was designed to verify the possibility of using SPME in its simplest form.No attempts were made to optimize the analysis conditions by matrix modifications or coating selection to avoid additional com- plexity. The basic 100 pm poly(dimethylsi1oxane) (PDMS) fibre was chosen, as it is usually the first choice in SPME. Experimental Test Procedure The following compounds were selected for the test (in elution order): dichlorvos, EPTC (eptam), ethoprofos, trifluralin, Presented at the 18th International Symposium on Capillary Chromatography, Riva del Garda, Italy, May 20-24, 1996. + On leave from the Faculty of Chemistry, Technical University of Gdansk, Gdansk, Poland. * To whom correspondence should be addressed. simazine, propaLine, d i a h o n , methyl chlorpyriphos, hep- tachlor, aldrin, metolachlor and endrin. Two solutions of the test compounds in methanol were prepared by Supelco (Rellefonte, PA, USA): a 10 pg ml-1 standard solution and a blind solution containing an unknown to the participants amount of each analyte.Each participant obtained a test kit containing a 30 m x 0.25 mm id X 0.25 pm SPB-S column (Supelco), an SPME holder and three fibres coated with 100 pm poly(dimethy1silox- ane) (Supelco), 5 X I ml of a 10 ppm standard solution of pesticides in methanol, 3 X 1 ml of a methanolic pesticide test mix, description of the procedure, sample chromatogram of the standard mixture, NlST library spectra of the analytes, table of quantitation ions and results report forms. Each participant was required to use the following equipment: a gas chromatograph equipped with a split-splitless, PTV or SPI (Varian, Palo Alto, CA, USA) injector; a quadrupole or ion trap ma\\ spectrometer coupled to the gas chromatograph; a data acqui4ition and processing system, preferentially capable of library searching; an ultrapure water system as the source of water for the preparation of aqueous solutions; and a quality magnetic stirrer ensuring a constant and repeatable stirring rate.The following laboratories took part in the test (in alphabet- ical order): Analytical Chemistry Chair, Technical University of Gdansk, CISM, Florence University, Italy; ENEA CRE Ambiente, La Spezia, Italy; Fraunhofer-Institut fur Toxikologie und Aerosolforschung, Istituto di Ricerche Farmacologiche, Milan, Italy; Joint Research Centre. Ispra, Italy; PEI Food Technology Centre, Charlottetown, PE.Canada; Research Institute for Chromatography, Kortrijk, Belgium; Soil Science Department, University of Manitoba, Winnipeg, South Carolina Department of Natural Resources, Charles- UFZ Centre for Environmental Research Leipzig, Ger- Each laboratory was assigned a code number on a random basis to be used throughout reporting. Laboratory numbers given in the tables are the code numbers, and bear no relation to the sequence presented above. The participants received their kits in November 1995 and were expected to deliver the results by the end of January 1996. Poland; Hanover, Germany; MB, Canada; ton, SC, USA; and many. The test procedure included the following steps: Conditioning of the column according to the manufactur- er's specifications and check of the column blank.Conditioning of the fibre according to the manufacturer's specifications and check of the fibre blank. For quadrupole MS users only: syringe injection of the 10 ppm standard in order to establish the retention times of the analytes and set up the selected-ion monitoring (SIM) acquisition method.1382 Anulyst, October- 1996. Vol. 121 (4) Preparation of a 30 ppb aqueous standard followed by ( 5 ) Repeated analysis of a freshly prepared 30 ppb aqueous (6) Analysis of 10 and I ppb aqueous standards in a similar (7) Determination of the calibration curves for all the analy- (8) Preparation of the aqueous solution of the blind sample (9) Calculation of the concentration of the analytes in the blind The test procedure specified that extractions should be carried out with samples vigorously stirred.The extraction time was set at 45 k 0.5 min. The chromatographic conditions were as follows: injector and transfer line temperature, 250 "C; temperature programme. 40 "C held for 5 min, increased at 30 "C min-1 to 100 OC, at 5 "C min-1 to 250 "C and at SO "C mio- I to 300 "C, held for 1 min. Each participant was required to report the results using the forms included in the test kit and to attach the following: chromatograms of the column blank and fibre blank; chromatogram of the syringe injection and spectra of the analytes (quadrupole MS only); chromatogram of the 30 ppb standard (all systems) and the spectra of the analytes (ion trap (IT) MS] only; chromatograms of the remaining standards (one for each level of calibration); sample chromatogram of carryover determination; chromatograms of the unknown sample; and a copy of the spreadsheet used for calculation of results.SPME/GC-MS analysis and subsequent carryover check. standard. way. tes. followed by SPME analysis. aqueous sample. Processing of Results In order to ensure that all the data were processed in the same way, the entire processing of results was carried out on the basis of raw data (peak areas) reported by the participants. A spreadsheet was designed for this purpose using Lotus 1-2-3 for Windows, release S (Lotus Development, Cambridge, MA, USA). The calibration curves were forced through the origin, as the blank values reported by the participants were all zero. In addition, the deviations from linearity at low concentration levels affect the course of the regression line calculated by the least-squares method (calibration curve) to a much lesser extent than similar percentage deviations at higher concentrations, which can lead to relatively large errors at low Concentrations if the calculated intercept is non-zero.The slope, its standard error and linear correlation coefficients were calculated for each regression line. The data were also plotted to permit visual examination of the possible trends. The entire procedure was aimed at eliminating data processing as the source of the observed variability. Further processing of the results was based on IS0 Standard 572.5.2I.22 This standard utilizes the ANOVA technique for the estimation of gross average, intralaboratory and interlaboratory variances, repeatability and reproducibility of the method.The following is a brief description of the procedure used. In order to use the ANOVA technique, the following assumptions must be true: (1) the data distribution within a given cell follows the normal distribution law; (2) the laboratory mean distribution follows the normal distribution law; and (3) intralaboratory variances are equal. i.e., the data distribu- tions are homoscedatic. Tests are applied to verify assumptions (1) and (2). The within-laboratory variability is examined using Cochran's test and the between-laboratory variability using tirubb's test. In the rest of the text the following notation is used: i denotes the laboratory number, varying from 1 t o p (p = 1 l ) , andj denotes the replicate number varying from 1 to yzi (IZ; = 3).Cochran's test uses the following statistic: C S i 2 i=l where si is the standard deviation of results from laboratory i and smax is the highest standard deviation in the set. If the test statistic is greater than its 5% critical value and less than or equal to its 1% critical value, the item tested is called a straggler, and is left in the set. If the statistic is greater than the 1 % critical value, the item is called an outlier and the result is rejected. After an outlier is detected, the test is repeated on the remaining values. Crubb's test is performed on the results that were not rejected in the first test. This test uses the following statistic for the largest value in the set: (2) GI, = (xP - i ) / S where l i ? x = -xx, i= I P and Y ' i=I (3) (4) For the lowest value, the statistic is Similar rules apply to detection of stragglers and outliers as in Cochran's test.If Grubb's test does not detect single outliers, a variation of it is applied to detect if two largest or two smallest observations can be outliers. The test statistic in this case is for two largest observations and for two smallest observations. where i=l n-2 i=l sl,** = pi - " , 2 ) 2 i=3Analyst, 01-toher 1996, Vol. 121 1383 Intralaboratory mean: - -Yl,2 = - zxr (12) 1=3 P - 2 - /=I "Y, = - nl If the test statistic is less than its 5% critical value and greater than or equal to its 1% critical value, the two results are called stragglers, and are left in the set. If the statistic is lower than the 1% critical value, the item is called an outlier and the result is rejected.Following the detection of outliers, the statistical analysis is performed. The following statistics are calculated: Estimate of the repeatability variance: Gross average: Estimate of intralaboratory variance: M \2 where * ss, = f i=l 2 j=l si = ni -1 j=I Repeatability: Reproducibility : N =z n, (total number of data), and p is the number of laboratories. r=I Estimate of interlaboratory variance: \ Repeatability is an estimate of method reliability. whereas reproducibility is related to a measurand rather than method.2' The accuracy of the results obtained by the participants was evaluated by comparing their confidence intervals (gross average 2 repeatability) with the confidence interval of the 'true' value.The latter was estimated based on the accuracy of preparation of the blind sample (+US% relative, as reported by Supelco), in addition to the accuracies of the syringe and pipette used for the preparation of aqueous standards from the methanolic solution (0.5 pl for a 25 pl syringe and 0.1 ml for a 25 ml pipette). The law of propagation of errors was used for this purpose. If the confidence intervals of the 'true' value and the gross average for a given compound overlap, the difference between the two values is not statistically significant and the method is accurate. \ n-1 I (15) 7 \ I sL,- = N' where SSI- = ss, - ss, f P I?, 1 Results and Discussion The participants did not report any problems with the test procedure. However, some of them did not follow the protocol.For example, laboratory No. 002 did not run the standard at the 1 ppb level, and made only two replicates for the blind sample (three were required). Laboratory No. 006 used 4, 10 and 20 ppb standard solutions instead of 1, 10 and 30 ppb. Laboratory No. 01 1 used a different column for the separation (DB-5, J & W Scientific, Folsom, CA, USA), which resulted in a reversed elution order for aldrin and metolachlor. Also, the column blanks from some other laboratories indicated that the new column obtained in the test kit was not used for the test. Some of the laboratories indicated in the follow-up after the test that the aqueous pesticide solutions (standard and blind) were all prepared at the same time and stored before the analysis for a certain time, instead of being freshly prepared directly before the extraction.This is another potential error-prone area, Estimate of reproducibility variance:as the stability of aqueous solutions i \ much poorer than that of methanolic solutions, and sorne of the pesticide5 are easily lost as a result of adsorption on the walls of non-silanized glass vials. As the low concentration standard and the blind sample were thc last ones to be analysed according to the protocol, this might have resulted in deviations from linearity observed on sonic of the calibration curves and in lowered results for the blind sample. Finally, vials of different sizes were used by the participants. Thc protocol recommended the use of 40 in1 vials, as it has been established previously that the use of sinall vials results in significantly worse precision.The choice of vials, however. was probably dictated by their availability in the participating laboratories. Several test participant\ (laboratory Nos. 004, 006, 007 and 009) reported instrumental problems. In at least one case they were successfully ovcrcome (laboratory No. 006), as indicated by the results reported. The results from laboratory Nos. 007 and 009 were submitted 1 month after the deadline, which might have affected the composition of the methanolic solutions (depending on the storage conditions). Table 1 Standard deviations of analyte determination for the participating iabordtories Standard deviation, ,s, Coin pou rid Lab.001 Lab. 002 Lab. 003 Dichlorvos EPTC Ethoprof-os Trifuralin S i iiiazi ne Propazine Di ;i/ i non Methyl chlorpyriphos Heptxh lor Aldrin Me to I ach lor Endrin I .075 0.507 1.017 0.567 1.120 0.333 0.566 0. I96 I .324 0.07 5 0.4x7 1.135 0.976 0.399 0.786 0.1 16 1.158 1.25 I 0.343 O.OX3 0.334 0.540 0.770 0.0 I 9 2.15x 02x4 0 .‘>04 0.20 1 2.585 0.58 I 0.4 I 1 0. I37 2.724 0.774 0.850 0.984 Dashes indicate I alues not reported by the participants. Lab. 004 2.999 1 .loo 3.400 0.101 4.686 0.259 1.261 0.639 2.578 0.322 2.373 1.718 Lab. 005 I .057 0.395 0.746 0.2 I6 0.423 0.227 0. I 84 0.05 5 1.795 0.552 0. I07 0.332 Lab. 006 1.489 0.44 1 0.5 16 0.153 1.73 I 0.299 0.038 0.054 0.82X 0.269 0.704 0.256 Lab. 007 __ 0.452 0.7 16 0.3 I7 2.59 1 0.699 0.155 2.233 I .249 2.758 - - Lab.008 2.689 0.380 0.37 1 0.2xx 2.019 0.567 0.743 0.1 I6 4. I75 0.370 0.736 0.X66 Lab. 009 3.067 0.86 1 1.169 0. I70 0.750 0.94 1 0.178 0.846 0.25 I 0.342 0.9 I 1 _- Lab. 010 0.525 0.100 0.7 60 0.268 1.813 2.344 0.121 0.03 3 0.464 0.3 14 1.124 0.324 Lab. 01 1 1.467 0.264 0.679 0.094 2.1 5 0 0.216 0. I Oh 0.060 0.350 0.1 12 0.53x 0.397 Level 1 Level 2 Con1 pou nd Dichlorvos EP’I’C‘ Ethoprofos Trifluralin S i inazi iit: Propazine Diuinon Methyl chlorpyriphos Heptachlor Aldrin Mctolachlor End r i n a\,)? 3x. 159 3.249 0.749 46.40 I 15.359 4.1 83 0.539 43.306 4. I94 9.903 13.X34 I 7.92 I Lab. N o . 009 ( 104 004 00 1 004 007 004 ( 104 OOX 007 004 007 Crltlcal value c 5% 0.247 0.445 0.372 0.417 0.645 0.4 I7 0.429 0.4 17 0.473 0.478 0.437 0.417 0.380 0.41 7 0.733 0.4 17 0.402 0.4 17 0.372 0.4 17 0.569 0.445 0.513 0.417 Cri tical value 1 ql 0.536 0.504 0.504 0.504 0.573 0.504 0.504 0.504 0.504 0.504 0.536 0.504 Out1 ier x t rag g 1 e I- 01 - 0 u t 1 I er Straggler -_ Straggler - Outlier __ - Outlier Outlier Critical valuc 5% Z(S,)? Lab.No. C - - - - 6.360 009 0.215 0.445 - - - - 0.144 001 0.266 0.445 Cntical Outlier value or 1 % straggler - - 0.536 - - - 0.536 - 4.272 010 0.296 0.478 11.883 004 0.248 0.445 0.573 - 0.536 - Table 3 Mean coiiccntration\ (pg I I ) of the analytes determined by the participating laboratories (values rejected by Cochrdn’s test not included) Compound Lab. 001 Lab. 002 Lab. 003 Lab. 004 Lab. 005 Lab. 006 Lab. 007 Lab. 008 Lab. 009 Lab. 010 Lab. 01 1 D ic h lorvos 30.9 EPTC 10.9 Etlioprofos 19.7 Trifluralin I .9 S mi azi ne 26.8 Diazinon X.9 Methyl chlorpyriphw ! .X Aldrin 2.6 Metolachloi 17.9 Eiidrin 8.9 PropaLine 10.2 Heptxhlor 10.0 35.2 13.6 22.I I .8 30.8 10.8 11.3 1.9 5.4 3 .0 I X.8 15.5 20.6 8.9 13.5 I .2 19.2 7.8 7.2 1.3 9.6 I .9 13.2 6.2 33.8 103 0.5 21.9 8.4 5.3 3.2 0.6 3.9 __ - - 27.3 10.8 18.0 I .s 24.3 10.5 9.6 1.7 7.5 I .7 17.9 9.3 27.5 9.9 17.3 I .9 23.1 9.7 8.4 1.8 9.8 2.x 17.1 9.2 - 9.3 12.8 1.8 13.8 9.7 1.3 10.8 6.4 - 30.1 10.8 18.5 1.9 25.1 8.0 10.8 2.4 14.1 2.6 17.7 12.0 23.6 7.4 12.4 1.3 7.2 7.4 1.6 6.5 I .0 10.3 7.0 - 20.5 10.8 18.3 2.6 25.7 11.6 8.7 I .4 12.3 2.x 15.2 10.9 10.2 7.8 5.2 0.9 18.0 6.6 4.0 1 .5 7.8 1 .8 12.9 7.6 Dashes indicate values not reported by the participants or rejected by Cochran’s test.Aiiulyst, October- 1996, Vol.12 I 1385 ~~~~~ ~ ~~ Table 1 gives the standard deviations of analyte determina- tion for the participating laboratories. The values in this Table were used for the calculation of Cochran’s test statistics. presented in Table 2. Cochran‘s test performed on all the data (Level 1 in Table 2) indicated the presence of two straggler\ (trifluralin in laboratory No. 00 I and propazine in laboratory No. 007), and four outliers (ethoprofos, methyl chlorpyriphos and inetolachlor in laboratory No. 004 and endrin in laboratory No. 007). The stragglers were left in the set and the outliers were rejected. No further stragglers or outliers were found after repeated application of Cochran’s test to the remaining values (denoted Level 2 in Table 2). Table 3 presents mean analyte concentrations in the blind sample, as determined from the raw data reported by test participants, using the spreadsheet designed for thi5 purpose.Values rejected by Cochran‘s test were not used in the calculations. The values reported in Table 3 differ slightly in some caqes from the values reported by the participants, as the calibration curves were not determined in exactly the same way. The values presented in Table 3 were used for the calculation of Grubb’s test statistics. Table 4 presents the results of application of Grubb’s test for single largest and single lowest value in the set. No stragglers and one outlier (aldrin in laboratory No. 007) were found. The outlier was rejected from further data processing. Table 5 presents the results of application of Grubb’s test for two largest and two smallest observations in the set.This variation of Grubb’s test is applied only when no single outlying observations are found, and therefore it was not used for aldrin. No stragglers or outliers were found when analysirig two extreme observations. In general, the linearity of the calibration curves was excellent for all the analytes. Table 6 presents mean correlation coefficients and the highest and the lowest coefficient value\ reported for each compound. For all the analytes the mean linear correlation coefficients were better than 0.99000. The highest values reported ranged from 0.99967 to 1.00000 (with five significant digits). The lowest values ranged from 0.95555 to 0.99265. In most cases they originated from laboratories that reported instrumental problems (five compounds in laboratory No.004, two compounds in laboratory No. 007 and one compound in laboratory No. 009). In some cases the course of Table 6 Mean, maximum and iiiinimiim values of the correlation coefficients of the calibratioii curvcs reported for each analyte Coi-relation codficient Compound Dichlorvos EPTC Ethoprofos Tri tlul-din Simazine Propazi tie Diazinon Methyl chlorpyriphos Heptachlor A1 dri n Metolachlor Endrin Mean 0.99640 0.99737 0.99425 0.99667 0.99794 0.998 12 0.99559 0.99823 0.99410 0.99 17 1 0.99249 0.9964 1 Max. 0.99995 0.99998 0.99999 0.99997 0.99985 0.99997 0.99995 1.01)000 0.99967 0.99976 I .00000 0.99999 Min. 0.9787 8 0.97726 0.95765 0.97 122 0.99200 0.99265 0.97769 0.99204 0.955 5 5 0.969 12 0.95822 0.97 180 Table 4 (irubb‘s test results lor mgle extreme observation Compound Dichlorvos EPTC Ethoprofos Tritlural in Siniazine Propazine Diazinon Methyl chlorpyriphos Heptachlor Aldrin Metolachlor Endrin Mean/pg I- I 27.47 10.03 15.76 1.57 23.89 9.50 8.29 1.65 8.82 2.47 15.67 9.05 \ 5.005 1.626 4.637 0.559 3.693 2.039 2.I00 0.309 2.995 1.447 2.773 3.083 Lab. No. 002 002 002 010 002 007 002 008 008 007 002 002 G high 1.547 2.194 1.37 1 1.850 1.873 2. I14 I .434 2.2118 1.752 2.728 I . 146 2.100 Lab. No. G low 01 I 1.66 1 009 1.631 01 1 2.287 004 1.994 01 1 1.603 01 I 1.445 01 1 2.041 003 1.311 004 I .88 1 004 1.88 1 009 1.940 004 1.669 Critical value 5% 2.290 2.355 2.290 2.355 2.2 IS 2.355 2.355 2.290 2.35s 2.1155 2.215 2.290 Critical value Outlier or 1% 2.482 - 2.564 __ 2.482 - 2.564 __ 2.387 - 2.564 ~ 2.564 - 2.482 - 2.564 - 2.564 High outlier 2.387 - 2.482 - s t ragg I er Table 5 Grubb’4 test results for two extreme observdtions Compound Dichlorvos EPTC Ethoprofos Trifl uralin Simazine Propazine Diazinon Methyl chlorpyriphos Heptachlor Aldrin Metolachlor Endrin .S$ 250.52 29.07 215.05 3.43 122.7 1 45.74 48.50 0.95 98.70 69.20 95.02 - .$/I ~ I./,Z 125.89 13.36 145.86 2.05 52.79 18.47 29.9 I 0.32 50.19 49.80 33.30 - G 0.502 0.460 0.678 0.597 0.430 0.404 0.6 17 0.339 0.509 0.720 0.35 1 ~ Lab. No. 002. 004 002. 00 I 002,001 010,008 002, 001 007, 0 I0 002, 008 008,002 008, 010 002,001 002, 008 - .y I ,Z2 105.71 14.4 1 66.33 1.39 49.75 28.89 15.52 0.6 1 46.1 1 23.03 52.65 - G 0.422 0.496 0.308 0.404 0.406 0.632 0.320 0.644 0.467 0.333 0.554 - Lab.No. 01 1,003 009, 01 1 01 I , 009 004, 01 1 01 1. 003 01 1.009 011, 004 003, 007 004,002 009, 01 I 004, 003 - Critical value 5% 0.149 0. I45 0. I86 0.145 0.149 0.14.5 0.145 0. I49 0.145 0.149 0.186 - Critical value Outlier or 1% straggler 0.085 - 0.22 1 - 0.1 15 - 0.22 I - 0.085 - 0.22 1 - 0.22 1 - 0.oxs - 0.22 1 - 0.085 __ 0.1 15 ~ - -1386 Anulyst, Octohei- 1996, Vol. 121 Table 7 Statistical characteristics of the results obtained by the participating laboratories for the blind sample. s, Repeatability standard deviation; sL, interlaboratory standard deviation; sR, reproducibility standard deviation; r , repeatability; R, reproducibility; GA, gross average; CI, confidence interval of the gross average; TV, confidence interval of the ‘true’ value.All values expressed in pg 1-1 Compound Dichlorvos EPTC Ethoprofos Trifluralin Simazine Propazine Diazinon Methyl chlorpyriphos Heptachlor Aldrin Metolachlor Endrin s I SL J I? r- 2.06 5.04 5.44 5.83 0.56 1.56 1.66 1.57 0.82 4.79 4.86 2.32 0.27 0.57 0.63 0.76 2.34 3.45 4.17 6.61 1.21 2.04 2.37 3.42 0.63 2.13 2.22 1.79 0.12 0.32 0.34 0.35 2.03 2.89 3.53 5.75 0.54 0.73 0.91 1.53 0.73 2.83 2.92 2.07 0.87 3.00 3.13 2.47 R 1 5.40 4.70 13.74 1.79 11.79 6.7 1 6.29 0.97 10.00 2.58 8.28 8.85 GA 27.3 9.9 15.5 1.6 23.6 9.5 8.2 1.6 8.9 2.0 15.7 8.8 CI 27 f 5.8 10f 1.6 16 f 2.3 1.6 f 0.76 24 f 6.6 10 f 3.4 8 f 1.8 1.6 f 0.35 9 k 5.8 2 f 1.5 16f2.1 9 * 2.5 TV 25 iz 1.35 10 k 0.54 17 iz 0.92 2F0.11 2s k 1.35 1040.54 I0 f 0.54 2k0.11 10 f 0.54 2k0.11 17 iz 0.92 10 f 0.54 the calibration curves indicates that aqueous solutions might not have been freshly prepared. Table 7 presents the statistical characteristics of the results obtained by the participants for the blind sample.In general, the results are characterized by good repeatability, which proves that SPME is a valid method for the determination of a very diversified group of semi-volatile compounds at trace levels. As expected, the reproducibility standard deviations are higher, but still satisfactory. The values of SR were significantly affected by the results from the laboratories that reported instrumental problems. Even though several results from those laboratories were rejected, the remaining ones, not fulfilling the statistical criteria for rejection.contributed significantly to the observed value of reproducibility standard deviation ,TR. The results in Table 7 indicate that SPME is an accurate method. In all cases the confidence intervals of the gross average and the ‘true’ value overlap, which indicates that any differences between the two respective values are due to random factors. Interestingly, for 10 out of 12 compounds the values of the gross average are slightly lower than the ‘true’ values. This might be due in part to the losses of analytes through adsorption (as described at the beginning of this section) in cases when the aqueous solutions were not prepared directly before the analysis. Conclusions The laboratories taking part in the test included both those which use SPME on a regular basis and those which used this technique for the first time.No significant differences in performance were observed between the two groups. The results obtained indicate that SPME is a valid method for the determination of trace amounts of semi-volatile pesticides in water, even though the method was used ‘as is’, without any attempt to optimize the conditions. The repeatability, repro- ducibility and accuracy of the results were satisfactory in all cases. Taking into account the diversity of the compounds studied, it can be concluded that it should be possible to use SPME successfully for the determination of many other classes of semi-volatile compounds. Compared with other currently used methods, SPME offers several very significant advantages, including complete elimination of solvents, very low cost related to reusability of the fibre and no requirements for dedicated instrumentation, over-all simplicity and time savings.The sensitivity of the method is very good, and can be further improved by optimization of the analytical procedure (coating selection, matrix modification, etc.). idea of the test. We gratefully acknowledge all test participants for their contribution to the success of this study. Financial support of Supelco Canada, Varian and the National Sciences and Engineering Research Council of Canada is also ac- knowledged. References 1 2 3 4 5 6 7 8 9 10 I 1 12 13 14 15 16 17 18 19 20 21 22 Arthur, C. L., Killam, L. M., Motlagh, S., Lim, M., Potter, D. W., and Pawliszyn, J., Environ. Sci.Technol., 1992, 26, 979. Potter, D. W., and Pawliszyn, J., J. Chromatogr., 1992, 625, 247. Wittkamp, B. L., and Tilotta, D. C., Anal. Chem., 1995, 67, 600. Arthur, C. L., Pratt. K., Motlagh, S., Pawliszyn, J., and Belardi, R. P., .I. High Resolut. Chromatog r.. 1992, 15, 741. Langenfeld, J. J., Hawthorne, S. B., and Miller, D. J., Anal. Chem., 1996, 68, 144. Nilsson, T., Ferrari, F., and Facchetti, S., in Proceedings qf the 18th liiterncttional Symposi~im on Capillary ChromatoKraphy, R i w del Gar-da (Italy), May 20-24, 1996, ed. Sandra. P., and Devos, G., IOPMS, Korfrijk, 1996. p. 618. Chai, M., Arthur, C. L., Pawliszyn. J., Belardi, R. P., and Pratt, K. F., A/?afyst, 1993, 118, 1501. Hawthorne, S. B., Miller, D. J., Pawliszyn, J., and Arthur, C. L., J . Chromatogr., 1992, 603, 185. Potter, D., and Pawliszyn, J., Environ. Sci. Techno/., 1994, 28, 298. Buchholz, K., and Pawliszyn, J.. Environ. Sci. Technol., 1993, 27, 2844. Buchholz, K., and Pawliszyn, J., Anal. Chem., 1994, 66, 160. Schaefer, B., and Engewald, W., Frescnius’ J . Anal. Chem., 1995, 352, 535. Boyd-Boland, A. A.: and Pawliszyn, J., J . Chromatogr.. 1995, 704, 163. Eisert, R., Levsen, K., and Wuensch, G., .I. Chromutogr., 1994,683. 175. Popp, P.. Kalbitz, K., and Oppermann, G., .I. Chromatogr., 1994,687, 133. Lee, X., Kumazawa, T., Taguchi, T., Sato, K., and Suzuki, O., Hochudoku, 1995, 13, 122. Eisert, R., and Levsen, K., Fresenius’ J . Anal. Chem., 1995, 351, 555. Levsen, K., and Eisert, R., ,I. Am. So(*. Muss Spectrom., 1995, 6, I 119. Graham, K. N., Sarna, L. P., Webster, G. R. B., Gaynor, J. D., andNg, H. Y. F., J . Chromatogr., 1996, 725, 129. Magdic, S., and Pawliszyn, J., J . Chromatogr., 1996, 723, 11 1. Feinberg, M., Trends A n d . Chem., 1995, 14, 450. International Standards Organisation, Accuracy (Trueness and Preci- sion) of Measurement Methods and Results, IS0 5725-2: 1994, ISO, Geneva, 1994. The authors thank Dr. A. Boyd-Boland for her input during the planning stages of the test and Professor P. Sandra for giving the Paper 6103649J Received May 28, 1996 Accepted July 19, 1996
ISSN:0003-2654
DOI:10.1039/AN9962101381
出版商:RSC
年代:1996
数据来源: RSC
|
14. |
Flow injection method for the determination of arsenic(III) at trace levels in alkaline media |
|
Analyst,
Volume 121,
Issue 10,
1996,
Page 1387-1391
Joseph H. Aldstadt,
Preview
|
PDF (819KB)
|
|
摘要:
Analyst, October 1996, Vol. 121 (13677-1391) 1387 Flow Injection Method for the Determination of Arsenic(iii) at Trace Levels in Alkaline Media Joseph H. Aldstadt and Alice F. Martin Environmental Research Division, Ai-gonne National Laboratory, 9700 South Cuss Avenue, Argonne, IL 60439, USA A flow injection (FI) method for the determination of trace levels of trivalent arsenicals in environmental samples is reported. The method is applicable to arsenic compounds that can be base hydrolysed to yield arsenious acid, which is then detected electrochemically. By using the constant-current mode of potentiometric stripping analysis (PSA), a method was developed for the determination of arsenious acid in basic media at Ep = -475 mV versus Ag/AgCl (3 moll-' NaCl). The method parameters were optimized, including pH, supporting electrolyte, deposition potential, deposition time, stripping current and stripping delay.Additionally, possible electrochemical interferences, including AS\', BPI1, Cd", Cu", Hg", Pb", Sb"', SeIV and Sn", were studied. A detection limit of 0.21 pg I-' was achieved for As"' in aqueous samples (100 pl) over the range 0.10-50 pg 1-I in the stopped-flow FI manifold. The precision (RSD) of the method at 5 pg 1-1 (n = 8) is ~ 5 % . The method was applied to the indirect determination of dichloro(2-chloroviny1)arsine (Lewisite), a chemical warfare agent that is difficult to measure in the environment because it rapidly decomposes to form its geminal diol, 2-chlorovinylarsonous acid (CVAA). CVAA can be base hydrolysed to form arsenious acid for detection by FT-PSA.Keywords: Flow injection; potentionietric stripping; arsenite; arsenious acid; 2-chloi.ovinylai.sonous acid; Lewisite Introduction Arsenic is widespread in the environment, both naturally and as a result of activities such as fossil fuel combustion, glass manufacture and non-ferrous metal smelting. 1 The determina- tion of arsenic species in the environment encompasses many techniques.2 Arsenic has been determined over the past several decades with various electrochemical techniques, including amperometry,? cathodic stripping ~oltammetry,~ anodic strip- ping voltammetry,s and potentiometric stripping analysis (PSA).6-10 Particularly because of the requirement in voltam- metry to deoxygenate the sample, all but PSA are cumbersome to implement in the field.Furthermore, PSA has been shown to possess advantages in sensitivity and selectivity over the voltammetric techniques.' 1 PSA is a sensitive and selective method for electrochemically detecting several heavy metals and metalloids such as arsenic in aqueous environmental samples. 12-14 In acidic solution, As"I exists as metaarsenious acid, HAs02, a weak acid with pK, = 9.29.1s The reductive electrochemistry of arsenious acid involves a three-electron transfer to metallic arsenic. I h This reaction occurs in an acidic solution at a formal potential (E"') of +240 mV, as shown in the following simplified reaction: HAs02+3 H++3 e-=As+2 H20 (1) Metallic arsenic can then be dissolved on to a gold electrode surface as a gold-arsenic amalgam [i.e., As(Au)]. The analytical signal is measured by oxidizing elemental arsenic from the cold surface by using either a chemical oxidant in solution or an applied constant current (or both).The time required to pass through the AS''' redox potential is directly proportional to the concentration of ions in solution. In addition, this approach can be used to speciate trivalent arsenic from pentavalent arsenic by maintaining the potential during the deposition step at a more positive value than that required to reduce the high valence species. In this work, we explored the use of a flow injection (F1)-PSA system for quantifying arsenic(iI1) species (as arsenious acid) in alkaline media as an alternative to the published methods for determining As"' under acidic condi- tions.The development and optimization of the field-portable methodology for this method is described. Experimental Reagents and Supplies All chemicals were of analytical-reagent grade or better and were prepared in high-purity ( I 8 MQ) water (Barnstead NANOPure system, Barnstead/Thermolyne, Dubuque, IA, USA). Optima-grade acids were obtained from Fisher Scientific (Pittsburgh, PA, USA). All solution containers were acid washed to remove background contamination. Instrumentation A PSU22 TraceLab potentiometric stripping analyser was obtained from Radiometer America (West Lake, OH, USA). FI system components were obtained from Global FIA (Gig Habor, WA, USA): an Alitea XV four-channel peristaltic pump, VICI six-port, two-position injection valve and VICI serial valve interface controller and software.Batch PSA experiments were used for detection parameter optimization using the Radiometer SAM20 sample station. For batch PSA studies, a saturated calomel reference electrode (SCE) and Pt wire counter electrode with a salt bridge from Radiometer were used along with a 3 mm id gold disc working electrode (BAS, West Lafayette, IN, USA). For FI studies, a wall-jet (0.33 mm id jet capillary) tlow- through electrochemical cell was used.17 The spacer in the flow cell was cut from a 0.70 mm thick Teflon sheet; the volume of the flow cell above the plane of the working electrode was calculated to be 110 pl. The 3 mm id gold disc working electrode and Ag/AgCl (3 mol I-' NaCl) reference electrode were obtained from BAS.The stainless-steel outlet to the tlow cell served as the auxiliary electrode. The working electrode was stored in 6 mol I-' HN03 and polished daily with 0.05 vm y-Alz03, followed by sonication in water at 55 kHz for 5 min. All potentials reported herein for the FI system are versus the Ag/AgCl reference electrode at 20 "C unless indicated other- wise. Discoloration of the Ag/AgC1(3 moll-' NaCl) reference electrode occurs slowly over time because of the formation of AgOH,,, on the Ag wire from the basic solution. However, the variation in measured potential was negligible under these circumstances.1388 Amlyst, October 1996, Vol. 121 5.0 (d (d Y (d !?? 4.0 3.0 U ?I 8 .- - g 2.0 z 1.0 Procedures Sumple piqaiutio~i Arsenic(rrr) standards were prepared from reductimetric stan- dard-grade As203 as described previously.I* All arsenic- containing solutions with concentrations less than 1 mg 1-1 were prepared on the day of use. Solutions were not deoxy- genated. 5 - I I - TI! /4 - I I , I I - 1 . 1 . 1 . 1 . Potentionietric strippirig Optimum deposition conditions determined during the course of this work were an initial potential of -1050 mV, a final potential of +300 mV and a deposition time of 180 s over the 0.5-100 pg 1- I calibration range. Shorter deposition times can be used t o extend the calibration range to higher values. Stripping was performed in the constant-current mode (typic- ally +O.IO pA applied current) and stripping data were treated by using a digital filter (eight-point boxcar average, followed by a nine-point third-order Savitzky-Golay filter) and digital curve fitting for background subtraction.F ION' injec tiori The FI carrier solution (0.75 ml Inin-') was 10 mmol 1 - 1 sodium hydroxide at 0.90 ml min-1 using 0.030 in id Teflon manifold tubing. Samples were loaded into the six-port, two- position injection valve fitted with a 100 pl injection loop. Im- mediately prior to each deposition period, the electrode was cycled (from +SO0 to - I 100 mV at 3 s intervals for ten cycles) under carrier flow to clean electrically the surface of the solid gold working electrode. The carrier flow was stopped during the deposition and stripping steps. Culibi*crtion A linear calibration model was constructed by using arsenious acid standards prepared in carrier solution at 0, 0.5, 5 and 50 pg 1 - 1 in random order.Carryover between standards was less than 1% under these conditions. Results and Discussion Development of Detection Method For the determination of As"' species in basic media, a complication arises because the caustic solution must be acidified in existing PSA methods for inorganic arsenicals.6-'0 The acidification step would complicate the FI manifold design (see below) and also the safe operation of a field-portable instrument. Further, the use of a gold-film electrode would also complicate the instrument because separate lines for Au"' plating solution would be required in the FT manifold. Although these modifications would be routine for a bench-top FI system, they present practical problems for a field-portable instrument.Therefore, our goal was to develop a method for determining arsenious acid directly in basic solution by using a solid working electrode. Above pH 8, metaarsenious acid dissociates to form its conjugate base, the arsenite ion (As02-). IC) Attempts by previous workers to determine arsenite ion by voltammetry or polarography in basic solution resulted in poorly defined reduction waves, presumably caused by an equilibrium shift toward orthoarsenious acid, HA SO^.^^) Arsenite ion reduces from alkaline solution 011 to a mercury surface, as shown in the following reaction (EO' = -680 mV):19 As02-+2 H 2 0 + 3 e-=As+4 OH- (2) The complex nature of the cathodic deposition of arsenite ion in basic solution has been studied by polarography, in which arsenic was thought to undergo crystallization polarization on a mercury surface;21,22 however, we found that reaction (2) could be reliably implemented on either a gold film (on glassy carbon) or solid gold disc working electrode with EO' = -680 mV.The solid gold disc working electrode was chosen for further optimization because it would be easier to support in the field. To our knowledge, the work presented here is the first reported use of reaction (2) for the determination of arsenic species in solution. Optimization of the method for arsenite ion in base was focused on solution (pH and supporting electrolyte concentra- tion) and detection (deposition potential, deposition time and stripping current) conditions. The optimum pH level was found to be 12, as the response was linear from pH 9 to 14 (results not shown).The expected Nernstian cathodic shift in potential was observed at increasing pH levels. Although an improvement of the signal was found under more alkaline conditions, the uncertainty of the response was found to be unacceptable. Either 10 mmol 1 - 1 Na2C03 or 10 mmol 1-1 NaOH was found to provide adequate signal-to-noise ratios. The effect of added supporting electrolyte on the PSA signal was also studied, and an enhancement at concentrations above 10 mmol 1- for KC1 and above 100 mmol 1-1 for KN03 was observed (results not shown). The enhancement is apparently due to the decreasing solubility of dissolved oxygen with increasing ionic strength.23 In the constant-current mode of PSA, dissolved oxygen contributes significantly as a chemical oxidant in addition to the applied constant current.At higher ionic strengths, a lower dissolved oxygen concentration is reflected in longer stripping times. Concern over the oxidation of the surface of the gold working electrode by the formation of chlorine at the auxiliary electrode made KC1 a poor choice.24 The signal-to-noise ratio for KNOT proved to be too low to justify its further use because impurities in the KN03 reagent became troublesome above a concentration of 100 mmol 1-1. Cathodic reduction of arsenite ion at pH 12 indicated an optimum deposition potential of -1050 mV (Fig. 1) and a deposition time of 180 s (Fig. 2). At longer deposition times in the batch cell, a monolayer of metallic arsenic (along with ultra- trace amounts of interfering species) that prevents further amalgamation apparently forms.In the FI system, we observed a linear increase in the signal even at long deposition times ( > 15 min). We suspect that the presence of pure carrier solution in the flow cell during the analytical step as a result of medium exchange is responsible for this behaviour. In this way, trace levels of impurities in the sample matrix that could 'poison' the working electrode surface are absent during the stripping of arsenic. The optimum stripping current was a more complex parameter because negative currents resulted in significantAnalyst, October 1996, Vol. 12 I 1389 signal enhancement. Beyond i = -2 pA, the response was uninterpretable (results not shown). Further, the response under a negative current was more variable, and the associated background (blank) response was approximately tenfold higher than that observed with positive currents.The successful use of negative currents for an enhanced response in other PSA methods25 indicates that the electron transfer kinetics of arsenite ion are apparently too slow for redeposition on the gold surface under a reducing current. The design of the flow cell was important in understanding the reactivity of the gold surface; in a three-electrode cell using 1 mol I-' HCl for the determination of AsIi',6-IO chlorine is generated at the auxiliary electrode concurrently with arsenic deposition at the working electrode. Chlorine can then diffuse to the working electrode, where it will readily oxidize the gold surface, thereby inactivating it.The use of an auxiliary electrode with a salt bridge helps to minimize this effect in a batch cell. Although we avoided the use of chlorides in our system, it is nevertheless useful to have the auxiliary electrode downstream of the working electrode to prevent chlorine oxidation of gold. In fact, the only automated ilow analysis system for trace levels of As"' found in the literature, which used voltametric stripping analysis based on the acidic deposition of arsenic as shown in reaction (l), used a flow cell design incorporating this feature? Development of Flow Injection Method Incorporating a PSA detector in an automated on-line flow analysis system is straightforward and would yield advantages that include higher throughput of samples, improved precision by minimizing the opportunities for human error and a lower risk of contamination. Medium exchange, the performance of the electrolytic stripping step in an electrolyte that is different from the solution (i.e., sample matrix) present during the deposition step, can also be implemented conveniently in a flow system.The miniaturization that is possible with FI methods is advantageous for analytical problems that benefit from compact instrumentation and minimum logistic needs. Although PSA is used widely for batch analysis, it has been commonly implemented in continuous-flow systems for the advantages of automation and reduced risk of sample contam- ination. Potentiometric stripping analysis has been adapted to on-line flow systems and has been demonstrated for more than 12 metal ion determinations, including As"' and AsV.27 Flow injection techniques are used principally as a 'front end' to PSA for the precise handling of microlitre samples in a closed system.In comparison with other flow methods, FI affords higher sample throughput, less complexity and virtually no carryover between adjacent samples. The miniaturization that is possible with FI is also advantageous for analytical problems I 40.0 (d 30.0 a, Q : 20.0 N .- - (d g 10.0 z 0 120 240 360 480 600 720 840 De post ion ti m e/s Fig. 2 Optimization of deposition time. Normalized responses for batch PSA (---) and FI-PSA (-) experiments with a -1250 mV deposition potential (Ed) and a 30 s stripping delay time. that benefit from a small instrument 'footprint' and minimal logistic needs.The sensitivity for an FT-PSA experiment can be improved by using larger sample volumes at slower flow rates, thereby allowing a large sample bolus to pass by the working electrode slowly for a long 'effective plating time'.28 Small sample volumes can be used if, in the extreme case, one stops the flow as the bolus reaches the working electrode (i.e., the 'stopped- flow' experiment).*9 We observed that an additional reason to use stopped flow in PSA is the signal enhancement observed as a function of stripping delay time (i.e., the delay between stopping the flow and beginning the stripping step) (results not shown). The decrease in the diffusion of oxygen to the electrode surface in the absence of convective forces causes the signal enhancement.While other workers have observed the enhance- ment of the PSA signal by delaying stripping by 10-30 s,25,30,31 we found that up to 120 s resulted in a tenfold increase in the response. The wall-jet flow cell design used in this work provides well defined hydrodynamics, low dead volume, high mass transfer rates, high signal-to-noise ratio, a simple design that is easy to maintain, working electrode placement that minimizes ohmic drop and flow interference and a stable reference electrode.32 The optimum flow cell volume was determined to be 1 10 pl, as the linear least-squares regression model for the dependence of the PSA signal on flow cell volume was y (ms) = -0.5.51 (+0.101)x (pg I-') + 576 (+24.1), with a standard error of the estimate ( 3 , ) equal to 1.58 pg 1-1 and a coefficienl of determination ( I ' ) equal to 0.984 over the range 110-330 pl.The ratio of the jet capillary to the spacer thickness was approx- imately 3.2, thereby approaching the accepted criterion (2.3) for creating the 'wall-jet effect' in the flow cell.27 Thinner spaces (i.e., volumes < 110 p1) were impractical because the occa- sional air bubbles that formed in the system could not readily pass through the flow cell. A restrictor (1 m) made of 0.020 in id Teflon tubing placed at the flow cell outlet eliminated the formation of air bubbles in the flow cell. A schematic diagram of the computer-controlled FI-PSA system is shown in Fig. 3. Method Performance We studied possible electrochemical interferences, including AsV, BiI'', Cd", Cu".Hg", Pb", Sb"', Se'" and Sn". Of these, only Bill1, Pb" and Sb"' produced a PSA signal under the conditions used. Bi"' and Pb" will place on to gold adjacent to the As"' peak potential, but both can be baseline resolved from As'll, as shown in Fig. 4 for 10 ppb concentrations of each ion. An increase in the response for both Bi"' and Pb" is apparent in Fig. 4, while the As"' signal is reproducible. This indicates irreversible adsorp- Peristaltic Sample Injection Valve Pump Sample Carrier Fig. 3 Schematic diagram of the FI-PSA instrument. Flow direction is indicated by arrows. The prototype instrument in a custom-designed carrying case for field portability weighs approximately 30 kg.I390 Analyst, O('tohcr 1996, VoE. 121 7 I > < 750.0 a, cn 0 8 500.0 [I tion of Bill1 and Pb" on to the solid gold surface.Typically, stepping the potential anodically removes the plated metal by oxidation into solution,'3 but under the alkaline conditions used here, these metals remained tenaciously adsorbed to the electrode surfdce. The performance of the working electrode following BillL and Pb" experiments could be restored only by repolishing. For this reason, attempts to incorporate Bill' or Pb" into the method as internal standards were abandoned. In Fig. 5 , the signals for 10 pg 1-1 Sb"' and Asi1' are shown. The two Sb"' peaks were unexpected, and the more cathodic peak overlaps with the As"' peak. Our observation of two peaks in the stripping of Sb"' may be the result of Sb(Au) dissolution not being as facile as As(Au) dissolution, the formation of antimonous oxides in solution, the presence of trace levels of impurities or a combination of these effects.Nevertheless, for situations in which Sb"' is present, the anodic peak can be monitored to indicate (and possibly correct for) the degree of Sb"' interference. Under acidic conditions, it has been shown7 - 4000.0 7 I > 3000.0 cn a, v) 0 \ 8 2000.0 [I 1000.0 0 -200 -300 -400 -500 -600 PotentiallmV(versus SCE) Fig. 4 Repetitive determination of a mixture of Pbiii. Bi"l and As"' (10 pg 1-1 each). Batch PSA experiments (12 = 4) at pH 12 using 50 pg 1-' As"', a 120 s deposition time (fd), a +1.0 pA stripping current (i,) and a - 1050 mV IVISZIS SCE deposition potential ( E d ) . For other experimental conditions, see text.1250.0 1000.0 250.0 1 0 i I I I I -200 -300 4 0 0 -500 -600 PotentiallmV(versus SCE) Fig. 5 Interference of Sb'" (---) with the Asrii signal (-). Batch PSA experiments as in Fig. 4 with As1" and Sbl" at 10 pg 1-1 in pH 12 NaOH solution. that Sb"l can be resolved from the As"' peak if high levels of chloride ion ( e . g . , 4 mol 1-' HC1-4 mol 1-1 CaCI2) are present during the stripping step. Thus, another way to circumvent the Sb"' interference is by using medium exchange to change the stripping solution from 10 mmol l-I NaOH to 4 mol 1-1 HC1-4 mol 1-1 CaC12. Medium exchange allows the analyst to control the selectivity of the experiment because the stripping solution can be designed to complex an interfering species, to separate the peak potentials of overlapping analytes or to eliminate electroactive interferenc~s.'~ Our results with medium ex- change under these conditions yielded one Sb[" peak that did not overlap with the As"' peak (results not shown).Nevertheless, because antimony is (like arsenic) normally found in the environment in the higher oxidation state (Sb"), a deposition potential could be selected that would not plate SbV while plating As111. We also observed that although Cu" and Hg" did not give rise to a signal, they attenuated a 10 ppb As"' signal when present at more elevated levels ( > 100 ppb). We corrected this attenuation effect (S45% reduction in signal) by using the method of standard additions when analysing samples of unknown composition. This effect may arise from Cu or Hg depositing on the gold surfxe, which alters the surface and consequently the efficiency of As"' deposition and stripping's Thus, the basic method for As"' described here possesses advantages over other methods performed under acidic condi- tions in that the effect of interfering metals can be eliminated.In the basic method, the medium exchange solution is simpler, as only fresh carrier enters the flow cell, and the solid disc working electrode is easier to maintain than the gold film electrode used in other PSA methods for As"'. We applied the method to the determination of trace levels of 2-chlorovinylarsonous acid (CVAA), the degradation product in the environment of dichloro(2-chloroviny1)arsine (Lewisite), a chemical warfare agent developed during World War I that is of interest to environmental remediation efforts at military installations and to the verification of international arms control agreements36 Lewisite undergoes the following series of reactions.37 CZH2AsC13 + 2 H20 C~H~ASCI(OH)Z + 2 HCI (3) Lewisite CVAA C*H2AsCl(OH)2 + OH- G C2Hz + HAs02 + H20 + CI- (4) In the first step, slightly water-soluble Lewisite (560 mg 1 - l ) hydrolyses to form its geminal diol, CVAA, which is the predominant form of Lewisite found in the environment.38 The kinetics of reaction (2) were too fast to study under the conditions used.The concentration range for measuring CVAA in aqueous samples was linear up to 100 pg 1 - I . The RSD at 5 pg 1-1 (n = 8) is <5%. A typical calibration model using a least-squares regression was y (ms) = 24.9 (k0.338)~ (pg I-I) + 0.357 (+1.70), I - = 0.999, and s, = 5.67 pg 1-1 over the range 0-100 pg 1-1 ( n = 4 at each level).The application of a more sophisticated regression model to fit better the slight parabolic curvature of the data has been suggested,g although the use of a linear model appears to be adequate. The average detection limit for three different calibrations over a 3 month period was 0.2 10 f. 0.173 pg 1- 1; the average sensitivity over the same period was 28.4 If: 3.55 pg I-'. The carryover between a 100 pg 1-1 CVAA sample and a subsequent blank injection is < 1% (results not shown). Five minutes are required per sample determination, which includes 60 s for working electrode conditioning, 180 s for deposition and approximately 60 s for stripping of arsenic metal from the gold disc and data analysis.Using a four-point calibration and periodic quality control samples, approximately 30 samples (100 pl each) can be processed in (duplicate) per 8 h period, generating less than 250 ml of waste under these conditions. We are currently studying the application of thisAnulyst, October- 1996, Vol. 121 1391 method to other arsenicals of environmental interest in a variety of sample matrices. This work was supported by the Arms Control Technology Program, Defense Special Weapons Agency, US Department of Defense, under contract 953010. We thank Dr. H. J. O’Neill (Energy Systems Division) for providing CVAA standards. Argonne National Laboratory is operated by the University of Chicago for the US Department of Energy under contract W- 3 1 - 109-Eng-38.References 1 2 3 4 5 6 7 8 9 10 I 1 12 13 14 15 16 Metals and Their Conipounds in the Eni~iwnnient, ed. Merian, E.. VCH, Weinheim, 1991. Irgolic, K. J., in Hazarrloirs Mcfals in the Eni~irwimcnt, ed. Stoeppler, M., Elsevier, Amsterdam. 1988, ch. 8. Lown, J. A., and Johnson, D. C., Anal. Chini. Acttr, 1980, 116, 41. Holak, W., Anc-rl. Chem., 1980. 52, 2 189. Forsberg, G., O’Laughlin, J. W., Megargle, R. G., and Koirtyohaim, S. R., Anal. Chcni., 1975, 47, 1586. Lexa, J., and Stulik, K.. Talanttr, 1983, 30, 845. Huiliang, H., Jagner. D.. and Renman, L., Anal. Chim. Actu, 1988, 207, 37. Jagner, D., Sahlin. E., Axelsson, B., and Ratana-Ohpas, R.. Anal. Chim. Acta, 1993. 278, 237. Jagner, D., Renman, L., and Stefarsdottir, S.H., Elc.ctroanalysis, 1994, 6. 20 I. Miller, E. L., Quantifyin<q Arsonic in Aqueous Solution by Anodic Stripping Potcntiometry, EPA Draft Method 7472, US Environ- mental Protection Agency, Washington, DC, 1994. Aldstadt, J. H., and Dewald, H. D., Anal. Cheni., 1993, 65, 922. Bruckenstein, S., and Nagai, T., Anul. Chcm., 1961, 33, 1201. Jagner. D., and Graneli, A., Anal. Chim. A(.ta, 1976, 83, 19. Hussam, A., and Coetzee, J., A m l . Chcm., 198.5, 57, 581. Smith, R. M., and Martell, A. E., Critical Stability Constants, I)701ume 4: Inor-ganic Coniple.res, Plenum Press. New York, 1975, p. 132. Tomilov. A. P.. and Chomutov, N. E., in Encyclopedia of Electro- chernistr:). qf’the Elements. ed. Bard, A. J., Marcel Dekker, New York, 1974, vol.2, pp. 4 3 4 5 . 17 18 19 2 0 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 Aldstadt, J. H., King, D. F., and Dewald. H. D., Analyst, 1994, 119, 1813. Kolthoff, I. M., Sandcll, E. B., Meehan, E. J., and Bruckenstein, S., Quantitative Chemical Analysis, Macmillan, New York, 1969. Van Muylder, J., and Pourbaix, M., in Arlns nf Elec~tl-crc~licrnir~al Eyuilihria, ed. Pourbaix, M., Pergamon Press, Oxford, 1966, Section 18.3. Vasilyeva, E. G., Zhdanov, S. I., and Krjukova, T. A., E / P ~ - tivkhiriii-ya, 1968. 4, 25. Cozzi, D., and Vivarelli, S., Anal. Chini. Acta, 1951, 5 , 215. Kothegarov, V. M., and I_.omakina. T. P., Elrktrokhinriyu, 1966. 2, 1220. Dyrssen, D., Eskilsson, H . , and Haraldsson, C., .I. E/(~c~trouna/. Chenz., 1989, 262, 161. Adams, R. N., Elec~trochmzistlg at Solid Elec~tmdcs, Marcel Dekker, New York, 1969, ch. 7. Zie, Y., and Huber, C. O., Anal. Chini. Acta, 1992, 263. 63. Wang, J.. and Greene, B., .I. Elerti-ounal. Chem., 198.1, 154, 261. Luque de Castro, M. D., and Izquierdo. A., Elec,ti-oaiicrl~sis ( N Y ) , 199 I , 3, 457. Hu. A., Dessy, R. E., and Grankli, A., Anal. Cheni., 1983, 55, 320. Anderson, L., Jagner. D.. and Josefson. M., Atiol. Chem., 1982, 54, 1371. Mannino, S., Anulyst, 1984, 109, 905. Locascio, L. E., and Janata, J., Anal. Cl7in7. Actu, 1987, 194, 99. Gunasingham, H., Trends Anal. Chm?.. 1988, 7, 2 17. Jayaratra, H. G., Curt.. Sep., 1994, 12, 173. Christensen, J. K., Kryger, L., and Pind, N., Anul. Chim. Ac,ta, 1982, 136, 39. Wang, J., Stripping Analysis: Principles. Iiistl-unzentation, and Applications, VCH, Deerfield Beach, FL, 1985. Robinson, J . P., The Prohleni of Clzenzicul and Riologicul Wcirfarc: the Rise of CB Weapons, Stockholm International Peace Research Institute, New York, 197 I . Waters, W. A., and Williams, J. H., .I. Chrm. So(.., 1950. 18. Clark, D. N., Rcview, of Rractions of Chcnzicul Agents in Wafer, Battele Memorial Institute, Columbus, OH, 1989. Pupel- 6101896C Received March I9, I996 Accepted July 5, I996
ISSN:0003-2654
DOI:10.1039/AN9962101387
出版商:RSC
年代:1996
数据来源: RSC
|
15. |
Simultaneous assay of nitrite, nitrate and chloride in meat products by flow injection |
|
Analyst,
Volume 121,
Issue 10,
1996,
Page 1393-1396
I. M. P. L. V. O. Ferreira,
Preview
|
PDF (723KB)
|
|
摘要:
Analyst, October 1996, Vol. 121 (1393-1396) 1393 Simultaneous Assay of Nitrite, Nitrate and Chloride in Meat Products by Flow Injection I. M. P. L. V. 0. Ferreira", ,J. L. F. C. Lima", M. C. B. S. M. Montenegro",* R. Perez Olmosh and A. Riosb (I CEQUPIDepar-tal??eIito u'e Qui'mic-a Fisica, Faculdade de Farmkia (UP), Rua Aizi'bal Cuuha, 164, 4050 Porto, Portugal T k n i i u Industi-ial, Plaza de la Castilla, No. 3, 48012 Bilbao, Spain Departanziento de Quimica Anali'ticu, Escuela Univcrsitdria de Ingenieria A flow injection (FI) analytical method for the simultaneous assay of nitrite, nitrate and chloride in meat products is reported. The method is based on the potentiometric determination of chloride using a tubular ISE and on the spectrophotometric determination of nitrite. The FI system consisted in splitting the flow after potentiometric detection using a tubular detector and the subsequent confluence of the flow before reaching the spectrophotometric detector.This allowed the reduction of nitrate to nitrite in part of the sample plug on an on-line copper cadmium reductor column. Since each channel had a different residence time, two peaks were obtained for nitrite and nitrite plus nitrate. Spectrophotometric determination was made after a diazotization-coupling reaction. The results obtained were in good agreement with reference procedures and showed adequate precision (RSDs less than 6% for chloride and nitrite and 2% for nitrate). A high sampling rate was obtained (120 determinations per hour corresponding to 40 samples per hour). Keywords: F~OMJ injection; nitrite; nitrate; chloride; meat products; spectrc,pl.lntomett~; potentiomctry Introduction Nitrate and nitrite, as inhibitors of potential pathogenic microorganisms, are substances commonly used in meat products as additives.'-3 Their toxic effects, derived from the formation of nitrosamines, require their frequent determination in the food industry." Common salt is usually added to cured meat to preserve it and improve its taste. Since a correlation between salt ingestion and hypertension has been found, its control in food products is therefore justified.The reference method for the determination of chloride in food products5 is liable to many interferences and is based on a potentiometric titration using a second kind of chloride electrode as indicator.The official methods used for nitrite and nitrate quantification637 recommend spectrophotometric analyt- ical methodologies of tedious execution, involving a high consumption of reagents and their handling, incurring consider- able cost. Flow injection (FI) analytical systems incorporating tubular selective electrodes allow combination with other on-line detection systems for multidetermination, since the aforemen- tioned detectors do not produce significant changes to the flow hydrodynamic characteristics.x In addition to an improvement in the sample processing there is also a significant reduction in reagent consumption and it is possible to increase the analysis ' To whom correspondence should be addressed. efficiency, since several sequential determinations can be carried out on the same sample.Although FI sequential analyses have recently been fre- quently used for the quantitation of the above species, the analyses are usually confined to two determination maxima on the same sample.9--'* The developed systems have been used in the determination of nitrate and nitrite in different matrices and used two different spectrophotometric detectors9 or two injec- tion valves coupled in the interior of the system, with a reductor column incorporated in the loop of the second valve,'O a commutator injector operating in two positions' or a reagent injection into the sample carrier system.I2 However, the determination of more than two parameters on the sample has not been referred to before. This work reports the development of an automatic methodology using FI, to perform the simultaneous assay of nitrate, nitrite and chloride in meat products.The determination of chloride was carried out using a tubular potentiometric chloride detector, with a homogeneous crystal- line mernbrane.I3 In the developed set-up, the sample plug is split into two streams after the potentiometric determination, which prevents significant changes in the sample plug, and each stream is allowed to flow through two channels with different diameters and lengths, one of them containing a copper- cadmium reductor column. Afterwards, confluence of the flow occurs and the colour developing reagent is added before the flow reaches the spectrophotometric detector. Since each channel has a different residence time, two peaks are obtained for nitrite and nitrite plus nitrate.Therefore, it is possible to perform the appropriate sample treatments of each detection system inside the FT system, except for preparation of the extracts, allowing the simultaneous assay of the above spe- cies. Materials and Methods Reagents and Solutions All reagents were of analytical-reagent grade and the de-ionized water had a specific conductivity less than 0.1 pS cm-I. Chloride standard solutions (0.1 mol 1 - I ) were prepared from solid sodium chloride previously dried in an oven at 100 "C for at least 1 h. More dilute standard solutions, used in the calibration curves, were obtained from the concentrated solutions by dilution. Standard solutions of nitrate ( 1 ml = 1000 mg N03--N) and nitrite (1 ml = 1000 mg N02--N) were prepared by dissolving 6.07 and 4.92 g of NaNO3 and NaN02 (dried for I h at 100 "C), respectively, carefully weighed and diluted to 1000 ml.The nitrite solution was standardized against a 0.1 mol 1-' permanganate solution. The solutions were treated with some chloroform drops to prevent the development of microorgan- isms and were stored in a refrigerator. Working standard1394 Analyst, October 1996, Vol. 121 solutions containing nitrate and nitrite were prepared by appropriate di 1 uti ons. The sample carrier solution used in the FI manifold consisted of 5 X 10-2 mol 1-1 sodium sulfate and 2 x 10-5 mol 1-1 sodium chloride. A buffer solution was introduced through the auxiliary channel (R2), and was prepared by dissolving a mixture of 100 g of ammonium chloride, 20 g of sodium tetraborate and I g of Na? EDTA in 1000 ml of water.This solution remains stable for a long time. I I The colour reagent solution was prepared by dissolving 20 g of sulfanilamide and 1 g of N-( 1 -naphthyl)ethylenediamine dihydrochloride i n 100 ml of 80% phosphoric acid and diluting to 1000 ml with water." This solution was stored in a refrigerator. The cadmium column used in the reduction of nitrate to nitrite was prepared as described elsewhere14 by using a glass tube (3 mm id) filled with cadmium-copper filings, held in position by glass wool plugs. Apparatus and Electrodes The absorbance was measured with a Hitachi U 2000 UV/VIS spectrophotometer, equipped with an 8 pl Hellma type 178.713 flow cell, connected to a BD 1 12 Kipp & Zonen (Bohemia, NY, USA) recorder with two needles.A Crison model 2002 potentiometer (sensitivity o f f 0.1 mV) also coupled to the above recorder was used to perform the potentiometric measurements. A tubular electrode, with a homogeneous crystalline mem- brane sensitive to the chloride anion, was used as potentiometric detector13 with an Orion (Cambridge, MA, USA) 900200 as reference electrode. The support of the reference electrode and the earth connection contact as well as the support of the tubular electrode sensitive to the chloride anion were constructed as described elsewhere. I x I 5 The potentiometric titrations, used as a reference method in the determination of the chloride anion, were carried out in an automatic titration system composed of a Crison model Micro BUR 203 1 burette, controlled by a Hyundai computer equipped with a Advantech Model PCL 720 card and connected to an Epson LX 800 printer.A Metrohm 6.0762.100 Ag/AgCl electrode was used as the reference electrode and a silver cation- sensitive electrode with an homogeneous crystalline membrane as indicator electrode.16 An Ismatec model S 820 peristaltic pump was used in this FI system. The insertion of samples and standards into the system was carried out with a Rheodyne SO20 six-port valve. Omnifit Teflon tubings (0.8 and 0.3 mm id), Gilson (Worthington, OH, USA) end fittings and connectors were used. Sample Preparation Different types of meat products, including ham, sausages, smoked pork sausages and smoked ham were purchased at random from the local market. Samples were pre-treated by homogenization in a triturator according to the process recommended by the International Standards Organisation.6.7 Samples were extracted with hot water, followed by purification and filtration.Reference Methods As no certified reference materials were available and in order to evaluate the accuracy of the results obtained by FI, the reference methods of the International Standards Organisation IS0 3091-197s and IS0 2919-19756-7 were used for the determination of nitrite and nitrate levels, respectively, in meat products. For the chloride determination in the same products, the referred norms do not mention any specific method. Therefore, the process described by AOAC, relating to food products,s which is based on potentiometric titration using a silver cation solution as titrant was used.A selective electrode with a homogeneous crystalline membrane sensitive to silver cations was used as indicator electrode in this potentiometric titra- tion.16 Results and Discussion Determinations of chloride, nitrite and nitrate in meat products without any sample pre-treatment other than the preparation of the extracts, required the development of a flow/ manifold similar to that represented in Fig. 1. The FI system was constructed to enable the sequential measurement of three chemical parameters while allowing the performance inside the system of all procedures of the sample adjustment to each measuring system used. These procedures included the reduc- tion of nitrates to nitrites, which was accomplished by circulating part of the sample plug through a copper-cadmium reductor column, by the colorimetric reaction with N-( 1 -naph- thy1)ethylenediamine dihydrochloride, and also by the adjust- ment of the ionic strength of the standards and samples. In order to perform chloride, nitrite and nitrate measurements with high sensitivity and with a good sampling rate, some of the FI manifold parameters were optimized, namely, the injection volume, the flow rate and the sample dispersion level.Optimization of FI Manifold To obtain a better response performance from the FI system (high sample throughput, adequate sensitivity and working range and minimum waste of carrier solutions), a univariant optimization procedure was applied by varying the injection volume, the flow rates of the RI, R2 and R3 solutions and the lengths of the coiled tubes (L1, L2, Ls, Lq, Ls). The tubular configuration of the detector sensitive to the chloride anion allows the sample plug, in which the poten- tiometric measurement will be performed, to flow through it without changing the flow hydrodynamic characteristics. More- over, potentiometric determination does not require a sig- nificant change of the chemical composition of the sample but only an ionic strength adjustment. These two aspects enable, after the measurement of the chloride anion by the tubular detector, the sample plug to reach the point of the system where nitrite and nitrate are determined, without undergoing sig- nificant chemical variations.R31 Q3 I Fig. 1 F1 manifold used in the sequential multiparametric determination of chloride, nitrate and nitrite in tneat products. V,, injection volume; B, peristaltic pump; Li, length of the reaction coils (cm); Q,. flow (ml min-I), Q1 = 2.9, Q2 = 0.5, Q3 = 1.0; ER, reference electrode; DT, tubular potentiometric detector; P, decimillivoltmeter; R, double needle recorder; ET, earth contact; CR, copper-cadmium reductor column; D, spectrophoto- metric detector; X and Y, confluences; R , , 5 x 10-2 mol 1- sodium sulfate solution and 2 X 10-5 mol I-' sodium chloride; RZ, a solution containing 100 g of ammonium chloride, 20 g of sodium tetraborate, 1 g of Na2EDTA in 1 1; Rj, a solution containing 20 g of sulfanilamide, 1 g of N - ( 1 -naphthyl)ethylenediamine dihydrochloride and 100 ml of 80% phos- phoric acid diluted to 1 1 with water.Analyst, October- 1996, Vol.121 139s - ____ _ - _ - ~ ~~~~ _.__ ~ The FI system was optimized in order to allow, with the highest sampling rate, determinations of the chloride anion within the concentration range 10-2 to 10-1 mol 1-1, determinations of nitrate within the concentration range 1 to 4 ppm in N03-N and of nitrite within the range 0.03 to 0.2 ppm in N02-N. The sample carrier solution of the established set-up, composed of S x 10-2 mol 1-1 sodium sulfate and 2 X 10-5 mol 1 - sodium chloride, was selected with the aim of enabling the adjustment of the ionic strength and of stabilizing the line base of the potentiometric measurement. The reference elec- trode was placed in a lateral channel by a confluence, in which the solution worked as a salt bridge between the indicator and the reference electrode.The sequential arrangement of the tubular detector and the reference electrode could produce a non-controlled sample dispersion, which could affect the reproducibility of the subsequent determinations of nitrite and nitrate by spectrophotometry, besides causing an excessive dilutidn of the sample plug. The injection volume used (180 pl) derived from the optimization based on a compromise between the best sampling rate, a good reproducibility and the highest sensitivity. The highest injection volumes of the sample caused two peaks for nitrate and nitrite that slightly overlapped and therefore were difficult to quantify.The total separation of the peaks was only possible if the length of the tube (L4) which included the reductor column was greatly increased, which would com- promise the sampling rate. Teflon tubing with a 0.8 mm id was used in the manifold, except after the splitting of the stream, where tubes with different diameters were tested. When the diameter of the tube which did not incorporate the copper-cadmium column (L3) was reduced, the sensitivity of nitrite plus nitrate determination increased. A diameter of 0.3 mm was enough to obstruct the sample flow through that part, compelling its flow through the reductor column. In order to minimize the sample dispersion level, before spectrophotometric detection, the lowest possible length (30 cm) of Ll was selected allowing, however, the adjustment of the ionic strength by mixing with the carrier solution, and therefore ensuring good reproducibility.After the potentiometric measurement, the sample plug was mixed along the reaction coil (L2) with R2 reagent, which was used as buffer and catalyst of the reduction reaction. The length of 20 cm established for L2 (see Fig. 1) was enough to promote the mixture of the sample plug with R2 reagent. The spectrophotometric measurements of nitrate and nitrite were performed on the sample plug after the system was split through two channels with different lengths and diameters, one of which included a copper-cadmium reductor column. After- wards, the confluence of the flow occurred just before the colour reagent was added through the R3 auxiliary channel.Since each channel had different lengths and ids, the residence time of the flow in each one was different and therefore two peaks were obtained for nitrite and for nitrite plus nitrate. At point Y, the colour-developing reagent, responsible for the diazotization that occurred along L5 reaction coil, was added. After the reaction, the measurement of the absorbance peak relating to the nitrite level was performed first and then that corresponding to the nitrite plus nitrate levels. In the tests carried out with the purpose of optimizing the L? and L4 tube lengths, when the L3 tube length was reduced, the separation of the peaks for nitrite and nitrite plus nitrate was facilitated. In the same way, when increasing the L3 tube length (which contained the reductor column), the separation of the peaks was also improved.However, as the L4 length increased, its peak increased too, whereas the second peak besides becoming shorter also become wider, which means that there was a dispersion increase and consequently a decrease in the sensitivity of the determination. Moreover, the sampling rate decreased. The length for L5 (1 00 cm) was selected after its optimization so that the colour-developing reaction for the spectropho- tometric detection was quantitative, thus enabling the highest sensitivity. When shorter lengths (50 and 70 cm) were used, the > E 0 c\I - t c 0) cn .- 1 min Time + Fig. 2 Record of a typical calibration curve and of a series of concentration samples of the injected solutions.A, 0.05 ppm “I2-- N + 1 ppm NO?--N + X x mol I-’ C1-; R , 0.05 ppm N02--N + 2 ppm NO?--N + 2 x 10-2 moll 1 C1-; C, 0.05 pprn N02--N + 3 ppm NO?--N + 5 X 10-2 moll-’ CI-; D, 0.05 ppm N02--N + 4 ppm NO? -N + 8 X lo-* moll C1-; E, 0.03 ppm N02--N; 1 to 5, samples. Table 1 Results obtained for chloride, nitrate and nitrite determination in meat products by FI and reference methods Chloride f Nitratel Nitrites Sample* Cf c r S ct c, \ ct c, J 1 2 3 4 5 6 7 8 9 10 45.12 27.76 42.48 64.96 21.81 41.81 73.27 42.83 69.36 109.2 44.83 28.4 I 4 1.54 63.06 23.19 40.18 70.27 41.12 70.47 110.1 0.58 -2.28 2.26 3.0 1 -5.95 4.06 4.27 -0.82 4.16 - 1.58 40.48 42.03 31.97 40.45 31.37 35.05 47.69 36.76 26.52 3 1.06 40.090 42. I20 3 1.670 39.950 3 1.560 35.070 47.240 37.450 27.190 3 1.820 0.970 -0.210 0.950 I .250 -0.600 0.950 0.953 - I .840 -2.046 -2.380 0.98 1.15 2.07 1.33 1.51 1.49 1.01 0.907 0.9 1 I .35 0.93 1 1.13 2.02 1.275 1.59 1.53 0.982 0.956 0.954 1.275 5.26 1.77 2 .4 3.38 --5.03 -2.6 1 2.85 --5.13 -4.6 1 5.88 * Smoked pork sausages (samples I , 9); sausages (samples 2, 3, 6); ham (samples 4, 5 , 7); smoked ham (samples 8, 10). All analyses were carried out in duplicate. t Concentration expressed in g of NaC1/1000 g of meat product. l Concentration expressed in mg of KNO3-N/1000 g of meat product. 3 Concentration expressed in mg of NaN02-N/1000 g of meat product.1396 A i d y s t , October 1996, Vol. 121 colour development was incomplete, whereas with longer lengths (1 50 cm) the analytical signal diminished significantly because the sample dispersion increased.Q l , Q2 and Q3 flow levels were selected so that the flow within the spectrophotometric cell did not exceed 4.5 ml min- I , in order to avoid overpressurization in the manifold. The flow of R3 colour-reagent solution was established using a 0.2 ppm N02-N solution and varying the flow from 0.5 to 1 .S ml min-1, thus obtaining a maximum absorbance within a 0.9-1.1 ml min-1 range. Analytical Applications After being optimized, the developed manifold enabled analy- ses to be performed within a linear response range from 10-2 to 10-1 mol I-' for chloride, I to 4 ppm for NO-3-N and from 0.03 to 0.2 ppm for N02--N, with a sampling rate of 40 samples per hour, corresponding to 120 determinations per hour. Fig. 2 represents an FI record corresponding to a linear calibration graph for chloride, nitrite and nitrate, and several injections of samples of different meat products, the preparation of which was mentioned above.The results obtained are also shown in Table 1 as well as the values obtained from the analysif of the same products using the reference proce- dure~"6.7 With the purpose of evaluting the accuracy of the FI methodology, a correlation between the values obtained from the FI method (C,) and those given by the respective reference methods (C,) was established in 10 samples of meat products. By establishing a linear regression (C, = C, + SC,) between the results given by the FI and corresponding reference methods, the intercept value (C,,), the slopes (S) and 14 obtained for each species, represented in Table 2, indicate that the results are in reasonably good agreement, an average of relative error less than about 6% for chloride and nitrite and 2% for nitrate being obtained.The precision of the results obtained by FI analysis was assessed by measuring the RSD corresponding to 10 consecu- tive determinations of nitrate, nitrite and chloride levels in a sample of intermediate concentration. RSDs of 2.5% for chloride, 0.7% for nitrate and 1.1% for nitrite were obtained. Conclusions The F1 system proposed for the multiparametric analysis of nitrate, nitrite and chloride in industrial meat products consisted of an on-line potentiometric tubular detector and a spec- trophotometric detector; a non-complex manifold, easy to operate, robust and fairly economic.The method compares well with reference methodologies usually applied to the analysis of these products, since it avoids sample pre-treatment, as it is Table 2 Comparison between the results for chloride, nitrate and nitrite determination in 10 samples of meat products by FI (Cf) and the Reference procedures (Cr) C,' S" I' Chloride 0.399 1.00 0.997' Nitrite 0.0005 1.00 0.9836 Nitrate - 1.78 1.04 0.9917 -* Parameters of the equation C,- = C, + SC, (see text). adjusted to the measurement system used by the manifold itself. Moreover, while the reference procedure for the determi- nation of nitrite and nitrate requires control of the time of the colour development, which affects the measurement reprodu- cibility, the method proposed in this work performs the determination within a constant time interval, dependent on the flow rate of the FI system.These aspects result in smaller analysis times, a sampling rate of 40 samples per hour being obtained. As the amount of the sample that flows through the cadmium reductor column is much less in the proposed method than that used by the reference method, the durability of the column is consequently higher. The quantity of material necessary to prepare the sample is greater in the reference method, which increases the cost of the determinations. The sensitivity of the proposed method is high and the results given by the FI method are comparable with the corresponding values given by the reference method. Additionally, the F1 method gives more accurate results with relative errors of less than 6% for chloride and nitrite and 2% for nitrate.This work was supported by the Research of the Basque Government, Spain, Project PI 94/08. References 1 2 3 4 5 6 7 8 9 10 1 1 12 13 14 15 16 Casas, C., Sanchez, G., Moreno, T., and Sanz, B., Aliment. Eyitipos Tecnol., 1991, September, 127. Report of the Scientific. Comniittec for the Human Nutrition about Nitrates and Nitrites, Serial no. 26, Report EUR 1391 3 PT issn 1018. Bosch, N., Martinez Alvarez J. R., and Perez Rodriguez, M. L., An. Bromatol., 1993, XLIII 2-3, 215. Schweinsberg, F., in Catalvsis c$ Nitrosamine Synthesis, ed. Bo- govski, p., and Walkers. E. A., Nitroso Compounds in the Environment, International Agency for Research on Cancer, Lelio, Association of Official Analytical Chemists (AOAC), Ofliciul Methods uf Analysis, Association of Official Analytical Chemists, Washington, DC, 14th edn., 1984. International Standards Organisation, Viands et produits ci base de limn&. I>r'ter-minution de la teneur en nitrites. Methode de reference, IS0 2019, 1975. International Standards Organisation, Viands et produits d base de viande. De'ternzination de la teneur en nitrates. Mkthode de refkrence, IS0 3001, 1975. Ferreira, 1. M. P. L. V. O., and Lirna, J. L. F. C., .I. Flow Injection Anal., 1993, lO(1). 17. Anderson, L., Anal. Chim. Acta, 1979, 110, 123. Bermudez, B., Rios, A., Luque de Castro, M. D., and Valcarcel, M., Tula/?ta, 1988, 3 3 lo), 8 10. Gink, M. F., Bergamim F , H., Zagatto, E. A,. and Keis, B. F., Anal. Chim. Acta, 1980 114, 191. Johnson, K. S., and Petty, R. L., Limnol. Oceunogi.., 1992, 28(6), 1260. Ferreira, I. M. P. 1.. V. O., Lima, J. L. F. C., and Rocha, L. S. M., Fresenius' J . Anal. Chem., 1993, 374, 3 14. Henriksen, A., and Selmer-Olsen, A. R., Anulyst, 1970, 95, 514. Alegret, S., Alonso, J., Bartroli, J., Machado, A. A. S. C., Lima, J. L. F. C., and Paulis, J. M., Quim. Anal., 1987, 6, 278. Ferreira, I. M. P. L. V. O., Lima, J. L. F. C., and Rangel, A. 0. S. S., Food Chem., 1994,50, 324. 1974, 80-85. Paper 6/02 728H Receii'ed April 18, 1996 Accepted .June 21, I996
ISSN:0003-2654
DOI:10.1039/AN9962101393
出版商:RSC
年代:1996
数据来源: RSC
|
16. |
Turbidimetric flow method for the enantiomeric discrimination ofL- andD-aspartic acid |
|
Analyst,
Volume 121,
Issue 10,
1996,
Page 1397-1400
Monika Hosse,
Preview
|
PDF (676KB)
|
|
摘要:
Anulyst, October 1996, Vol. 121 (1397-1400) 1397 Turbidimetric Flow Method for the Enantiomeric Discrimination of L- and D-Aspartic Acid Monika Hosse, Evaristo Ballesteros, Mercedes Gallego and Miguel Valcarcel* Department qf Anulytical Chemistry, Faculty oj’ Sciences, University of Cdrdoha, E-14004 Cbrdoba, Spain The proposed enantiomeric resolution of aspartic acid is based-on the inhibition of the crystallization of L- and D-histidine. A flow-through system permits the turbidimetric multi-detection of the signal produced in the crystallization of histidine from a supersaturated solution. The presence of L- or maspartic acid delays the growth of L- or D-histidine crystals, respectively, the delay being proportional to the concentration of aspartic acid. Calibration graphs are linear down to 40 mg 1 - 1 of L- and D-aspartic acid, with a precision (repeatability, as RSD, n = 11) of 2.5%.The method was applied to the determination of I,-aspartic acid in pharmaceutical preparations (spiked with D-aspartic acid) and the resolution of a racemic sample of D,L-aspartic acid. The results obtained were consistent with the nominal contents. Keywords: Tui-hidimetry; jlow-tlil-ougli system; L- and D-aspartic acid; c h i d resolution ; pharmaceuticul prepaimkm; racemic. sunzple Introduction From Pasteur’s very first optical resolution of a racemate’ to today’s fast chromatographic techniques, there has been a formidable accumulation of stereochemical knowledge. The importance of optically active compounds for the elucidation of reaction mechanisms and the dynamic behaviour of chiral molecules in organic chemistry has grown dramatically.The pharmaceutical industry is becoming increasingly interested in methods for resolving racemates into optical antipodes in order to be able to subject them individually to pharmacological testing2 In recent years, HPLC has bccome the most popular choice for the resolution of enantiomeric compounds.3--” However, this technique has some disadvantages, such as the high price of chiral stationary phases. Alternative techniques for the resolu- tion of racemic mixtures are based on fractional crystallization of conglomerates. Growing crystal surfaces can be thought of as being composed of ‘active sites’ that interact stereospecifically with molecules in solution, in a manner similar to the interactions of enzymes and substrates or antibodies and antigens.6 The molecules that interact with the active sites are known as ‘tailor-made’ inhibitors. Grases and March7 reviewed the potential applications of these crystallization inhibitory processes in analytical chemistry.They have developed various methods for the determination of L-amino acids in which a supersaturated solution of a substrate is obtained by changing the solvent composition; the analyte (L-amino acid) inhibits crystallization of the substrate.8.9 The applicability of enantiose- * To whom correspondence should be addressed. lective biosensor systems depends on the availability of suitable group-specific/enantiospecific and D- and I -enantiospecific enzyme pairs. These enzymes can be combined with suitable potentiometric, amperometric, calorimetric and optical trans- ducers.Schugerl et al. l o recently discussed the potential of biosensors for enantio-sensitive analysis with non-enantio- meric, enantiomeric and D- and L-enantiomeric enzyme pairs. They reported that biosensors for enantiomeric analysis can be used for process monitoring and control of the enantiomeric purity of organic chemicals, whereas HPLC is suitable for quality control of the product but not for process monitoring. Recently, our group developed several turbidimetric methods for the continuous determination of s,-lysine. 1,-arginine and I.- ornithine in pharmaceutical preparations.] 1 . i 2 in this work, two enantioiners (L- and D-) of aspartic acid were discriminated by their inhibitory effect on the crystal growth of L- and I)- histidine, respectively.For this purpose, a continuous method was developed that permits the sequential determination of L- and maspartic acid in the presence of other amino acids with no need for a prior separation. The method was applied to the analysis of pharmaceutical samples, where the determination of L-aspartic acid is of special interest. Since the method is integrated in a flow-through system, it can be implemented on- line for process control purposes. Experimental Chemicals and Apparatus Propan-2-01, ethanol, methanol and acetonitrile, all of HPLC grade, were purchased from Romil Chemicals (Loughborough, UK). Amino acids (1,- and maspartic acid, racemate of I),[,- aspartic acid, L- and D-histidine and the other amino acids used for the study of interferences) were supplied by Sigma (St.Louis, MO, USA). Sodium hydroxide and hydrochloric acid were obtained from Merck (Darmstadt, Germany). Stock standard solutions containing 1 g 1-1 of L- or 11- aspartic acid were prepared in Milli-Q-purified water and stored -stoppered bottles. Solutions containing 3.0 or 2.8 g I-’ of L,- or D-histidine were used as substrates. These solutions remained stable for at least 1 week. Turbidimetric measurements were made on a Unicam 8625 UV/vIS spectrophotometer (Unicam, Cambridge, UK) equipped with a Hellma (Jamaica, NY, USA) flow cell (path length 10 mm, inner volume 18 ~ 1 ) . The ab5orbance was continuously recorded at 550 nm by a Radiometer (Copen- hagen, Denmark) REC-80 Servograph recorder. The flow manifold consisted of two peristaltic pumps [ Gilson (Wor- thington, OH, USA) Minipuls-21 fitted with poly(viny1 chlo- ride) and Solvaflex pumping tubes for aqueous and organic solutions, respectively.A Rheodyne (Cotati, CA, USA) Model 504 1 injection valve, two Rheodyne Model 530 1 switching13% Analyst, October 1996, Vol. 121 30 .r 25- E \ -0 a, n .g 2 0 - 1 5 - .- + 0 $ 10- - 5 - valves and PTFE tubing of 0.5 mm id for coils were also used. - Sample Preparation Five tablets (BOl-K aspartic acid, Laboratory BOI, Barcelona, Spain) or ten pills (Aspartono, Laboratory MEDIX, Madrid, Spain), chosen at random from several samples, were placed in a mortar and ground to a fine mesh. A portion of about 0.5 g (pills) or 5 g (tablets) of the resulting powder (containing approximately 300 mg of L-aspartic acid) was accurately weighed and dissolved in I00 ml of Milli-Q-purified water with electromagnetic stirring for 60 min.The solution was filtered, the residue washed with water and the filtrate was diluted to volume with water in a 250 ml calibrated flask. Aliquots of 100-200 p1 of these final solutions were placed in 10 ml calibrated flasks and diluted to volume (pH 3-10) for analy- sis. Procedure for the Resolution of Enantiomers The continuous-flow procedure for the resolution of L- and D- aspartic acid involves several steps (Fig. 1). For the determination of L-aspartic acid [Fig. 1(a)], an aqueous sample containing 3 4 0 mg 1-1 of L- and D-aspartic acid at pH 3-10 is continuously aspirated at 0.3 ml min-1 and mixed with a substrate stream containing 3.0 g 1-1 of L-histidine at 0.3 ml min-1.Then, the mixed stream is merged with another of propan-2-01 circulated at 1.3 ml min- I . The mixed solution is homogenized in coil C2 and then passed through the injection valve (IV). Simultaneously, the open system is filled with carrier (propan-2-01) by means of the second pump (P2) at 0.7 ml min- I . Then [Fig. 1 ( h ) ] , 100 pl of the loop contents of IV are injected into the carrier and SV2 is switched to close the circuit. At that moment, the absorbance curve, which reflects the time course of crystal growth, is recorded at 550 nm. Finally, SV2 is switched to flush the open-closed system with Milli-Q-purified water. For the determination of D-aspartic acid [Fig.l(a)], SV, is switched and the same aqueous sample is mixed with D- histidine solution (2.8 g 1-I). The procedure is then executed as described above for r--aspartic acid. The signal profile obtained during the crystallization of L- or D-histidine is shown in Fig. 1. The induction period ( t ) was used as the analytical parameter. I I L-Histidine D-Histidine R 11 Fig. 1 Manifold for the continuous-flow discrimination of 1,- and D- aspartic acid. ( a ) Introduction of sample (L- and D-aspartic acid), substrate and propan-2-01 into the system. ( h ) Closed system and signal multi- detection of crystal growth of L- or D-histidine. P, pump; SV, switching valve; IV, injection valve; C, coils (C,, C2 and C3,40, 100 and 40 cm long, respectively); W, waste; D, spectrophotometer; R, signal recording; t, induction period.Results and Discussion Selection of the Substrate and Organic Solvent The continous-flow system depicted in Fig. 1 was used to select the most suitable reagent and solvent. Initially, a sample containing 10 mg 1-1 of L- and D-aspartic acid was introduced into the flow system and merged with a solution containing 3.0 g 1 - l of L- or D-substrate solution. Two amino acids (histidine and lysine) were examined as substrates. I,- and D-aspartic acid exhibited an inhibitory effect on the crystal growth of L- and D- histidine, respectively; however, using lysine to determine aspartic acid was inadvisable as L- and D-aspartic acid showed no inhibitory effects on the crystal growth of L- and D-lysine.For this reason, L- and D-histidine were selected as the substrates for the discrimination of L- and u-aspartic acid, respectively . A previous study' 1 revealed that some organic solvents (methanol, ethanol, propan-2-01 and acetonitrile) induce crys- tallization of the substrate. Such solvents, all miscible with water, were used to prepare supersaturated solutions of the substrate. Propan-2-01 was finally selected as the organic solvent because the resulting induction period ( 5 min) was shorter than that obtained with other solvents (longer than 20 min) and the analysis time was thus the shortest. Optimization of the Working Conditions In aqueous solution, amino acids are present as cations. zwitter- ions or anions, depending on the pH; their crystallization is especially favourable in the zwitterion form.The effect of pH on the crystallization of L- and D-histidine was studied between pH 2 and 12 (adjusted with 0.01 mol 1-1 HNO3 or NaOH). For this purpose, aqueous solutions of L- and D-aspartic acid ( 10 mg 1- 1) or blanks (Milli-Q-purified water) at different pH values were processed in the system. As can be seen in Fig. 2, the induction period in the crystallization of L-histidine (for sample and blank) remained constant over the pH range 3-10, outside which the period was considerably longer. Similar results were obtained for the D-enantiomer (Fig. 2). The induction periods were similar (approximately 6 min) for both blanks. A water blank and a sample pH of 5-6 were selected, which were directly obtained by preparing the samples in water.In order to select the best L- and D-histidine concentrations, several calibration graphs for I,- and D-aspartic acids ( 3 4 0 mg 1-1) were run at a various concentrations of the substrate between 2.5 and 3.5 g 1-1. Fig. 3(a) shows the influence of the substrate concentration on the sensitivity (slope of the calibra- B 4 01 I I 1 1 1 3 5 7 9 11 PH Fig. 2 Effect of pH on the crystallization of L- or D-histidine (3.0 and 2.8 g 1 - I , respectively) for a sample containing 10 mg 1-1 of L- and D-aspartic acid (1 and 2, respectively) and for the blanks (water) of the L- and D- enantiomers of histidine (3 and 4, respectively).Anulyst, October 1996, V d . 121 I399 tion graph) of the method. These experiments allowed the following conclusions to be drawn: ( I ) the sensitivity increases as the concentration of substrate decreases; (2) the sensitivity is slightly higher for L-aspartic acid than for the D-enantionier at all substrate concentrations (probably because the crystalliza- tion of whistidine in propan-2-01 is more favourable than that of L-histidine); and (3) increased sensitivity can be obtained at the expense of longer induction periods [Fig.3(h)]. Concentrations of 3.0 and 2.8 g I-' for L- and D-histidine, respectively. were selected because the sensitivity for L- and D-aspartic acid was similar and the sample throughput acceptable. The flow variables affecting the perforniance of the con- tinuous-flow system were optimized by using an aqueous sample containing 10 mg 1- 1 01' 1,- and ixispartic acid.water as the blank and a substrate solution containing 3.0 or 2.8 g I-' of L- or 1,-histidine, respectively. The influence of the flow rates used in this system was studied by varying one at a time while keeping all others constant. Increasing the sample flow rate increased the induction period for the substrate crystallization through an increased concentration of L- and n-aspartic acid and increased dilution of the substrate in the final mixture. Conversely, increasing the substrate flow rate decreased the induction period. Increasing the flow-rate of propan-2-01 had a similar effect to diluting the sample and substrate in the final mixture. Flow rates of 0.3, 0.3 and 1.3 ml min-I were therefore selected for the sample, substrate and propan-2-01, respectively, as compromises between acceptable sensitivity and sample throughput.The effect of the flow rate of the carrier (propan- 2-01) in the open-closed system was studied over the range 0.3-1.5 ml min-I; the signal remained constant above 0.5 ml min-1. A flow rate of 0.7 ml min- 1 was chosen to transfer the contents of the sample loop into the closed system. The crystallization of L- and D-histidine was significantly affected by the injected volume of valve IV throughout the range studied (50-250 pl); the induction period decreased with increasing volume injected through decreased dilution of the sample and substrate. An injected volume of 100 pl was selected. The influence of the length of the coils for mixing the sample and substrate [C, in Fig.I(a)], and both with propan-2-01 (C,) was studied between 25 and 150 cm (0.5 nim id). Lengths of 40 and 100 cm were selected for C I and C2, respectively, as they proved sufficient for homogenizing the solutions. Finally, the length of the coil inserted in the closed system [C, in Fig. I (b)] was varied between 25 and 100 cni. Increasing length of C, increased the dilution of the sample and substrate in the closed system and hence increased the induction period for histidine 0.5 0.4 c '5 0.3 .- c .- v) C $ 0.2 0.1 0 --+ , 2.4 2.6 2.8 3.0 3.2 3.4 Concentration of L - or D-histidine/mg I-' Fig. 3 ( u ) Sensitivity (slope of the calibration graph) for L.- and n-aspartic acid (solid and dashed lines, respectively) and ( h ) induction period for a sample containing 10 nig I - ' of L- and o-aspartic acid at various concentrations of L- or n-histidine.respectively. crystal growth. A length of 40 cm was chosen as a compromise between acceptable sensitivity and sample throughput. Analytical Performance The performance and reliability of the continuous-turbidimetric system (Fig. 1) were tested by determining the sensitivity (slope of the calibration graph), analyte detectability, linearity range and RSD for the determination of L- and D-aspartic acid. For this purpose, solutions containing various concentrations of I.- and waspartic acid were introduced into the system and merged with a substrate stream containing 3.0 or 2.8 g 1-I of L- or D- histidine, respectively. The calibration graphs obtained by plotting the induction period [f, difference between the induction period for the sample and blank (about 6 min), in minutes] against the aspartic acid concentration (mg 1- I ) were t = -0.08 + 0.26 [L-aspartic acid] t = 0.03 + 0.19 [n-aspartic acid] I - = 0.997; linear rangc 3 4 0 mg 1 r = 0.998; linear range 4-40 mg 1-1 The detection limits, calculated a\ three times the standard deviation of the induction period for 10 determinations of the same blank (Milli-Q-purified water), were I and I .8 mg 1.- for L- and D-aspartic acid, respectively. The RSD, obtained by measuring 1 1 samples containing 10 mg 1- I of each enantiomer, was 2.1 and 2.5% for the L- and D-enantiomer, respectively. Study of the Interference of Amino Acids The influence of L-amino acids frequently encountered in pharmaceutical products was studied by adding different amounts of each potential interferent to samples containing 1 0 mg 1- I of L-aspartic acid.As can be seen from Table I , most of the species tested were tolerated at high concentrations in the determination of 1,-aspartic acid. The most serious interferences were those of the L-enantiomers of cysteine, glutamic acid and diaminocarboxylic acids (lysine, ornithine and arginine), which interfered at concentrations below that of L-aspartic acid. The maximum tolerated concentrations of D-amino acids in the determination of 10 mg 1-1 of n-aspartic acid are also included in Table 1. All amino acids that interfered increased the induction period for the substrate crystallization. An amino acid was considered to interfere when the induction period for the 10 mg 1- I standard (about 8 min) was increased by 0.4 min.As can be seen, the greatect interferences were those from diamino- carboxylic acids, as in the determinalion of L-aspartic acid. A more detailed study of potential interferences in the determina- tion of L-aspartic acid was carried out than for D-aspartic acid because pharmaceutical preparations contain predominantly the ].-form. The D-enantiomer did not interfere in the determination of L-aspartic acid at concentrations seven times in excess; on the other hand, L-aspartic acid is tolerated at a concentration only 2.5 times higher than that of D-aspartic acid (analyte). However, this tolerated ratio permits the discrimination of the two isomers. Analysis of Pharmaceutical Preparations and u Racemic Sample Amino acids present in pharmaceutical preparations are L- enantiomers, except for phenylalanine, histidine, methionine and tryptophan, which are also active in the D-configuration.3 Therefore, the proposed method can only be applied to the determination of L-aspartic acid in pharmaceutical samples. Only two commercial samples were available for this purpose, which were dissolved as described under Experimental. Table 2 gives the average results for five determinations of r>-aspartic1400 Analyst, October 1996, VoE. 121 Table 1 Effect of various amino acids on the determination of L- and D-aspartic acid (1 0 mg 1 I ) [.-Aspartic acid determination D-Aspartic acid determination Amino acid Maximum tolerated concentrationlmg 1-' Amino acid Maximum tolerated concentrationlmg 1-1 11-Aspartic acid 70 L-Aspartic acid 25 L-methionine, L-valine, D- Alanine 40 L-Isoleucine, L-phenylalanine, D- V a1 i ne > 100 D- Asparagine 35 L-Alanine, i -leucine 7 s D-Glutaniic acid 5 L-Serine, L-asparagine 50 D-Lysine, n-orni thine, D-arginine 1 L-Threonine, L-glutamine 25 L-Cysteine 5 L -Lysine, L-ornithine 1 .5 L- Arginine 1 L-tyrosine > 100 L-Glutamic acid 4 Table 2 Determination of L- and D-aspartic acid spiked in pharmaceutical preparations L-Aspartic acid D-Aspartic acid Nominal content/ Found/ Added/ Found/ Trade name mg per tablet or pill BOI-K aspktico 350 356+ 10 225 225 f 6 (tablet) 300 290 f 6 mg per tablet or pill mg per tablet or pill mg per tablet or pill 600 610f 12 Aspartono 410 (Pill) 395 + 13 270 360 720 274 k 7 353 * 7 716+ 14 ' 1.-Aspartic is present as asparlate in the original sample.acid and their standard deviations. Because no real samples containing D-aspartic acid were available, the above samples were spiked with this enantiomer. Three standard additions were made for each preparation following dissolution and dilution (the final concentrations of D-aspartic acid in the diluted samples, were 7.5, 10 and 20 mg 1-1). The recoveries obtained from five individual additions of D-aspartic acid were close to 100% in all instances (Table 2). The potential of the proposed method for the discrimination of L- and D-aspartic acid was assessed on a real racemic sample (D,L-aspartic acid from Sigma). For this purpose, solutions containing various concentrations of the racemate were ana- lysed.The mean content and standard deviation ( n = 3 ) of each enantiomer at three different racemate concentrations (20, 30 and 40 mg 1-I) were 10.2 & 0.2, 14.9 & 0.3 and 20.1 & 0.4 mg 1-l (L-aspartic acid): and 9.7 k 0.2, 15.3 f 0.4 and 20.0 & 0.5 mg 1-1 (D-aspartic acid), respectively. Conclusions The inhibitory effect of aspartic acid on the crystallization of histidine can be successfully exploited for the indirect determi- nation of L- and D-aspartic acid. The chief advantages of the proposed method are as follows: ( 1 ) D- and L-aspartic acid can be determined in the same sample simply by changing the reagent (L- or D-histidine, respectively); (2) none of the chiral columns or mobile phases required in HPLC are needed, so analysis costs are reduced; and (3) the method can be used for on-line control of the enantiomeric purity of pharmaceuticals.The principal limitation of the method is that it cannot be applied to amino acid mixtures and can only play a role when the composition of the other amino acids in the material is known . The authors are grateful to the Direccion General de Investiga- ci6n Cientifica y Tkcnica (Project No. PB95-0977) for financial support. References 1 2 3 4 5 6 7 8 9 10 I 1 12 13 Pasteur, L., A m . Chim. Phys., 1850, 28. 56. Allenmark, S. G., Chromut~~grz~phic Enantiosepurotion: Methods urzd Applications, Ellis Horwood, Chichester, 1988. Brueckner, H., Haasmann, S., Langer, M., Weesthauser, T., Witter- ner, R., and Godel. H., J . Chromatogr., 1994, 666, 259. Toyooka, T., and Liu, Y. M., .I. Chromatogr., 1995, 689, 23. Sukbuntherng, J., Hutchaleelaha, A., Chow, H. H., and Mayersohn, M., .I. Anal. To.xicnl., 1995, 19, 139. Weissbuch, I.. Addadi, L., Lahav, M., and Leiserowitz, L., Science, 1991,253, 637. Grases, F., and March, .l. G., Trends Anul. Cheni., 1991, 10, 190. Grases. F., and Genestar, C., Taluntcr, 1993, 40, 1589. Grases, F., Costa-Bauzj, A., Forteza. R., and March, J . G., Afztcl. L c f f . , 1994,27, 278 1. Schugerl, K., Ulber, R., and Scheper, T., Tieizd.7 Anul. Chenz., 1996, 15, 56. Ballesteros, E., Gallego, M., Valcarcel, M., and Grases, F., Anal. Cheni., 1995, 67, 3319. Ballesteros, E., Gallego, M., Valciircel, M., and Grases, F., And. Chim. Acta, 1995, 315, 145. Del Pozo, A., Fcirnzucia GalPnica Esprciul, Romargraf, Barcelona, 1978. Paper 6103425.1 Rec*eivod May I6, 1996 Accepted June 17, I996
ISSN:0003-2654
DOI:10.1039/AN9962101397
出版商:RSC
年代:1996
数据来源: RSC
|
17. |
Selective recovery of uranium(VI) from aqueous acid solutions using micellar ultrafiltration |
|
Analyst,
Volume 121,
Issue 10,
1996,
Page 1401-1405
Edmondo Pramauro,
Preview
|
PDF (742KB)
|
|
摘要:
Analyst, October 1996, Vol. 121 (1401-1405) 1401 Selective Recovery of Uranium(v1) From Aqueous Acid Solutions Using Micellar U It raf i It ra t i on Edmondo Pramauroa, Alessandra Rianco Prevota, Vincenzo Zelanoa, Monica Gulminia and Guido ViscardP u Dipartiniento di Chirnica Aizaliticw, Univei-sit2 di Toi-ino, 101 25 Turin, Italy h Dipartiniento di Chimica Gmerale e Organicu Applicata, Universita di Torino, I01 25 Turin, Italy Preconcentration and removal of uranyl ions from aqueous solutions were achieved by using micellar-enhanced ultrafiltration with complexing micellar aggregates composed of Triton X-100 and different hydrophobic ligands. Derivatives of 4-aminosalicylic acid (PAS) and of 1-(2-pyridylaz0)-2-naphthol (PAN) having tuned alkyl chains were used as suitable chelating agents.The selective recovery of uranyl from acid samples containing also Sr" and Cd" is possible using the multi-step ultrafiltration approach with the PAN derivatives, whereas effective uranyl retention can be obtained with salicylates only in neutral to basic media. Keywords: Chelating nzicelles; micellar ultrafiltration; uranyl recovery Introduction Micellar aggregates have been succesfully employed in differ- ent fields of analytical chemistry and separation science. Their use for environmental purposes, in particular in water decontamination, also offers promising perspectives.* Among the surfactant-based separation techniques, micellar- enhanced ultrafiltration (MEUF) is very interesting as it can be easily applied to recover or remove a variety of solutes by exploiting their binding to suitable micellar aggregates, which are in turn separated from the aqueous bulk using a membrane having an appropriate pore size.For example, metal ions have been separated from aqueous streams using oppositely charged micelles or polyelectrolytes,9--' 1 although this approach suffers from poor selectivity. The application of chelating micelles, formed from usual unreactive surfactants and properly designed amphiphilic ligands having tunable hydrophobic groups, allows one to achieve more selective demetallation treatments. I2-I5 Theoretical models able to explain and predict the performances of such processes have been developed.I6 The aim of this work was to investigate the potential application of MEUF for the preconcentration and selective removal of uranium from dilute acid wastes containing also other metals, as an alternative to the use of classical extraction processes with organic solvents.Previous studies on uranium enrichment from aqueous media based on ~ltrafiltration'~ and on cloud-point extractions were performed on solutions containing only the uranium species. Practical interest in these. procedures arises from their possible applications in the management of nuclear wastes and during decommissioning treatments.20 Our attention has been focused on two main points: (i) the feasibility of the MEUF-based concentration of U"' from acid samples using chelating micelles obtained from the examined lipophilic ligands and common non-ionic surfactants and (ii) the possible separation of U"' from model mixtures containing also Sr" and Cd", exploiting their different tendencies to form complexes with those ligands.Radioactive strontium is a typical fission product which must be separated from uranium, and cadmium represents an example of metals coming from other sources, such as the acid attack of plant vessels. The succesful separation of metals which are usual components of stainless steel via MEUF has been described in a previous paper.2' Experimental Reagents and Materials The hydrophobic ligands examined have the structures shown in Fig. 1. Their synthesis and purification have been described in detail elsewhere.21.22 The surfactants Triton X- 100 [polyoxyethylene(9 5)-p- 1, 1,3,3-tetramethylbutyIphenol] and N-hexadecyl-N,N,N-tri- methylammonj um bromide (HTAB) of maximum purity grade, purchased from Merck (Darmstadt, Germany), were used as received to form the mixed micelles.Standard solutions of uranyl, cadmium and strontium were prepared starting from analytical-reagent grade UOz(N03)2.6H20 (Carlo Erba, Milan, Italy), Cd(N0&.4H20 (Merck) and Sr(N03)2 (Fluka, Buchs, Switzerland), respec- tively, in 0.5 mol I-' HN03. An ionic buffer solution composed of CsCl (0.1 g 1 - I ) from Merck was used in the determination of Sr. Arsenazo I11 { 2,2'-[ 1,8-dihydroxy-3,6-disulfo- 2,7-naphthylenebis(azo)]diphenylarsonic acid} from Merck was used for the spectrophotometric determination of uranyl. Sodium hydroxide solution (0.1-1 mol I - I ) , nitric acid (concentrated and 1 mol 1-1) and sodium nitrate, purchased from Merck, were of analytical-reagent grade.Doubly distilled, de-ionized water was used throughout. 8 PAN-C, H 0' Fig. 1 micellar aggregates. Structures of the hydrophobic ligands used to form the coinplexingI402 Aiialyst, October 1996, Vol. 121 Procedure Prepai-ation qf samples Owing to their very low solubility i n water, the ligands were first dissolved in aqueous surfactant solutions and then added to the metal-containing samples. The complex formation reaction takes place essentially at the micelle-water interface. performance. For example, for the mixture U022f-Sr2+, S is given by Ultt-ufiltt-ution I - M M . ~ Cylindrical cells (S-43-70) (Spectrum, Los Angeles, CA, USA). equipped with a magnetic stirring bar, were used for the ultrafiltration runs.Spectra/Por-C 1 0 disc hydrophilic cellulose membranes (molecular tnass cut-off I0 000 Da) were used in the experiments. Each ultrafiltration w a ~ usually performed on 30 ml of feed solution placed in the ultrafiltration cell (capacity 70 ml) and forced to pass through the ultrafiltration membrane by applying a constant pressure delivered by an inert gas, usually nitrogen. A constant volume (25 ml) of permeate solution was collected after about 1 h under 3 atm pressure. A simplified scheme of the ultrafiltration process is depicted in Fig. 2. The membranes were previously washed with water (about 30 ml) in order to eliminate the wetting agents incorporated in the filter (usually glycerin or polyethylene glycol). The rotation speed of the stirring bar, supported just above the membrane, was also constant in all the runs (aboul 300 rpm).The metal ions and the ligands were determined in the permeate. Several determinations were also performed on the retentate solution in order to check the mass balance. The ultrafiltration efficiency was calculated in each step by evaluating the rejection factor, R, defined as follows: (1) where C, is the analyte concentration in the permeate and Co is the initial concentration in the solution to be filtered. The efficiency of consecutive ultrafiltrations for each analyte was evaluated through the retentate/feed ratio, I-//, defined by the following expression: t‘,, = t71,,/??2() (2) where M, and nzo are the number of moles (or the mass) of the analyte in the retentate after ri consecutive ultrafiltrations and in the feed solution.respectively. Before starting a successive UF step, the volume of retentate (5 mi) was adjusted to 30 ml by adding aqueous surfactant solution at the c.m.c. The separation factor, S , defined as the ratio between the undesired component yield and the target analyte yield after each ultrafiltration step, was used to evaluate the separation R (96) = ( 1 - C,,/Co) X 100 Fig. 2 Schematic representation of an MEUF-based process with chelating micelles, where: I., retentate; p, permeate; and m, membrane. The solid circles represent the metal ion. Complete binding of the hydrophobic ligand (represented by a tailed square) is assumed. Each reported ultrafiltration parameter was the mean of the results obtained from three or four independent experiments.Ligand derernzinntion The rejection factors of the applied ligands were determined by measuring their concentrations in the permeate by HPLC. A liquid chromatograph composed of an L 6200 pump and an LC- 4200 UV/VIS detector (Merck-Hitachi, Tokyo, Japan) was used. Aliquots of 10-20 PI of permeate were injected into the column (LiChroCART CIS, Merck) and eluted with aceto- nitrile-water (40 + 60 v/ v ) at constant flow rate ( 1 ml min-I). The absorbances of the PAS-C,, ligands were monitored at 305 nm, whereas those of the PAN-C,, series were measured at 470 nm. A calibration curve was obtained under the same condi- tions. E \wluution of 1igand-mic.elle hinding Partition data concerning PAN-C, ligands in Triton X- 100 micellar solutions have been reported previousIy.2l The binding constants (Ks) of PAS-C2 and PAS-C4 compounds to Triton X- I00 aggregates were determined in this work from their retention data at different surfactant concentrations in the eluent, according to the micellar HPLC method.23 LiChrosorb RP-8 250-4 columns (10 pm) from Merck were used in these determinations, the detector wavelength being fixed at 305 nm.Uririiyl detcmiiimtiorz The analyses were performed on the permeate solution by using the Arsenazo 111 spectrophotometric method.24 Usually, 2 ml of chromogenic reagent ( 1 X 10 -3 mol 1-1 Arsenazo 111 dissolved in 1 mol 1-1 HCI) were added to 2 ml of permeate solution (previously acidified to pH 3 with HCl). The uranyl determina- tion in the retentate requires a preliminary dilution ( I + 9) with water.The absorbance of the corresponding uranyl complex was measured at 652 nm. A good linear calibration plot was obtained in the working range (corresponding to uranyl concentrations up to approximately 100 pmol 1-1). Strontium and cmhnium determination A. Model 1 100 B spectrometer purchased from Perkin-Elmer (Uberlingen, Germany) equipped with a standard air-acetylene burner was used for the AES determination of Sr. An Intensitron hollow-cathode lamp (Perkin-Elmer) was used for the AAS determination of Cd. All the samples were brought to the proper acidity (0.2% m/v HN03) prior to the spectrometric analysis. The spectrometer conditions are reported in Table 1. Table 1 Experimental conditions employed in the spectroscopic analysis of Sr and Cd Analyte Parametcr Sr (ABS) Cd (AAS) Waveleng th/nm 460.7 228.8 Slit width/nm 0.4 0.7 Aspiration tlow/ml min I 3.0 3.0 Air flow/l min-’ 8.0 8.0 Acetylene tlow/l min-1 2.5 2.0Analyst, October 1996, Vol.12 I 1403 Results and Discussion Determination of PAS-C, Binding Constants to Triton X-100 Micelles The evaluation of the ligand-micelle binding constants is very important as a direct relationship exists between this parameter and the MEUF efficiency. The retention volumes of the ligands investigated were determined as a function of the micellized surfactant concentration in the eluent phase, according to the Armstrong-Nome equation:23 where the chromatographic factor, Cf, is defined as: C f = V,/(V, - V,), V,, V, and V,,, are the volumes of the stationary phase, the eluted solute and the mobile phase, respectively, Psw is the partition coefficient of solutes between the stationary phase and the aqueous bulk and CD, the concentration of micellized surfabtant, is given by Cto, - c.m.c.From the slope and intercept of the linear plots of Cf IYI-SUS CD, the corresponding KB values were determined (see Fig. 3). The measured KR, together with the calculated value of the more hydrophobic ligand PAS-Cx, are listed in Table 2. For PAS-Cx, the binding constant was extrapolated from PAS-C2 and PAS-C4 data, taking into account that the contribution of each aliphatic carbon atom to the free energy of transfer of the molecule (from water to the micelles) is additive.2s These KB values compare fairly well with those determined previously in the presence of C l2EX [polyoxyethylene(8) dodecyl ether] micelles,l4 a surfactant which contains a number of polyoxyethylene units close to that in Triton X-1 00.Partition 0 0.005 0.0075 0.0100 0.0125 0.0150 0.0175 CD Fig. 3 the micellized surfactant concentration, CD, according to eyn.(4). Plots of the chromatographic retention factors (C,) as a function of ~~~ ~ ~ Table 2 Binding data for PAS-C,, and PAN-C,, ligands in non-ionic micellar solutions Ks/l mol-1 Compound Triton X- 100 C XEx+' PAS-C;?* 120 120 PAS-C4* 290 350 PAS-Cx",+ I 250 I800 PAN* 430 PAN-C,I 2 300 PAN-C,'.t 12 000 * This work. + Value calculated from the additive contribution of the aliphatic carbon atoms to the free energy of transfer of the whole f Data from ref.21. Data from ref. 14. data for PAN-C, ligands are also reported for comparison purposes. Micellar Ultrafiltrations With PAS-Cs The ligand PAS-C8 was chosen on the basis of its favourable binding with the host surfactant micelles, which ensures its negligible release in the ultrafiltrate. Quantitative rejection of the solutes is, in fact, observed when their KR values are higher than about 1000-1 500 1 mol- I . The ultrafiltration performances of these ligand-containing micelles were examined, with particular attention to the recovery yield of uranyl from the samples and to the separation factors from Sr" and Cd". Ultrafiltration blank experiments with Triton X- I00 (without any added ligand) were also performed. Rejection factors in the range 3-5% were determined for the investigated analytes, thus indicating a very small contribution to the retention arising from the membrane-surfactant system.The maximum ligand concentration in the MEUF runs was 7 X mol 1-1 (in 0.02 moll-' Triton X-100) and a constant ligand-to-metal ratio (10 : 1) was maintained in all the experi- ments. It must be emphasized that the ligand hydrophobicity, although favouring the chelate binding, limits the ligand excess in the system.The pH range examined was 3-6. Fig. 4 shows the rejection factor profiles obtained after the individual ultra- filtrations of each investigated analyte. The measured uranyl rejection at pH 5 (about 91%) indicates that a quantitative recovery of this metal in the collected retentate is possible by recycling and re-filtering the permeate, after addition of the proper amounts of surfactant and ligand to re-establish the initial conditions.At the same pH value, the corresponding rejection factors for Sr" and Cd" are still relevant (between 10 and 20%). However, the observed differences in %R should be in principle large enough to permit a separation based on multi- step ultrafiltrations, as demonstrated previously for other metal mix tures.21 Table 3 summarizes the ultrafiltration results obtained at pH 5 with ternary mixtures containing Uvl, Sr" and Cd". The ligand- to-metal ratio was kept constant at 9.33 : 1 (ligand concentra- tion, 7 X 10-4 mol 1-1; concentration of each metal ion, 2.5 X 10-5 mol 1-l). It must be noted that the rejection factor of uranyl is almost constant, whereas those of Sr" and Cd" increase significantly with n.Taking into account the initial composition of the (equimolar) feed solution, a beneficial effect is observed, but the relevant increase in the rejection factors of the undesired components during the process drastically reduces the separa- 100 4 80 60 8 cy: v 40 20 P O - l 3 4 5 PH 6 Fig. 4 Variation of %R as a function of pH in MEUF experiments with Triton X-100/ PAS-Cx aggregates. Surfactant concentration, 0.02 mol I-'; ligand concentration, 0.7 mmol 1- I and metal concentration, 0.07 mmol 1- 1 .1404 Ancrlyst, Octohei. 1996. Vol. 121 tion performance. In fact, only a smooth decrease of the separation factors with n is observed. A limited increase in R (96) in multi-step ultrafiltrations could be expected as a consequence of some surfactant (and ligand) accumulation at the membrane due to moderate adsorption effects.Moreover, an increase in retention as a consequence of the increased ligand-to-metal ratio in each consecutive step is also predictable, in particular for those metals which are only partially complexed. These unfavourable effects, which can be miniinized only in the presence of a large excess of ligand and/ or when the cotnplex fortnation is quantitative, are probably operating in the system under investigation. It is in fact difficult to increase the ligand concentration in the system owing to the low solubility of such hydrophobic compounds. Taking into account that most real wastes are acidic solutions, an ultrafiltration treatment performed at lower pH values should, in principle. be preferable.However, at lower pH values, the rejection o f the target analyte rapidly decreases (it is about 22% at pH 3). For this reason, an efficient separation is not feasible under these conditions and the use of other chelating agents was examined. Ultrafiltrations With PAN-C, Mixed Aggregates It is known that the ligand PAN has the tendency to form more stable complexes than salicylate with most metal ions.26 Previous studies on transition inetals2I indicated that PAN-Cd and PAN-Cx are quantitatively bound to Triton X-100 ag- gregates, whereas the unsubstituted ligand PAN is partitioned. In particular. the more soluble PAN-C4 ligand was chosen since its corresponding K R is high enough (2300 1 mol-l) to ensure negligible leakage into the permeate and, moreover, the complex formation is faster.Indiiiidircrl MEUF e.pv-inwnts Most individual ultrafiltrations were performed on aqueous solutions containing 1 X rnol I-' of each metal ion and 1 X 10- mol 1-1 of each ligand dissolved in 0.04 mol 1-1 Triton X-100. Ultrafiltrations performed at pH 3 showed that the uranyl accumulation in the retentate is in the range 92-97%, increasing ac expected with increasing ligand hydrophobicity. Fig. 5 shows the variation of %R with pH for Uvl, Cd" and Sr". It can be seen that an almost quantitative recovery of UVI (99%) from these aqueous acid samples is possible after a single ultrafiltration step, at pH 5 , using PAN-Cx. The ultrafiltration yield is only slightly lower when PAN-C4 is used.The separation of uranyl from Sr" and Cd" appears, in principle, to be feasible at pH 3 using PAN-Cd. All the investigated mixtures were therefore successively treated with chelating micelles containing the above ligand. Table 3 Ultrafiltration of ternary mixtures of UL', Cd" and Sr" using PAS- C8-Triton X- 100 mixed niicelles at pH 5 91 0.93 19 0.33 27 0.39 0.35 0.42 92 36 75 0.86 0.15 0.3 1 0.18 0.36 95 71 85 0.83 0.11 0.27 0.14 0.33 Stoichiometry of the UV1-PAN-C,, complexes The formation of highly hydrophobic uncharged chelates is preferred in MEUF-based separations with non-ionic host aggregates.I4 Since the stoichiometry of the metal chelates can vary on passing from homogeneous to micellar media, its determination under the working conditions is often necessary.In this work, the stoichiometry of the retained uranyl complexes was determined (at pH 3) using the Job method. It was found that zero-charge 2 : 1 ligand-metal complexes are formed with both PAN-C4 and PAN-C8 compounds, thus ensuring a strong binding of these species to the micelles. Ultrujiltrution of uranyl-containing binary mixtures The results obtained with binary mixtures are shown in Fig. 6, where the evolution of the corresponding separation factors with the number of successive ultrafiltration steps is reported. After four steps, only about 4% of the Sr" initially present in the feed solution is still present in the retentate phase, whereas the separation from Cd" is even better (approximately 1% of the initial amount).Ultrafiltration of ternury mixtures The rejection factors determined in these mixtures are different from those measured in the presence of a single component or in binary systems. In particular, a relevant increase in rejection 1 O - l 3 4 5 6 PH Fig. 5 Effect of pH on metal rejection using Triton X-100PAN- C4 aggregates. Surfactant concentration, 0.04 moll- 1; ligand concentration, 1 mmol I-'; and metal concentration, 0.1 mmol I-'. The uranyl rejections measured in the presence of PAN (curve 1) and PAN-C8 (curve 2) are also shown. v, 0.5 0.4 0.3 0.2 0.1 0 1 2 3 n A Fig. 6 Variation of the separation factor between uranyl and the undesired components in binary mixtures as a function of the number of UF steps. 1, SSr/urdnyl; 2, SC.d,urdnyl. Other conditions: as in Fig.5 .Analyst, October 1996, Vol. 121 1405 with n is observed for Cd" in ternary mixtures (see Table 4), whereas for uranyl and Sr" the rejection data for binary and ternary mixtures are comparable. From examination of the above results, the separation of uranyl from both Sr" and Cd", at pH 3, appears feasible. After three ultrafiltrations Ssr,uranyl and were decreased to approximately 0.1, whereas less favourable separation factors were obtained at this stage on working with PAS-Cg at pH 5 (in particular, the concentration of residual cadmium in the retentate was about three times higher). Conclusions The present experimental data indicate that preconcentration of uranyl ions at trace levels from dilute aqueous samples can be performed by MEUF using moderate amounts of cheap and safe surfactants (concentration range 10-25 g 1-1) and small amounts of hydrophobic ligands (0.2-0.3 g 1 - 1 ) readily obtained from common chelating compounds.Handling of larger amounts of more dangerous and usually toxic organic solvents is usually required when standard liquid-liquid extraction procedures are applied. Quantitative recovery of Uvl is possible, after two consecutive steps, using either PAS-Cg or PAN-C4 ligands at pH 5-6. The present performances are better than those obtained previously using the same method, but working with more complex chelating agents and in thc presence of auxiliary ligands. 17 The separation of uranyl from other metals is also possible by exploiting the same approach. In particular, the use of PAN-C, compounds allows a better separation of the target analyte from Sr" and from Cd", present at the same initial concentrations in aqueous acid samples.Financial support from CNR and MURST (Rome) is gratefully acknowledged. Table 4 Ultrafiltration of ternary mixtures of Uvl, Sr" and Cd" using PAN- C4-Triton X-100 mixed micelles at pH 3 Parameter Step 1 Step 2 Step 3 94 30 24 0.95 0.4 1 0.36 0.43 0.38 94 34 32 0.90 0.19 0.16 0.20 0.17 95 39 48 0.87 0.09 0.09 0.10 0.10 References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 I8 19 20 21 22 23 24 25 26 Hinze, W. L., in Solution Chemistry ofSuifac'tants, ed. Mittal, K. L., Plenum Press, New York, 1979, vol. 1, p. 79. Armstrong, D. W., Sep. Purg. Methods, 1985, 14, 213. Pelizzetti, E., and Pramauro, E., Anal. Chim.Acta, 1985, 169, 1. Pfuller, U., Mizellen, Vesikel, MikroemulJionen. Tensidassoziate und ihre Anwendung in Analytik und Biochemie, VCH, Berlin, 1986. Ordered Media in Chemical Separations, ed. Hinze, W. L., and Armstrong, D. W., ACS Symp. Ser., 342, American Chemical Society, Washington, DC, 1987. Surjactant-Based Separation Processes, ed. Scamehom, J. F., and Harwell, J. H., Marcel Dekker, New York, 1989. McIntire, G. L., Crit. Rev. Anal. Chem., 1990, 21. 257. Pramauro, E., and Bianco Prevot, A., Pure Appl. Chem , 1995, 67, 561. Scamehorn, J. F.. Ellington, R. T., Christian, S. D., Penney, B. W., Dunn, R. O., and Bhat, S. N., AZChE Symp. Ser., 1986, 82,48. Sasaki, K. J., Bumett, S. L., Christian, S. D., Tucker, E. E., and Scamehorn, J. F., Langmuir, 1989, 5, 363. Hafiane, A., Issid, I., and Lemordant, D., .I. Colloid Interface Sci.. 1991, 142, 1. Pramauro, E., and Pelizzetti, E., TrAC, Trends Anal. Chem. (Pers. Ed), 1988, 7, 260. Klepac, J., Simmons, D. L., Taylor, R. W., Scamehorn, J. F., and Christian, S. D., Sep. Sci. Technol., 1991, 26, 165. Pramauro, E.. Bianco A., Barni, E., Viscardi, G., and Hinze, W. L., Colloids Surf., 1992, 63, 291. Tondre, C., Son, S. G., Hebrant, M., Scrimin, P., and 'recilia, P., Langmuir, 1993, 9, 950. Dharmawardana, U. R., Christian, S. D., Taylor, R. W., and Scamehorn, J. F., Langmuir, 1992, 8, 414. Pramauro, E., Bianco Prevot, A., Pelizzetti, E., Marchelli, R., Dossena, A., and Biancardi, A., Anal. Chim. Acta, 1992, 264, 303. FemBndez Laespada, M. E., PCrez Pavon, J. L., and Moreno Cordero, B., Analyst, 1993, 118, 209. The Treatment and Handling of Radioactive Wastes, ed. Blasewitz, A. G., Davis, J. M., and Smith, M. R., Springer, New York, 1983. International Atomic Energy Agency, Technical Reports Series, No. 286, IAEA, Vienna, 1988. Pramauro, E., Bianco Prevot, A., Zelano, V., Hinze, W. L., Viscardi, G., and Savarino, P., Talanta, 1994, 41, 1261. Savarino, P., Viscardi, G., Barni, E., Pelizzetti, E., Pramauro, E., and Minero, C., Ann. Chim. (Rome), 1987, 77, 285. Armstrong, D. W., and Nome, F., Anal. Chem., 1981, 53, 1662. Onishi, H., and Toita, Y., Bunseki Kagaku, 1965, 14, 1141. Bunton, C. A., and Sepulveda, L., J . Phys. Chem., 1979, 83, 680. Sillen, L. G., and Martell, A. E., Stability Constants of Metal-Ion Complexes, Chemical Society, London, 1971, suppl. 1, part 11, pp. 482 and 738. Paper 6104256B Received June 18, 1996 Accepted July 18, 1996
ISSN:0003-2654
DOI:10.1039/AN9962101401
出版商:RSC
年代:1996
数据来源: RSC
|
18. |
Lead isotopic analyses of NIST Standard Reference Materials using multiple collector inductively coupled plasma mass spectrometry coupled with a modified external correction method for mass discrimination effect |
|
Analyst,
Volume 121,
Issue 10,
1996,
Page 1407-1411
Takafumi Hirata,
Preview
|
PDF (828KB)
|
|
摘要:
Analyst, October 1996, Vol. 121 (1407-141 I ) 1407 Lead Isotopic Analyses of NIST Standard Reference Materials Using Multiple Collector Inductively Coupled Plasma Mass Spectrometry Coupled With a Modified External Correction Method for Mass Discrimination Effect Takafumi Hirata Laboi-atoi;~~foi* Planetar-y Scieiices. Tokyo Institute of Teclinology, 0-Okayama 2-12-1, Meguro, Tohyo 152, ,Ja~3aii A correction method for the mass discrimination effect was developed for isotopic analyses using multiple collector inductively coupled plasma mass spectrometry (MC-ICP-MS). For Pb isotopic analysis using MC-ICP-MS, the correction factor for the mass discrimination effect on Pb is based on the addition of TI to the sample solution and measurement of TI isotopic ratios; the correction factor obtained using TI is directly applied to the Pb isotopes (conventional external correction).However, the series of measurements of discrimination factors for several elements, including Rb, Sr, Ru, Nd, Hf, Re, Os, TI and Pb (mass range 80-210 u), clearly reveal that the mass discrimination factors observed using MC-ICP-MS were a linear function of mass, suggesting that the correction factors observed using T1 isotopes were not exactly identical with those for Pb isotopes. Therefore, the correction factors obtained with TI isotopes should be corrected for mass, and then applied to the Pb isotopes. The resultant Pb isotopic ratios for NIST Standard Reference Materials show excellent agreement (within 0.03% for 2ohPb/2*4Pb and 20 ppm for 207Pb/206Pb) with the data obtained by the thermal ionization mass spectrometry.The correction method presented clearly demonstrates the wide versatility of the external correction technique for the precise isotopic analysis using MC-ICP-MS. The possible cause of the ‘exceptionally large’ mass discrimination effect observed for Ru and 0 s is discussed. Keywords: Indiic*tively coupled plasniu triass spectrometry; multiple cvllector; isotopic ratio measurement; lead isotopes; niuss dis cr ini iTi u t ioi I Introduction Mass spectrometry using an inductively coupled plasma as an ion source (ICP-MS) is now accepted as a versatile analytical technique for elemental and isotopic analyses. ICP-MS had been most frequently utilized as a tool for sensitive elemental analysis, since the precision of the isotope abundance ratio achieved by the present quadrupole analyser-based ICP-MS (ICP-QMS) is not always sufficient for geological dating and determining nuclear properties.Z4 It has been suggested that the precision and accuracy of the isotopic ratio measurement achieved by ICP-QMS are severely restricted by the lack of detector linearity, background contributions and/or the mass discrimination effect.4 To date, thermal ionization mass spec- trometry (TlMS) has been accepted as a ‘benchmark’ technique for isotopic analyses.However, TIMS is time consuming and often requires higher levels of chemistry skills. Furthermore, the accuracy of TIMS analysis is limited by a time-dependent sample fractionation effect. Thus. although the high precision and accuracy of isotope ratio measurements can be achieved by TIMS, its utility has been restricted. Recently, the measurement of isotope ratios by multiple collector inductively coupled plasma mass spectrometry (MC- ICP-MS) has been described.”l2 Multiple collectors allow each isotope to be monitored simultaneously, thus removing signal instability as a limitation on analytical precision.Furthermore, when using Faraday cups as ion detectors, the analogue detection mode can offer a better dynamic range and precision than the high-gain pulse counting mode. Isotope abundances for several elements including Nd,8 M o . ~ Sn,9 Te,9 Hf,X%lo W,’,‘ 1,12 Pb6-8 and U h , l j by MC-ICP-MS have been reported. Typical RSDs of 0.005-0.0 1% (20) could be achieved by this technique and the levels of analytical precision and accuracy achieved by MC-ICP-MS were comparable to those exhibited by TIMS.However, in the MC-ICP-MS measurement, correction of the mass discrimination effect is strongly required for accurate isotopic analysis, since the correction factors for mass discrim- ination effects observed in MC-ICP-MS are typically larger than those observed in TIMS. Several processes are considered to contribute to the mass discrimination effect, including the space-charge effect in the plasma or vacuum interface re- gions. J 4 The space-charge effect results in the preferential transmission of the heavier ions. A typical mass discrimination effect for Pb observed in MC-ICP-MS is 1 % u-l, but is independent of time. This is in contrast to TIMS, which exhibit a time-dependent isotope fractionation effect.An external correction method for the mass discrimination effect is very important for improving the accuracy of the measurement, because the external correction method can remove the need for the analysis of calibrated isotopic standards for the corrections.1S-l6 Taylor et aI.l3 examined the correction method for the mass discrimination effect by means of a set of synthetic uranium isotope mixtures, and they concluded that a power law and exponential function result in the best correc- tion. The aim of this work was to investigate the mass discrimina- tion effect observed for MC-1CP-MS by measuring various elements. including Rb, Sr, Ru, Nd, Hf, Re, Os, T1 and Pb, and to develop a correction method for the mass discrimination effect for Pb.Superior levels of precision and accuracy of the lead isotopic data could be achieved with the present correction method.1408 Analyst, October 1556, Vol. I21 Experimental Instrumentation The MC-ICP-MS instrument used was a VG Elemental (Winsford, Cheshire, UK) Elemental Analysis Plasma-54. Details of the instrument and the parameters are given in Table 1. Sample solutions were allowed to aspirate freely into the plasma via a glass expansion nebulizer. Operating conditions such as torch position, argon gas flow rate, ion energy, extraction voltage and quadrupole lens settings were tuned to maximize the transmission of analyte signals. The sampling and skimmer cones were both held at 5900 V, which provides the acceleration potential for the analyte ions as they enter the forward-arranged double-focused mass spectrometer.The mass spectrometer achieved a 540 mm dispersion and incorporated nine adjustable Faraday collectors. The position of each Faraday collector was adjusted to suit the isotopic composition of each analyte as appropriate. Faraday cups and electrometers with lo1' 52 feedback resistors were used to measure each isotopes. The gain of the preamplifier associated with each Faraday collector was calibrated with respect to the axial (central) collector. No correction for collection efficiency was made. ~ ~~ ~ ~ ~ ~ ~ Table 1 ICP-MS instrumentation and parameters used in this study MC-ICP-MS instrument- ICP ion soiirce- VG Elemental Plasma-54 ICP Power Argon gas tlow rates- Cooling Auxiliary Nebulizer Sample injection Solution uptake rate Spray chamber temp.27.12 MHz 1.35 kW forward, < 20 W reflection 13 I min-' 0.5-1 1 min-1 0.88 1 min-1 Glass expansion nebulizer 0.6 ml min-1 (not pumped) 4 "C Vacuum inter$ace- Sampling cone Skimmer cone Pressure Quadrupole lens housing Ni, 1 mm orifice Ni, 0.5 mm orifice 4 x 10-2 Pa 2 x 10-6 Pa ESA housing Analyser tubing 2 x 10-7 pa Lens setting- Ion energy Extraction 5900 V 3800-4100 V Mass spectrometer- Resolution 400 (5% height) Analysis mode Static Ion detection Analogue by Fdraday Typical transmission Integration time 5 S 1.2-1.5 v (pg g-I)-' Signal acquisition- Scan settled time 5 s Standardization and correction- Mass discrimination Calc. Normalization value- R b/87 Rb X6Sr/X8Sr 99R~/102R~ 146Nd/144Nd 179Hf/177Hf lX5RelIX7Re lXSOs/l92Os 205T1/203T1 207Pb/206Pb Power Law 2.5927 (ref.18) 0.1 194 (NIST value, ref. lo) 0.4042 (ref. 19) 0.7219 (ref. 20) 0.7325 (ref. 21) 0.5974 (ref. 22) 0.3244 (ref. 23) 2.3871 (ref. 24) 0.914 64 (NIST value, ref. 18) Data Acquisition All measurements were carried out in the static mode. Each sample was analysed for 300 s and the analysis period consisted of 20 measurements, each of 5 s duration. The typical sample up-take rate was 0.6 ml min--*; a peristaltic pump was not used. Chemicals NIST SRMs 981 and 982 were selected for Pb isotopic measurements. Each SRM was dissolved in dilute HNO7 (approximately 0.5 moll-') at 80 "C for over 12 h in a Savilex PTFE bomb. The resultant solution was then diluted to 1000 pg 8-1 using Milli-Q SP water and was used as a stock standard solution.In the Sr isotopic study, NIST SRM 987 Sr was used for the measurement. Strontium reference material (SrC03) was dissolved in concentrated HN03 for 1 h and then diluted to 1000 pg g-1 as a stock standard solution. Nitric acid used for the dissolution was commercially available superclean acid (TAMAPURE AA-100 HN03), which was purified by sub- boiling distillation. Each stock standard solution was diluted to a concentration of 2 pg g-1 for the actual measurement. The Rb analytical solution used in this study was diluted from a Cica- Merck atomic absorption standard (1000 pg g-'). Analytical solutions for Ru, Nd, Hf, Re, 0 s and TI were diluted from Johnson Matthey Specpure ICP/DCP standard solution (1000 pg g-1).Results and Discussion Lead Isotopic Analysis As mentioned in the previous section, the mass discrimination effect observed in ICP-MS is mainly caused by the space- charge effect within plasma and vacuum interface regions. Taylor et 01.13 examined the correction method for the mass discrimination effect by means of a set of synthetic uranium isotope mixtures, and concluded that a power law and exponential function result in the best correction. In this study, the power law method was applied for the calculations. The following equation has been shown to predict the mass discrimination bias: R,,,,, = ~ r n e a s (1 + Wrn (1) where R,,, = corrected isotopic ratio, R,,,, = measured ratio, C = bias factor and 6m = mass difference. Longerich et a1.15 and Ketterer et a1.16 have demonstrated that thallium can be used as an isotopic calibration standard for the mass discrimination correction for Pb (external correction).Comparison of the observed T1 isotopic ratio with the true ratio allows the calculation of the discrimination factor (C) and hence a simultaneous correction for mass discrimination exhibited by the lead isotopes. Walder and co-workersS--ll applied this correction technique to MC-ICP-MS and achieved RSDs of 0.005-0.01 %. Because the external correction method can remove the need for the analysis of calibrated isotopic standard for the corrections, an external correction method for the mass discrimination effect is very important for improving the accuracy of the measurements and sample throughput. Table 2 summarizes the isotopic data (20XPb/204Pb, 206Pb/2"4Pb, ZOXPb/ 206Pb and 207Pb/206Pb) for NlST SRM 981.Lead isotopic data reported by Walder et a1.8 are also given. The magnitude of the mass bias (C) was typically between 0.0090 and 0.0110 throughout this measurement. As can be seen from these results, after correction for the mass discrimination effect using the TI external power law, the mean isotopic data obtained here show good agreement with the data obtained by MC-ICP-MS reported by Walder et a1.8 However, scrutiny of the data reveals that Pb isotopic data obtained by MC-ICP-MS appear to deviateAnalyst, October 1996, Vol. 121 I409 from the NTST certified value.17 It should be noted that not only the present data but also the Pb isotopic data reported by Walder et al.* were found to be systematically lower than the certified value.For the 20XPb/204Pb ratio, the present MC-ICP-MS value (36.642 k 0.020) agrees with that reported by Walder et a1.X (36.69 1 k 0.01 S ) , whereas the Pb isotopic ratio obtained by MC- ICP-MS clearly disagreed with the value certified by NIST (36.721 f 0.036). This disagreement has been noted and discussed by Walder et al.8 The most plausible explanation for the systematic difference between the MC-ICP-MS result and the NIST certified value is that mass discrimination effect is not effectively corrected by TI. In order to test this, the discrimina- tion factors calculated using TI and Pb isotopes were directly compared. The resultant discrimination factors calculated from 203T1/*()5T1 ratios were plotted against the discrimination factors calculated from the Pb isotopic data assuming that the 207/206Pb ratio for NIST 981 is 0.914 64.As can be seen in Fig. 1, although the mass discrimination Factors calculated using 203T1/ 205Tl and using *"7Pb/2"Pb ratios were closely correlated, it is evident that almost all the data points were plotted system- atically above the 1 : 1 correlation line. This can be understood as reflecting a systematic difference in discrimination factors between Pb and T1. Following the comparison of the mass discrimination factors for T1 and Pb, discrimination factors for other elements covering a wider mass range (80-210 u) were measured in order to test the possible relationship between mass discrimination effect and mass.Discrimination factors for Rb, Sr, Ru, Nd, Hf, Re, Os, T1 and Pb were measured and are summarized in Table 3. The power law as defined in eqn. ( 1 ) was used for the calculation. The resultant discrimination factors calculated for these elements were plotted against the intermediate mass of the monitored isotopes of each elements (Rb, 86; Sr, 87; Ru, 100.5; Nd, 145; Hf, 178; Re, 186; Os, 190; T1, 204; and Pb 206.5) as shown in Fig. 2. As can be seen, seven points for Rb, Sr, Nd, Hf, Re, T1 and Pb fall close to a straight line, showing a clear dependence of discrimination factors on mass (discrimination factors for Ru Table 2 lead isotopic ratios for NIST SRM 981 corrected for the mass discrimination effect by the conventional thallium external method.Errors are 20. Normalization 2usT1/2()3TI = 2.3871 (ref. 24) Run No. 20Xpb/204pb 206pb/204pb 2OXPb/206Pb 207Pb/206Pb 1 36.625 16.927 2.1637 0.9 1433 ( n = 100) f0.009 f0.004 f0.0002 f0.00003 2 36.636 16.925 2.1646 0.9143 1 (n = 100) f0.007 f0.003 fO.OOO1 +0.00003 3 36.663 16.937 2.1648 0.91435 ( n = 100) +0.009 f0.004 f0.000 1 f0.00003 4 36.645 16.929 2.1646 0.91432 ( n = 100) f0.007 f0.003 fO.0001 f0.00003 5 36.639 16.925 2.1649 0.91431 ( n = 100) f0.004 kO.001 fO.OOO1 f0.00002 6 36.640 16.927 2.1649 0.91432 (n = 100) k0.004 f0.002 fO.OOO 1 +0.00002 7 36.644 16.925 2. I649 0.9 1432 (n = 100) f0.006 f0.002 f0.000 1 f0.00002 8 36.638 16.922 2.1651 0.9 1434 (I? = 100) k0.004 f0.002 fO.0001 f0.00002 Average 36.642 16.927 1 2.16469 0.914325 c (20) 0.020 0.0090 0.00086 0.000028 RSD (%) 0.055 0.053 0.040 0.003 1 Walder et a1.8 36.691 16.937 2.1662 0.9141 1 50.0 1s f0.008 f0.0002 f0.0002 1 NIST value 36.72 1 16.937 2.1681 0.91464 f0.036 fO.011 f0.0008 f0.00033 and 0 s appear to deviate from this line; the possible cause of this deviation will be discussed later).The slope and intercept can be calculated from the straight line drawn by the least- squares method with respect to the points for Rb, Sr, Nd, Hf, Re, T1 and Pb, as shown in Fig. 2. The resultant relationship between bias factors and mass is C = -1.329 X rn + 1.0356 (2) where C and rn are the discrimination factor and intermediate mass of the monitored isotopes, respectively. The relationship between the mass discrimination factors and the mass reveals that the mass discrimination effects for lighter elements are more serious than those for heavier elements.Eqn. (2) also demonstrates clearly that the correction factors for the mass 3 U P ! 1.006 k,,< I . . . . I . . . . I . . . . 1.006 1.008 1.010 1.012 1.014 Bias Factor Calculated by Lead Isotopes (Power Law Correction using "$b /206Pb) Q ' u. fn .- Q m Fig. 1 Mass discrimination factors calculated using T1 and Pb isotopic ratios. Discrimination factors were calculated by the power law function. Normalized values for- TI and Pb are 20sT1/203Tl = 2.3871 (ref. 24) and 207Pb/206Pb = 0.914 64 (NIST value), respectively. Error bars represent the standard deviation (20) on 60 repeated ratio measurements. Although discrimination factors calculated by using T1 are closely correlated with those calculated using Pb, it is evident that all the points lie above the 1 : I correlation line.Table 3 Summary of the discrimination factors for Rb, Sr-, Ru, Nd, Hf, Re, Os, T1 and Pb. Errors are 20 Element Rubidium Strontium Ruthenium Neodymium Hafnium Rhenium Osmium Thallium Lead Monitored ratio* 8sRb/X7Rb 86Sr/XXSSr 1 79Hf/l77Hf Bias factor: 1.0243 f0.0002 I .0239 fO.OO 12 1.0277 f0.0002 1.0161 f0.0002 1.0116 C0.0004 1.01 17 f0.0001 1.0124 f0.0002 1.0083 k0.0032 1.0080 f0.0032 * Normalization: Rb (ref. 18), Sr (NIST value, ref. lo), Ru (ref. 19), Nd (ref. 20), Hf (ref. 21), Re (ref. 22), 0 s (ref. 23), Th (ref. 24), Pb (NIST value, ref. 17). + Calculated by power law (average of five sets of repeated analyses).1410 Analyst. October 1996, Vol.121 discrimination effect for Pb can be different from that for TI, because the intermediate mass for 207Pb/"06Pb ( r n = 206.5) is heavier than that for 2osT1/20-3T1 (m = 204). Therefore, in order to obtain further accurate Pb isotopic data using the external correction method, correction Factors calculated using T1 isotopes should be corrected for mass dependence prior to the collection of mass discrimination data on Pb. In this qtudy, we calculated the difference in mass discrimination factors between T1 and Pb using a relationship defined by eqn. (2): Cpb/CI I = [ - 1.329 X 10-4 X (206.5) + 1.0356]/ [-1.329 X 10 X (204) + 1.03561 (3) where CPb and CTI are the discrimination factors for Pb and TI, respectively. Crl can be obtained by measuring T1 isotopic ratios and CPb can be calculated from the CT, using eqn.( 3 ) , and then Cpb is used as a correction factor for Pb. The resultant Pb isotopic ratio data (2i)xPb/204Pb, 206Pb/2"4Pb, 20gPb/2°6Pb and 2o7Pb/'OhPb) observed for NIST SRM 981 are summarized in Table 4 and the repeated determination of the z07Pb/206Pb ratio is shown in Fig. 3. The ?o7Pb/20hPb isotopic ratio calculated by the conventional external T1 correction technique is also shown in Fig. 3. The relative deviations of the Pb data corrected by the conventional T1 external method and mass-corrected T1 external method developed here are about -0.04% and -0.002%, respectively, showing a substantial improvement in the accu- racy of the measurement. Lead isotopic data observed for NIST SRM 982 (equal atom, 2o*Pb/206Pb = 1 ) are summarized in Table 5 .All the Pb isotopic data for N E T SRM 981 and 982 show excellent agreement with the certified values. Fractionations of Ruthenium and Osmium As can be seen in Fig. 2, the discrimination factors for Ru and 0 s appear to deviate from the line defined by the six points for Rb, Sr, Nd, Hf, Re, T1 and Pb. The most plausible explanation for this is that the normalization values used for Ru and 0 s are incorrect. For Ru, two different normalization values have been independently proposed for the isotopic study of Ru. The calculated discrimination factors for Ru based on each normal- ization value are 1.0277 (C)6Ru/i02Ru = 0.4042'9) and 1.0285 (96Ru/101Ru = 0.324 8Sl25)), and both values are significantly 1 t l o 75 100 125 150 175 200 225 Mass (u) Fig.2 Dependence of the inass discrimination Factors on mass. Mass discrimination factors for I<b, Sr, Ru. Nd, Hf, Re, 0 s . TI and Pb were calculated and plotted against intermediate mass of measured isotopes of each clement: Rh, 86; Sr, 87; Ru, 100.5; Nd, 145; Hf, 178: Re, 186; 0 s . 190; TI, 204; Pb 206.5. The power was used for the calculation. The points for Sr. Nd, Hf. Re, TI and Pb fall close to a straight line, showing a clear dependence of discrimination factors on mass. The discrimination factors for Ru and 0 s appear to deviate from this line (the possible cause of the deviations is discussed i n the text). Table 4 Lead isotopic ratios for NET SRM 981 corrected for the mass discri in inati on effect by the rnass-corrected thall i um external method.Eirors are 20. Normalization: 205TI/2('3TI = 2.387 1 (ref. 24) Run No. 208pb/204pb 206pb/204pb 20Xpb/20bpb 207Pb/206pb 1 36.664 16.93 1 2.1654 0.9 1463 2 36.675 16.929 2.1663 0.9 146 1 ( / I = 100) f0.007 f 0 . 00 3 fO.000 1 f0.00003 3 36.702 16.94 I 2.1665 0.9 1465 ( M = 100) f0.009 f0.004 fO.OOO 1 f0.00003 4 36.684 16.933 2. I663 0.9 1462 ( / I = 100) k0.007 f0.003 fO.OOO 1 rt0.00003 5 36.678 16.929 2.1666 0.9 1462 (n = 100) _+0.004 k0.00 1 ~0.000 1 +0.00002 6 36.679 16.93 1 2.1666 0.91459 ( n = 100) f0.004 f 0 .002 fO.OOO I f0.00002 7 36.683 16.929 2.1666 0.91462 ( n = 100) f0.006 f0.002 fO.000 1 +0.00002 X 36.677 16.926 2.166X 0.91464 ( n = 100) f0.004 f0.002 f0.000 1 f0.00002 (t? = 100) f0.009 f0.004 f0.0002 f0.000C~3 Average 36.680 16.93 I 1 2.16636 0.9 14623 (2u) 0.02 1 0.0090 0.00082 0.00003 7 RSD (%) 0.058 0.053 0.018 0.0040 NIS value 36.72 1 16.937 2.1681 0.9 1464 f0.036 fO.01 I f0.0008 f 0 .0003 3 a 0.9147 an 0.9146 0.9145 gn 0.9144 0.91 43 8 9 (v Fig. 3 Comparison of observed 2o7Pb/2O6Pb ratios for NIST SRM 981 using conventional TI external correction and mass-corrected TI external correction techniques. The relative deviation of the Pb isotopic data corrected by the conventional TI external method and mass-corrected TI external method are about -0.04%' and -0.002%, respectively, showing a substantial improvement in the accuracy of the measurement by using the present external correction technique. Table 5 Lead isotopic ratios for NIST SRM 982 corrected for the mass discrimination effect by the mass-corrected thallium external method.Errors are 20. Normalization: 2osT1/203T1 = 2.3871 (ref. 24) Run No. 208pb/204Ppb 206Pb/204Pb 2(lXPb/206Pb 207Pb/206Pb 1 36.723 36.726 0.9990 0.46704 ( n = 60) rt0.006 f0.006 f0.00004 fO.OOOO I 2 36.696 36.699 0.99989 0.46704 ( n = 60) f0.006 f0.005 f0.0000S fO.OOOO1 3 36.7 10 36.720 0.99979 0.46699 ( n = 60) f0.006 f0.005 f0.00005 f0.00001 4 36.693 36.693 0.99992 0.46707 ( n = 60) f0.006 20.005 f0.00005 fO.OOOO I 5 36.693 36.696 0.99984 0.46704 (n = 60) f0.006 +0.005 k0.00005 _+0.00002 6 36.7 I9 36.725 0.99985 0.4670 1 ( n = 60) k0.006 f0.005 fO.00005 k0.00001 7 36.715 36.722 0.99984 0.46702 (tz = 60) f0.007 f0.005 f0.00008 rt0.00002 8 3 6.704 36.71 1 0.99986 0.4670 I ( n = 60) f0.004 f0.003 k0.00007 f0.00002 Average 36.707 36.7 12 0.99986 0.46703 (20) 0.024 0.027 0.00008 0.00005 RSD (56) 0.066 0.074 0.0080 0.0 I05 NIST value 36.745 36.739 1.00016 0.46707 f0.039 f0.036 f0.00036 f0.00020Analyst, October 1996, Vol.121 141 1 above the line defined by Rb, Sr, Nd, Hf, Re, T1 and Pb. For Os, the normalization value of IX~OS/~~?OS = 0.32442-1 applied in this study has been widely used as a ‘golden number’ for the 0 s isotopic study (e.g., refs. 23-28), and therefore it is very difficult to test the accuracy of the normalization value. Masuda et al.29 determined 0 s isotopic abundances by means of a comparison of the isotopic sensitivity for each 0 s isotope and reported a 18*Os/1920s value of 0.3237 k 0.0005 (Os04 purchased from Strem Chem. Inc.), showing good agreement with the ratio reported by Nier.23 It should be noted that there still remains a small gap between the discrimination factors for 0 s calculated by the normalization value reported by Masuda et al.29 and the line defined by other elements.Furthermore, for the TIMS analyses, since the typical discrimination factors observed for heavy isotopes ( > 100 u) should be smaller that 0.1% u-1, the possible uncertainty for the normalized value for 0 s could be considered to be less than 0.1-0.2%, and therefore the deviation of the 0 s discrimination factors from the line defined by Sr, Nd, HF, Re, TI and Pb might not be caused by the uncertainty of the normalization value. Hence, we are led to an inference that the mass discrimination effects for Ru and 0 s isotopes are significantly larger than those for other elements.The remarkably high mass discrimination effect for Ru and 0 s might be explained by their unique physico-chemical features. Ru and 0 s easily form tetraoxides (Ru04 and Os04), both of which are highly volatile. In fact, the chemical forms of Ru and 0 s used for the series of measurements were Ru04 and Os04. Because Ru04 and Os04 begin to sublime even below the boiling-point of water (solution mist), in the case of these elements isotopic fractional evaporation from the sample mist can take place in the plasma, suggesting that the volatile oxides (Ru04 and Os04) enriched in the lighter isotopes should be evaporated and exposed to Ar+ at an earlier stage of atomization or ionization process compared with the oxides enriched in heavier isotopes.This results in preferential ionization of lighter isotopes in the plasma. As a result, lighter isotopes could have more seriously suffered from space-charge effects in the plasma, and hence the mass fractionation effects observed for Ru and 0 s were exceptionally larger than those for other elements examined in this study. However, this inference is open to the possibility that the normalization values for Ru and 0 s might not be accurate enough. In order to draw firm conclusions on this matter, measurement of discrimination factors using a double spike technique is strongly desirable. The main conclusion of this study is that the correction biases for the mass discrimination effect observed for MC-ICP-MS were well correlated with mass, and when the bias factor for TI was corrected by mass, the external correction technique could provide much more accurate Pb isotopic data compared than data calculated by the conventional TI external correction method.The resultant Pb isotopic data (207Pb/206Pb) observed for two NIST SRMs show excellent agreement with the certified values. The ability to correct for the mass discrimina- tion effect allows precision and accuracy of isotope ratios comparable to those given by TIMS. We are grateful to T. Shimamura (Kitasato University) and A. J. Walder (VG Elemental) for technical support and advice. Helpful comments from S. Scott (VG Elemental) are gratefully acknowledged. This work was supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science, Sports and Culture, Japan, and by a Kurata Research Grant, Japan.References 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 Falkner, K. K., Klinkhammer, G. P., Ungerer, C. A., and Christie, D. M., Annu. Rev. Earth Planet. Sci., 1995, 23. 409. Hirata, T., Akagi, T., Shimizu, H., and Masuda. A., Anal. Ch~m., 1989, 61, 2263. Hirata, T., and Nesbitt, R. W., Geochim. Cosmochim. Acta, 1995, 59, 249 I . Russ, G. P. I., and Bazan, J. M., Spectrochim. Actu, Part B , 1987,42, 49. Walder, A. J., and Freedman, P. A., J . Anal. At. Spectrom.. 1992, 7, 571. Walder, A. J., Koller, D., Reed, N. M., Hutton, R. C., and Freedman, P. A., J . And. At. Spectrom., 1993, 8, 1037. Walder, A. J., Abell. I. D.. Platzner, I., and Freedman, P.A., Spectrochim. Acta, Part B , 1993, 48, 397. Walder. A. J., Platzner, I., and Freedman, P. A., .I. Anal. At. Spectrom., 1993, 8: 19. Lee, D. C., and Halliday, A. N., Int. J . Mass Specmjm. I o n Pnxesses, 1995, 1461147, 35. Thirlwall, M. F., and Walder, Chem. Geol. (Isot. Geosci. Sect), 1995, 122, 241. Halliday, A. N., Lee, D. C., Christensen, J. N.. Walder. A. J., Freedman, P. A., Jones, C. E., Hall, C. M., Yi, W., and Teagle, D., Itzt. J . Mass Spectrom. Ion Processes, 1995, 1461147, 21. Lee, D. C., and Halliday, A. N., Nature (London), 1995, 378, 77 I . Taylor. P. D. P., De Bievre, P., Walder, A. J . , and Entwistle, A., J. Anal. Atm. Spectrom., 1995, 10, 395. Gilson, G. R., Douglas, D. J., Fulford, J. E., Halligan, K. W., and Tanner, S. D., Anal. Chem., 1988, 60, 1472. Longerich, H. P., Fryer, B. J., and Strong, D. F., Specmx,him. Actci, Part B , 1987, 42, 39. Ketterer, M. E., Peters, M. J., and Tisdale, P. J., J . Anal. At. Spectrom., 1991. 6, 439. Catanzaro, E. J., Murphy, T. J . , Shields, W. R., and Garner, E. L., J . Res. Natl. Bur. Stand., Sect. A. 1968, 72, 261. Catanzaro, E. J., Murphy, T. J., Garner, E. L., and Shields, W. R., J . Res. Nutl. Bur. Stand., Sect. A , 1969, 73, 5 1 I . Huthceon, I. D., Armstrong, J. T., and Wasserburg. G. J., Geochim. Cosmochirn. Acts, 1987, 51, 3175. Wasserburg, G. J., Jacobsen, S. B., DePaolo, D. J., McCulloch, M. T., and Wen, T., Geochim. Cosmochim. Acts, 198 1, 45, 23 1 1. Patchett, P. J., Geochim. Co.smochim. Acta, 1983, 47, 8 1. Gramlich, J. W., Murphy, T. J., Garner, E. L., and Shields, W. R., J . Res. Natl. Bur. Stand., Sect. A, 1973, 77, 691. Nier, A. O., Phys. Rev., 1937, 52, 88.5. Dunstan, L. P., Gramlich, J. W., Barnes, I. L., and Purdy, W. C., J . Res. Natl. Bur. Stand., 1980, 85, I . Poths, H., Schmitt-Strecker, S., and Begemann, F., Grochim. Cosmochim. Acta, 1987, 51, 1143. Creaser, R. A., Papan iou, D. A., and Wasserburg, G. J., Geochim. Cosmochim. Acta, 1991, 55, 397. Luck, J. M., and Allegre, C. J., Nature (London), 1983, 302, 130. Walker, R. J., and Fassett, J. D., Anal. Chem., 1986, 58, 2923. Masuda, A., Hirata, T., and Shimizu, H., Grochem. J., 1986, 20, 233. Paper 6102828D Received April 23, 1996 Accepted July 15, 1996
ISSN:0003-2654
DOI:10.1039/AN9962101407
出版商:RSC
年代:1996
数据来源: RSC
|
19. |
Determination of lead in soil samples by in-valve solid-phase extraction–flow injection flame atomic absorption spectrometry |
|
Analyst,
Volume 121,
Issue 10,
1996,
Page 1413-1417
Ponlayuth Sooksamiti,
Preview
|
PDF (758KB)
|
|
摘要:
Aizalyst, October- 1996, Vol. I21 (14I3-1417) 1413 Determination of Lead in Soil Samples by In-valve Solid-phase Extraction-Flow Injection Flame Atomic Absorption Spectrometry Ponlayuth Sooksamiti", Horst Geckeisb and Kate Grudpanc," j7 Institut fiii. Nukleai- Entsoi-Kungsteclznik, Forschungszentr-urn Karlsrulze. Postfach 3640, 0-76021 Karlsi-ulze, Germany ( Dqar-tnient (g Cliemisti-l,, Faculty of Science, Clziang Mai University, Cliiang Mai 502000, Thui1am-l Miriel-al Resour-ces Region 3 (Chiung Mai), Chiang Mai 50200, Tlzailund Solid-phase extraction (SPE) using an immobilized crown ether as an extractant was applied to the flow injection FAAS determination of Pb" with in-valve minicolumn preconcentration and separation. Lead(I1) was first loaded on to a column filled with the crown ether resin from a nitric acid solution (0.8-2.0 mol dm-3).Among the eluents studied (oxalic acid, ammonium oxalate, citric acid, sodium citrate and tartaric acid), 0.05 mol dm-3 ammonium oxalate was found to be most suitable. Calibration was made either by variation of the preconcentration time (using a single standard) or by using standard solutions of different concentrations. The detection limit ( 3 ~ ) was found to be 0.08 pg of Pb and an RSD of 4.1% was achieved for 0.8 pg of Pb" ( n = 15). An upper limit for the working range of 5 pg of Pb" per sample was found. Interferences of cationic and anionic sample components were found to be negligible. Application to the determination of Pb" in digested soil samples is described and the method was validated by using certified reference materials.Keywords: Solid-phase estr-a:c*tion; flame atomic ahsoiption specti-onietiy ; cwwn ether; $'owl injection; lecxd; soil. Introduction Lead belongs to those trace heavy metals which are of major interest in environmental protection owing to its cumulative toxicity. Lead is still emitted into the biosphere in considerable amounts owing to its application as a fuel additive.' Various analytical techniques are available for monitoring Pb" concen- trations in environmental matrices. For analyses at concentra- tions in the pg kg-I or pg dm-? range, ETAAS2 and, more recently, ICP-MS3,4 have been successfully applied, although both techniques suffer from their sensitivity to matrix interfer- ences due to the high solid concentration in the sample.Spectrophotometric methods536 and ion chromatography7 re- quire prior preconcentration steps, owing to their insufficient sensitivity. The same holds for the less sensitive atomic spectrometric techniques ICP-AES and FAAS. The latter technique, although relatively old, still offers advantages over other methods in terms of cost effectiveness and sample throughput. The combination of preconcentration procedures with FAAS yields high sensitivity but is usually accompanied by a decrease in sample throughput, owing to the tedious and time-consuming sample preparation. This problem can be overcome by using an on-line flow injection (FI) system for preconcentration and separation of the analyte from matri- ' To whom cot reymidence \hould be addressed.c e ~ . ~ - ~ O The application of an automated FI system increases the speed of the preconcentration process and additionally de- creases the potential of interferences caused by matrix effects, which are critical in trace analysis. Various methods of achieving preconcentration have been adapted to an FI manifold, including liquid-liquid extraction, precipitation, ion exchange, immobilization, electrodeposition and solid-phase extraction.9-23 An FI in-valve preconcentration column pro- vides the additional advantage of offering the possibility of a calibration using only a single standard solution.23-". The use of ion-exchange and chelating resins or the sorption of organic complexes on reversed-phased silica or polystyrene has been applied for the preconcentration of trace ele- ments.9,17,1x- 21-26 However, these procedures suffer from some drawbacks.First, sorption of trace metal ions usually requires pre-adjustment of the sample solution to an appropiate pH value. Depending on the chosen procedure, buffering of the sample solution to a certain pH value considerably higher than 2, usually around 5 , is necessary. After being pre-treated, sample solutions are acidic either owing to preservation, such as for water samples by addition of nitric acid, or because of digestion of soil samples using strong acids. Hence the above- mentioned preconcentration methods require addition of buffer- ing reagents and consequently make sample contamination possible. Second, the presence of complexing agents in the sample solution such as naturally occurring phosphate and hurnic substances or anthropogenic compounds such as EDTA could interfere with the separation by forming strong complexes that are not retained by the column.27 A method that allows the preconcentration of lead from acidic solutions could overcome these problems for obvious reasons: (1) sample solutions could be analysed as they are after acidic pretreatment and (2) interferences due to complexing agents are minimized as their complexation strength towards Pb" decreases at low pH values.A method for the separation of Pb" from nitric acid solutions has been suggested by Horwitz and co-workers,2x-Z9 who developed an extraction chromatographic resin consisting of bis-tert-butyl-cis-dicyclohexano- 1 8-crown-6, which is coated on Amberlite XAD-7 polystyrene, for the selective radio- chemical separation of 90Sr from other radionuclide~ They found that this resin, called Sr.Spec, also showed high distribution coefficients for Pb" at nitric acid concentrations from to 8 mol din-?, even exceeding those for Sr" by a factor of 102-103.Lead(ir) can be stripped from the column with a variety of reagents. 3() The main purpose of this study was to investigate the feasibility of the commercially available Sr.Spec resin for an FI procedure for on-line in-valve Pb" preconcentration and separation coupled to an FAAS detection system. The proce- dure. which is cost-effective, selective, sensitive and precise,was applied to digested soil samples. The accuracy of results was validated by analysing certified reference materials and comparison of the results with those obtained by ETAAS.Experimental Reagents Analytical-reagent grade chemicals and water purified in a Milli-Q system were used. HN03 (60% m/v), NHI (30%) m/v) and H202 (30% v/v) solutions (Merck, Dannstadt, Germany) were of Ultrapur, Suprapur and Extrapure grades, rerpectively. Standard rolution\ of metal ions were prepared by appropriate dilution of aqueous 1000 mg dtn-3 stock standard solutions (Johnson Matthey, Wayne, PA, USA) with 1 mol dm-3 nitric acid. The eluent solution used was 0.05 inol dm-? ammonium oxalate. Sr.Spec SPS (80-100 pm grain diameter) was pur- chased from Eichroin (Darien, IL. IJSA). Apparatus A Perkin-Elmer (Norwalk, CT. USA) AA 3300 atomic absorption spectrometer equipped with an air-acetylene burner, PC and Lab Benchtop software was operated under the conditions recommended by the manufacturer. The analytical wavelength was set at 217.0 nm.The niinicoluinn (25 mm X 3 mm id) (Omnifit, Cambridge, LJK) was prepared by filling with the Sr.Spec resin. The two ends of the column were plugged with porous polyethylene frits (pore size 20 pm) and covered with fittings for PTFE tubing (0.7 mm id). The column was connected with the injection valve rotor of a commercial FI system (FIAS 400, Perkin-Elmer, Uberlingen, Germany). The minicolumn was positioned to replace the \ample loop of the valve. A microwave digester (1200 Mega, Milestone MLS, Riviera Beach, FL, USA) was used for soil sample preparation. Flow Manifold The manifold used is shown in Fig.l ( u ) . All connections were made with 0.7 mm id PTFE tubing; a 2.5 cm length of the same C E W D s I E D Position 1 Position 2 Fig. 1 ((1) Manifold for FAAS deterinination of lead by in-valve minicolumn preconcentration and elution. (h) Valve positions: ( 1) loading and (2) clution. S, saniplc solution; E, cluent; P, pump; I, injection valve; C, minicolumn; B, bypass; D. detector (AAS); and W, waste. tubing was used for coupling between the injection valve and the FAAS system. Operation of valve position5 is depicted in Fig. l(b).31 Control for changing the valve position and flow rate for the sorption and elution steps i \ effected \,iu the computer program Lab Benchtop. Data acquisition is also managed by the software. Lead(~i) is \orbed on the resin by pasring a standard or sample solution (position 1 ) through the column 1 7 i c 4 the port\ a + 4 and 2 + c for a desired period, while the eluent flows through the bypas5 loop to the FAAS system 17iu the ports d + 1 and 3 + b.After switching the valve into position 2 for elution, the staiidard/sample solution flows through ports a + 5 to waste while the eluent pa\\e\ through the column 1 3 i ~ ports d + 2 and 4 + b. Lead(rr) is then desorbed and flows to the detector. The reverse flow direction\ of the sample loading and of the elution help to prevent blockage in the column, which may be caused by accumulation of the resin at one end of the column if the loading and elution pa\\ed through the coluinn in the same direction. The eluent tlow rate was optimized to 4.0 cm3 min-1, which wa\ slightly higher than the nebulizer uptake, and gave maximum absorbances.Procedure Record a calibration curve by passing a single standard solution through the minicolumn at a constant flow rate of 4 cm-i min- I for various time intervals (0.5-7 min) depending on the concentration of Pb" in the solution. Switch the valve position to pass the eluent stream through the column to the AAS system. Plot the absorbance of [he transient absorbance IWSUJ the total amount of Pb" desorbed from the column. Treat sample solutions analogously and convert absorbance into mass concentration units using the calibration curve. Application to Soil Samples Accurately weigh 0.3 g of sample (less than 200 tnesh, USS, dried at 1100C) into a digestion autoclave.Acid leach the \ample with a mixture of 5 ctn7 of concentrated HNO? and 1 cni? of H202 (30% v/v) in a microwave digester at 2.50 W for 5 min (to destroy the organic matter). After cooling, open the vessel to release the pressure and carry out the digestion for I0 min at 500 W and for another 5 inin at 650 W. After digestion, filter the wiiple by u\ing a syringe filter (0.45 pm pore size). Dilute the sample to 100.0 tin? with 1 mol din- nitric acid. Store the diluted \elution in a plastic bottle before analysis. This procedure. although satisfactory for Pb, will not be satisfactory for all elements. Aqua I-egia digestion is now prefered by most workers to give a 'pseudo-total' concentration of metals. Results and Discussion Optimization of Flow Rate Tn order to optimize the elution tlow rate, the following parameters were kept constant: the flow rate for loading (preconcentration) (4 cm3 min-I).the loading (or preconcen- tration) time ( 1 niin), the concentration of the standard solution ( 0.1 pg cm--' Pb" in 1 mol dm-? nitric acid) and the eluting solution (0.05 mol dm-3 ammonium oxalate). Fig. 2 shows absorbance values as a function of eluent tlow rate and time from the start of elution to the detected peak maximum. The elution flow rate selected was 4 cin7 min-I. Lower tlow rates caused higher dispersion and longer periods were needed for the determination. At higher tlow rates leakage of the system was observed. owing to the back-preswre of the column. The effect of the loading tlow rate on the signal was not so pronounced as the effect of elution flow rate.At low flow rates, a slight increase in signal with increasing flow rate was observed, which might be due to non-linearity of solution flow at low pumping rates. A slight decrease in signal was observed at flow ratesAnalyst, October 1996, Vol. 121 1415 higher than 4 cm3 min-1 owing to broadening of the Pb" sorption zone on the column, resulting in a higher dispersion during desorption. A loading flow rate of 4 cm3 min-1 was chosen as the optimum compatible with sensitivity and sampling frequency. Effects of Different Eluents Strong complexing agents have been suggested for the deso- rption of Pb" from crown ether re~in.~O Several chelating agents, namely oxalic acid, tartaric acid, citric acid and their sodium and ammonium salts, were examined for their suitability in the FI-FAAS procedure.Table 1 shows the absorbances obtained when the eluents at a concentration of 0.05 mol dm-3 were applied to desorb 2 pg of Pb" from the column under the optimum flow conditions. The effectiveness of the eluent was found to be dependent on the pH of the eluent solution. Adjusting an eluent solution of 0.05 h?ol dm-3 oxalate to different pH values (1.03,2.05, 3.01, 4.00, 5.05, 6.00 and 7.05) by stepwise addition of aqueous ammonia yielded increasing values for the absorbances (0.23, 0.44, 0.56, 0.62, 0.76, 0.80, and 0.80, respectively). Conse- quently, sodium and ammonium salts were found to desorb Pb" from the column more efficiently than the acids. This pH dependence of the elution is due to ( I ) the increasing deprotonation of the chelating reagent at a higher pH and, thus, a shift of the complexation reaction: Pb'2 + H2C204 C Pb(C204)22- + 4H' logK = 15.9'2 towards the right and (2) the decrease of the distribution coefficient for the Pb" sorption on the crown ether resin with decreasing nitrate concentration. However, with sodium salts, clogging of the burner was observed after a certain time.From all the eluents tested, ammonium oxalate was chosen as the most appropriate. The effect of ammonium oxalate concentration was further investigated and the results are displayed in Fig. 3. A solution of I I I ' 10 2 3 4 5 6 FIOW rate/cm3 min-' Fig. 2 Effect of eluent flow rate on absorbance (arbitrary units) and residence time (s).Table 1 Effect of various eluent solutions on absorbance Eluent Absorbance Component (arbitrary (0.05 mol dm-') PH units)* Oxalic acid 1.32 0.22 Tartaric citrate 1.82 0.40 Sodium citrate 7.35 0.77 Citric acid 1.90 0.21 Ammonium oxalate 6.47 0.82 * Average of triplicate results. 0.1 mol dm-3 or higher concentration caused clogging of the burner and high background signals, leading to lower absor- bances with noisy peaks and baselines. Solutions of concentra- tions less than 0.05 rnol dm-3 also yielded lower absorbances. Consequently, an eluent concentration of 0.05 rnol dm-3 was chosen as the optimum. Effect of Nitric Acid Concentration in the Sample Solution It has been reported that the selectivity and sorption capacity of Pb" on the sorbent are affected by the concentration of nitric acid.28-30 The effect was studied by passing a solution of Pb" t0.5 pg cm-3 in nitric acid solution (0.1-4.0 mol dm-3)] through the column at a flow rate of 7 cm3 min-1 for 1 min.The sorbed Pb" was eluted as described previously. The results (Fig. 4) show that the nitric acid concentration should be in the range 0.8-2.0 mol dm-3. Limit of the Analytical Working Range The upper limit of the analytical working range was determined by passing a 0.5 pg ~ m - ~ Pb" solution through the column at a flow rate of 4 cm3 min-l for various loading periods (1-8 min) followed by elution and detection of the peaks. A plateau is reached at 6 pg Pb". This indicates the maximum amount of Pb" which can be determined with the present FI-FAAS set- up.Under these conditions, 6 vg of Pb" can be sorbed on the column filled with 69.6 mg of sorbent (ix., 86.2 pg of Pb" per gram of sorbent). The limitation of the analytical range, however, is determined by the FAAS detection, not by the loading capacity of the column, which is 55.4 mg of Pb per gram of s0rbent.3~ The upper limit of the working range was fixed as 5 pg of Pb": up to this mass a linear relationship between the amount of Pb" and the absorbance was obtained. 1.0 ' I I I I I X I I 0.4 1 I I I I 0 0.1 0.2 0.3 0.4 Eluent/mol Fig. 3 trary units). Effect of ammonium oxalate concentration on absorbance (arbi- 1.0 ; I I I I , 4 Y 0.2 1 I I I I I I 0.0 0.8 1.6 2.4 '3.2 4.0 HNO, /mol Fig. 4 absorbance (arbitrary units). Effect of nitric acid concentration in the sample solution onThe recovery of Pb" throughout the separation step was checked by sorbing Pb" from various standard solutions (0.1-2.0 pg cm-' Pb" on the column and subsequently eluting with 0.05 mol dm-? ammonium oxalate solution.The eluent was collected off-line and the resulting solution analysed by FAAS. The mean recovcry of Pb" over the whole preconcen- tration procedure was found to be 9c).2L;b. In terje re n ce s Various ions that occur in environmental samples were monitored for their intluence on the absorbance signal of a 0.2 pg cm- + Pb" standard solution (using the recommended procedure, with a I min loading time). Table 2 shows the mass ratios of the potential interferents to Pb" which were examined. No effect on the Pb" signal could be detected which caused the relative error to exceed an acceplable value of 5%.Detei-niination of Lead(ii) in Soil Samples A calibration curve can be constructed by either time-based or volume-based calibration. A particular amount of' Pb" loaded on the column from one standard solution preconcentrated at different time periods or from Pb" solutions of different concentrations at a fixed loading time always gives peaks with the same absorbances. A comparison of the two calibration modes up to 4.5 pg of Pb" yielded [he same linear plot (v = 0.040.t - 0.001 ; 1.2 = 0.9958). The detection limit (30) was 0.08 pg of Pb" and the RSD was 4.1% for 0.8 pg of Pb" (17 = IS; 1 niin loading time. with a flow rate of 4 cm? inin-' for 0.2 pg cn-? Pb"). A quantitative assessment of the over-all procedure, using the parameters suggested by Fang,?? is shown in Table 3.Comparing these parameters with those in other reports show that the present procedure shows no ctriking advantage over other well known FI-precoiicentrHtion methods using chclating resin\ with respect to wisi tiv i ty enhancement and rapidity.33 However, the tolerance of the crown ether resin towards interferences duc to sample rnatricci i i higher than that for most other described FI on-line column preconcentration Table 2 Mass ratios of potentially interfering ions which may be tolerated ( < 5% error) in the analysis of a solution containing 0.2 pg cni - - ? of Pb" passing through the coluniii for I min at a flow rate of 4 ctn' minr I Ion : Ph" Inas\ ratio 2500 2000 200 200 250 Table 3 Chai-acteiixtic\33 of the proposed procedure Perform ance Preconcentration Preconcentration parameter ~ time 3 min time I niin FF 2 1.6 7.9 jh-l 17 4 0 CE'/rnin ~ I 6 4.8 C//min ~ I 0.56 0.56 DL/@ dlll 6.7 2 0 * EF.enrichment factor (the ratio 01' the linear calibration curve slopes without and with preconceiitration): ,/; samplc frequency (in samples per hour); CE conccntration efficiency ICE = EF (f/60)]; CI. consuinptivc index (CI = \',/EF: V, = sample volume in em-'); L)L, detection limit (30). methods. Most of the latter methods have been applied for the preconcentration of water samples with only a low matrix concentration. Consequently, the use of the crown ether resin is recorninended for sample solutions containing complex matri- ces such as acid-digested soil samples.Additionally. the other advantage of no preadjustinent of the pH of the sample solution before loading as described earlier allows a simpler FI manifold to be operated. In order to assess the accuracy of the results obtained by the described method, two certified reference soil samples werc digested as previously described. The solutions obtained were adjusted to I mol dm-? nitric acid and were then analysed by the proposed procedure. The results for the digested soil samples and the results for an ICP multi-element standard solution agree with the certified values (Table 4). Some soil samples collected from a paddy field near a zinc refinery in Tak Province, Thailand, were analysed. The results obtained by the proposed procedure after Cali bration by external standardization or standard addition compare well with those obtained by direct ETAAS (Table 5 ) .The recoveries for added standards were 9598%. The quoted detection limit (0.08 pg of Pb) can be achieved with sample/slandard throughputs of up to at least 25 h-1 . Conclusion The commercially available resin Sr.Spec, which has mainly been proposed for the separation and determination of radio- strontium, can be applied to an FT procedure for on-line in-valve Pb" preconcentration and separation coupled to FAAS. The method is especially recommended for the analysis of acidic solutions such as digested soil samples, as no extra addition of buffering reagents is necessary. The proposed method incorpo- rates the advantages of on- line separation and preconcentration, namely speed, precision, selectivity, sensitivity and a high tolerance to interfering ions.Enrichment factors of 10-50 can Table 4 Determination of Pb" in certified reference Pb contentlyg g I Sample Found Reference range IAEA soil 7 58.4 -t- 1.9 55-7 1 Thai \ o i l - 1 17.0 f 0.5 16.9-1 7.3 ICP, multi-eleiiicnt standard V1 (Merck) 9 8 f 0.6 9.5-10.5 ' Duplicate determination\. Table 5 Determination ot Pb" in ,oil \ample5 by FI-FAAS (calibrated by external 5tandardiiation or 4tandard addition) and ETA AS Pb content'/pg g- Sample External Standard NO. standardization addition ETAAS' I 2 3 4 5 6 7 8 9 10 22.5 f 0.4 18.2 k 0.2 20.3 f 0.3 21.2 k 0.2 25.2 k 0.3 19.5 f 0.4 lx.O f 0.2 22.2 f 0.3 24.2 f 0.2 20.2 f 0.3 22.4 f 0.2 18.2 f 0.4 20.4 f 0.3 21.2 * 0.5 25.2 * 0.4 1'4.4 f 0.4 18.1 i 0.2 22.3 f 0.2 24.1 f 0.2 20.4 f 0.4 22.3 f 0.3 18.0 * 0.4 20.3 f 0.3 21.4 + 0.6 25.0 k 0.5 19.6 f 0.6 1 x 3 f 0.5 22.2 f 0.6 24.0 f 0.4 20.2 * 0.5 ' Duplicate determinations.Following the procedure in ref. 34.Ancilyst, 0ctobc.i- 1996, Vol. 12 I 1417 be achieved. A single standard solution can be applied for calibration. Good signal stability and reproducibility can be obtained from the column over SO0 complete cycles. Require- ments for determination of Pb" in soil samples are satisfied using FAAS with FI in-valve minicolumn preconcentration and separation using a solid-phase extraction resin. The authors are grateful to the Deutscher Akademischer Austauchdienst (DAAD) and especially to Professor B. Kanellakopiilos for P.S.'s fellowship at the Forschungszentrurn Karlsruhe.The authors thank Dr. C. Taylor, Liverpool John Moores University, UK, for useful discussions. References 1 2 3 4 5 6 7 8 9 I 0 1 1 12 13 14 Tho Bioc~iicniisti~y of Lcatl iir rlir Etri~ii-orrni~~nt. ed. Nriagu, J. 0.. Elsevier, Amstei-dam. 1978. Sekerka. l . , and Lechiier, I.. ,41iul. Cliinr. A(.ttr, I99 I , 254, 99. Murphy. K. E., and Paulsen, P. J.. Fi.o.srnius' .I. Aiiul. Chew., 1995, 352. 203 Arunachalam, J.. Mohl, C.. Ostapczuk, P., and Enions, H., Fi-c)sririrr.s' .I. Anul. Clrcni.. I99 I , 340. 2 17. Savin, S., Petrova, T., Dzherayan. T., and Relkhshtat. M., Fi~,soi7irr,~' .I. AnuI. Clicni., 199 I , 340. 2 17. Rakhman'ko, R., Tsvirko G., and Gulevich: A,, Zlr. AIIFII. Khini., 1991, 48, 1525.Cardellichio, N., Cavalli, S., and Rivillo. J. M., .I. ChiwnutogI. .. 1993, 640, 207. Fang. Z.. RBiiEka, J., and Hansen, E. H., Anul. CIrini. Actu, 1984, 164, 23. Tyson, J. F.. Anuly.vt, 19x5, 110. 419. Fang, Z.. Xu, S . , and Tao, G., .I. Anul. At. Spc~cfiwn., 1996. 11. 1. Cabonell, V., Salvador, A,. and dc la Guardia. M., Fvr.senizi.s' .I. Anul. Ciwrn.. 1992, 342, 529. ~ l i ) ~ , - l t ! j i ~ ( ~ t i o t i Atoinic S1?'"ti'o.s(.0~)4', ed. Bul-guera. J . L.. Marcel Dekker. New York. 1989. RBiiCka, J.. and Hasen. E. 14.. F l o ~ i , lti,jc(.tion Anu/y,>i,\+ Wiley, New Yoi-k, 2nd edri.. 198X. Valcarcel, M., and Luque de Castro, M. D. FIi)M'-li?jC(.tioii A n u l ~ s i s . Prirrc.iplc u i i d Applic~rtion, Ellis Horwood, Chichester. 1987. ~ I5 16 17 I8 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 Valcarcel, M., in ,%mipie Iritrocluction iir Atonric., ~ / ) c . ( , f i . o ~ ~ ~ , [ ) ~ ) y (Anulyticul Spc.c.tl-oscopy Lilwui:): Vol. 4). ed. Snetldon. J ,. Elscvier, Amsterdam, 1990, ch. 1 I . Kurlberg, B.. and Pacey, G. E., Flobt' Zrr,jec~tiorr Anulj Guide. Elsevier, Amsterdam, 1989. Olsen, S., Pessenda, L. C. K., RbiiEka. J., and Hasen, E. H., Anulwt, 1983. 108. 905. Caroli, S., Alimanti. A., and Petrucci, F., Anul. Chim A r t ~ r , 199 I , 248, 241. tioklu, ti.. and Akman, S.. AiiuI. Lett.. 1990. 23. 569. Zhuang. Z., Wang, X., Yang, P.. Yang . C., and Huang, R. Can. .I. AppI. Spcctrmsc., 1994, 39, 100. Naghmush. A. M., Pyrzynska, ti.. and TrojanowicL, M., 'lirltrntu. 1995, 42, 85 1. RiiiiEka. J., and Ai-ndal, A.. Auul. C'liirri. A(.ta. 1989. 216. 243. Bysouth. S. I<., Tyson, J. F.. and Stockwell. P. B.. Ancxi. Chini. Actu, 1988,214, 329. Fang, Z.. and Welz. B, .I. AriuI. A I . Sprr'ttmv., 1989, 4, 543. Cirudpan. ti., Laiwraungrath. S.. and Sooksamiti, P., Ancrlyst, 1995, 120. 2107. Lancaster. H. L., Marshall. G. D., Gonzalo. E. R., KEiEka. J., and Christian, G. D.. Analwt. 1994, 119. 1459. Birrba. P., anti Willmer. P. G.. Wusscr. 1982. 58, 43. Chiarizia, R.. Horwilz, E. P., and Dietz. M. L.. Solivnt h t r . 1017 E.ur,h., 1992, 10, 313. Chiarizia. R., Horwitz, E. P., iind Dielz, M. I>., ,Col\~~nt Exfr. Ion E.i-cIi., 1992, 10, 337. Horwitz, E. P., DietL. M. L., Rhoads, S., Felinto, C., Gale. N. H., and tioughton: J., Anul. Chitii. A d a , 1994. 292, 263. FIAS 300 M~inziul. Perkin-Elmer. Norwalk, C?', 1993. NIST Stmiriui-d Kofri-rrices Datuhcxsc 46, National lnstitute of Standards and Technology. Gaitherburg, MD, 1995. Fang. Z., Fiow Injection Sepaiwtioii utid Pi.cr,onc.eiitr-atinrr, Verlag Chemie. Weinheini, 1993. pp. loft'. Anuiyticul hlcthocls,fi~r~ Fni.ricrc c /\AS R.3.32, Perkin-Elmer, Norwrzlk, CT.
ISSN:0003-2654
DOI:10.1039/AN9962101413
出版商:RSC
年代:1996
数据来源: RSC
|
20. |
Slurry preparation by high-pressure homogenization for the determination of heavy metals in zoological and botanical certified reference materials and animal feeds by electrothermal atomic absorption spectrometry |
|
Analyst,
Volume 121,
Issue 10,
1996,
Page 1419-1424
Yanxi Tan,
Preview
|
PDF (869KB)
|
|
摘要:
Arzulyst, Ortohel- 1996, Vol. 121 (1419-1424) 1419 Slurry Preparation by High-pressure Homogenization for the Determination of Heavy Metals in Zoological and Botanical Certified Reference Materials and Animal Feeds by Electrothermal Atomic Absorption Spectrometry High-pressure homogenization was evaluated for the preparation of slurries suitable for the determination by ETAAS of Cr, Cu, Fe, Mn, Ni and Se in soft organ tissues (liver and kidney), certified reference materials of biological and botanical origin and animal feeds. Frozen fresh organ tissue, (2 g) or certified reference material or dried, ground plant material (0.1 g) was blended, at high speed with 20 ml of ethanol-water (1 + 9 v/v) containing 0.25 % m/m of tetramethylammonium hydroxide and the resulting mixture was subjected to homogenization at 38.9 MPa.After four passes through the homogenizer, the resulting solution was suitable for analysis by ETAAS. Capping the flat valve head of the homogenizer with a ruby disc appreciably reduced (but did not eliminate) metal contamination introduced by the processing. Homogenization of botanical reference materials or dried animal feeds resulted in preparations with variable amounts of residual fibres and particulate matter in the resulting suspensions. Nonetheless, all the Cu and Mn and virtually all of the Fe had been transferred to the supernatant fraction and remained with that fraction for at least 10 d. The addition of EDTA to the solvent modestly increased the mobilization of Fe from the matrix but also increased the contamination from the homogenizer.The slopes of the calibration curves generated by the method of standard additions were not significantly different from those of calibration curves generated with aqueous standards in a homogenized blank indicating that there was no significant matrix effect for any of the analytes in the nine reference materials, liver or kidney or the five animal feed samples and that aqueous standards could be used to calibrate the instrumental response. Keywords: Holno,~Pni,cxtion; s1iu-i;~ inti-oditctioii; elec,tt-othcsrrzul atomic- uhsol-prior? spec.ti.ometr:~; hoturiic.ullhiologic~al c*ci*tififi‘Pd i-ejerencc niutcrials; unimul ,feeds: c.hrw?riun?; coppi-; ii-on, munguriesc; nickel Introduction Conventional sample preparation of biological materials prior to atomic spectrometry involves complete solubilization of the ’ To whoiii coi-respondcnce should be addreswl analyte and matrix, which is achieved typically by oxidative mineralization of the organic matter and solubilisation of the residue in a suitable solvent.Even for microwave-assisted digestions, whereas complete dissolution can usually be achieved by a suitable choice of digestion conditions, complete decomposition of the organic matrix in biological and botanical samples is appreciably more difficult. Often complete mineral- ization is achieved only with supplemental treatment of the digested matrix with H202 or even HF.5 However, these digestion procedures can be labour intensive, time consuming and prone to contamination errors.In consequence, there is a continuing interest in the development of simplified sample preparation techniques. The preparation and introduction of slurried samples continue to attract considerable attention because of the ease with which quasi-stable preparations can be generated and their compatibility with conventional liquid handling techniques. Within the general field of solid sampling analysis,6P1 it is the use of slurried samples that has become the most popular approach to trace element determination. Direct atomization from the solid state can provide excellent sensitiv- ity but the interpretation of the results can be complicated by molecular absorptions and/or scattering from the matrix, which can produce sufficiently large background signals to overwhelm the compensation capabilities of common deuterium back- ground correction systems.Additional difficulties can include sample inhomogeneities. the requirement for repeated microweighings and the lack of suitable calibration standards and techniques. A variety of sample pre-treatment procedures and addi- tives12--’X have been described and evaluated for the production of quasi-stable suspensions of samples prior to analysis by atomic spectrometry. Alternatively, suspensions with a ten- dency to segment rapidly have been reproducibly sampled by using ultrasonic agitation, 19 air or argon2() bubbling, vortex mixing21 or magnetic stirring.2’ Partial digestion procedures to produce carbonaceous slurries have also been successfully applied to the analysis, by ICP-AES, of a series of standard reference materials of biological origin.2-3 Various alkyl- ammonium hydroxide formulations have been used extensively to solubilise t i ~ s u e , ~ ~ ~ ’ ~ particularly those of zoological origin.The generic term ‘homogenizer’ has been applied to any piece of equipment that disperses and/or emulsifies (including a turbine blade mixer, an ultrasonic probe, a high-shear mixer, a colloid mill, a blender or even a mortar and pestle). One particular device for homogenization consists of a positive1420 Analyst, Octohei- 1996, Vol. 121 displacement pump and a homogenizing valve that forms a restricted orifice through which the product flows. The successful use of this device to prepare emulsions and/or dispersions of meats suitable for transmission FTIR spec- trometry has been reported.28 Recently, high-pressure homo- genization has also been evaluated for the rapid preparation of biological soft tissues (liver and kidney) prior to slurry introduction ETAAS.29 The homogenization of 2 g of fresh tissue or 0.1 g of powdered reference material together in 20 ml of ethanol-water (1 + 9 v/v) containing 0.25% m/m of tetramethylammonium hydroxide (TMAH) resulted in prepara- tions that contained no visible suspended solids and that were stable to Cd, Cu or Pb content for at least 6 d.The principle limitation of the technique was the introduction of metal contaminants into the sample by the homogenization process. The objectives of the current study were (i) to evaluate the slurry preparation and introduction technique for the determi- nation of other analyte elements (including Cr, Ni, Mn, Fe and Se) in biological tissues, (ii) to extend the applicability of the technique to selected botanical samples and (iii) to reduce the levels of heavy metal contamination introduced into the product slurry by processing at high pressure.Experimental Samples CRMs were purchased from the National Research Council of Canada (NRCC) or the US National Institute of Science and Technology (NIST). Samples of animal diet mixtures destined for a zoo were chosen to contain a variety of plant and animal materials including timothy grass, bamboo leaves, whole smelts, cricket chow and a panda bear mixture (contents unspecified). Sample preparation For CRMs or dried feeds (ground to pass a 0.5 mm screen in a Cyclotec sample mill, Tecator, Hoganiis, Sweden) accurately weighed sample (approximately 0.1 g) was added directly to 20 ml of ethanol-water ( 1 + 9 v/v) containing 0.25% m/m of TMAH in a 50 ml beaker.The mixture was maceratedrolended at high speed (20000 rpm, 60 s, in an SDT Tissuemizer; Tekmar, Cincinnati, OH, USA) and the resulting suspension was then processed through a 20 ml capacity flat valve homogenizer (EmulsiFlex Model EF- €33; Avestin, Ottawa, ON, Canada) capable of developing 138 MPa when provided with compressed air (689.4 kPa). Each slurried sample was re- processed through the homogenizer three more times. Frozen organ tissue was thawed and a 1 cm wide transverse section of the organ was excised with a stainless-steel scalpel, accurately weighed (approximately 2 g).blended at high speed with 20 ml of solvent mixture and homogenized. Dilutions, when required, were performed with solvent mixture that had not been homogenized. Homogenizer The valve stem of the screw-cap assembly of the EmulsiFlex Model EF-B3 homogenizer (Fig. 1) was modified by gluing a polished 4 mm diameter x 2 mm thick ruby disc manufactured from a ruby sphere (from an HPLC check valve) to the polished surface of the flat-faced head. Sample macerate was transferred into the sample chamber via the inlet port, which was then sealed with a fine-threaded screw-cap. The stainless-steel piston (connected to a pneumatic multiplier) then forced the fluid through an aperture and the homogenate was collected from the sample outlet.Each sample was reprocessed three more times with the valve stem retracted slightly so as to provide a slightly larger gap. FAAS Prior to determinations for Cu, Fe and Mn, feed samples were dried to constant mass and ground to pass a 1 mm screen. Accurately weighed aliquots of ground plant material (approx- imately 2 g) were digested at room temperature in a perchlorate fume hood with 25 ml of 70% HN03-HC104 (4 + 1 v/v) until gas evolution had ceased, then heated at 80 "C until a clear, yellow solution was obtained. The resulting digests were diluted prior to analysis. ETAAS Analyses for chromium, copper, iron, manganese, nickel or selenium were performed using a hot injection technique on a Varian (Palo Alto, CA, USA) Model 300 ETAAS system equipped with an autosampler, conventional hollow-cathode lamps and Zeeman-effect background correction.Analytical operating parameters for each analyte element are presented in Table 1. Calibration Quantification was performed by both the method of external standards (ES) and by standard additions (SA). ES consisting of appropriately diluted processed reagent blank, and up to four levels of standard were prepared automatically by the sample introduction device. Background-corrected integrated absor- bance, resulting from three replicate injections of each diluted standard, was used to define the best-fit regression equation. For SA calibrations, 10 pl aliquots of processed fluid was fortified with 2, 5 , 10 or 20 pl of aqueous standard chosen to result in a range of peak areas including signals which were half and at least twice the signal for the fortified sample.The data were modelled by least-squares linear regression. Quantification was performed by dividing the y-intercept of the regression equation by the slope of that equation and the over-all standard error of the estimate (SEest) was calculated from Student's t-test was used to identify significant differences between the slopes or between the y-intercepts of regression for different sample matrices. The f-test was used to detect significant differences between regression models. Results and Discussion Previous studies20 with high-pressure homogenization had demonstrated that copper, cadmium and lead concentrations could be reliably determined in soft organ tissues or in zoological reference materials by direct slurry introduction into the ETAAS system.Moreover, there was no evidence, over 6 d, of any analyte metal segmentation within solutions which resulted from a combination of high-speed blending and Sample in! Ruby disk, k \Fine threaded screw-cap Exploded view of the homogenizing valve and sample compart- Fig. 1 ment of the Model EF-B3 homogenizer.Analyst, October 1996, Vid. 121 1421 homogenization. In addition, the instrumental response to the analyte metal could be calibrated with aqueous external standards. However, the applicability of the technique was limited by the appreciable levels of contamination introduced into the sample by the homogenization step. It was postulated that stainless-steel surfaces that were exposed to the homoge- nized fluid, particularly the flat face of the demountable valve head, were the principle sites responsible for the contamination.Further, capping the valve head with an inert surface capable of withstanding the impact of the jet of fluid exiting the homogenizing orifice might reduce the levels of contamination appreciably. It had been reported") previously that zirconium oxide beads used to reduce the particle size and to mix particulate solids introduced appreciable levels of Fe, Cr and A1 but that silicon nitride or boron carbide provide good abrasion resistance and offer little likelihood of contamination. However, even for the relatively lower pressure requirements of pistons and check valves for HPLC, sapphire, ruby and zirconium oxide are preferred over other ceramic materials for their superior resistance.A disc of ruby that was manufactured by grinding and polishing a 4 mm diameter ruby ball, was glued temporarily to the flat face of the demountable valve head. Solvent mixture (20 ml) was blended and then homogenized four successive times (in the presence or absence of the ruby disc) prior to ETAAS analysis for Cr, Cu, Fe, Mn, Ni and Se. Analyte concentrations (Table 2) were expressed as if the solvent had contained 0.100 g of sample. The decrease in heavy metal content of the homogenized fluid was, in all cases, reduced appreciably (2-30-fold). Nonetheless, the level of contamina- tion remained appreciable even in the presence of the ruby cap. The slurry preparation technique, with the ruby-capped homogenizer, was then applied, in preliminary studies, to five biological CRMs and to four frozen liver and kidney tissues from moose and ring tail deer.As was reported previously, the resulting homogenates appeared to be uniform and contained no visible particulate material. Analyses for chromium, iron, manganese and nickel in the five CRM homogenates (with calibration by standard additions) provided estimates (Table 3) that did not differ significantly from their certified concentra- tions. No analyte Ni was detected in the bovine muscle CRM (0.05 pg g-I), however, this certified concentration was less than half that introduced into the homogenate during prepara- tion. The repeatability of the analyses, as measured by the relative standard error of the estimate (RSE,,,), did not indicate any problems with the repeatability of transfer of any of the homogenates to the graphite tube; however, as the concentration of analyte approached the level of contamination, the difference (in all cases <20%) between the experimental and certified concentrations tended to increase.Frozen liver and kidney tissue were analysed only for Cr and Mn. ETAAS estimates of [Cr] in homogenates were consis- Table 1 Graphite furnace operating parameters* for the determination of Cr, Cu, Fe, Mn and Ni Parameter Chromium Copper Iron Manganese Nickel W avelengthlnm Lamp current1A Slit w idthlnm Injection temperaturePC Preinjection Last drying step, 5 SPC Pyrolysis ,equence Cooling Alomization Measurement time/s Matrix modifier 3 5 7.9142 9.0 7 0.2 6 0 No None 20 s ramp to 1450 "C, 40 s hold 7 s ramp to 40 "C, 10 s hold 1.2 s ramp to 2400 "C, 8 s hold 8 3 pl of 20 g I- ' Mg(N03)? for 10 pl sample 3 24.71244.2 4 0.110.5 60 No None 5 s ramp to 900 "C, 20 s hold None 1.0 s ramp to 2300 "C, 2 s hold 5 pl of 1% m/m NH4N03 for 10 p1 sample 3 386.0 5 0.2 60 Yes 250 10 s ramp to 1100 "C. 21 s hold None 1.2 s ramp to 2400 "C, 4 s hold 5 pI 500 ing I-' Pd + 2.5% citric acid for 10 p1 sample 6 403,11279.5 5 0.2/0.2 60 Yes 250 10 s ramp to 1200 "C, 20 s hold None 0.7 s ramp to 2200 "C, 2 s hold 3 5 v1 of so0 mg 1-1 Pd + 23% citric acid for 10pl sample 232.0 4 0.2 60 No No 5 s ramp t o 800 "C, 20 s hold None 0.9 s ramp to 2400 "C, 4 s hold 4 None ' Each step of the furnace programmes (with the exception of the read step) was performed in the presence of argon purge gas (3 I min-I).Table 2 Analyte coiicentrdtionj (pg g ~ I sample) in 20 ml solvent mixture (+ EDTA) following various mixing treatments Concentrations in the homogenized tluid are expreswd 'ij it the 5olvent mixture had contained 0.100 g of jdmple Homogenization treatment* Aluminium Cadmium Chromium Copper Iron Lead Manganese Nickel Selenium 4 Sequential homogenizations, SS head 42.12 0.02 0.89 15.0 56.94 1.38 2.3 I 3.57 I .34 homogenizations, RD 2 1.52 0.03 0.20 0.70 13.99 0.28 0.39 0.1 1 0.68 homogenizations, RD+EDTA 23.10 0.03 0.35 1.16 14.59 0.23 0.43 0. I0 0.67 solvent blank 0.42 1l.d. "- 0.07 0.3 1 0.26 0.08 0.05 n.d.1 n.d.: 4 Sequential 4 Sequential EDTA-fortified SS = stainless steel; RD = ruby disc.1 None detected above the background in the unhomogenized solvent blank.1422 Ancllyst, October 1996, Vol. 121 tently higher (mean 14.4%) than estimates determined by conventional acid digestion ICP-MS. However, the [Cr] in these tissues was approximately half of the contamination concentra- tion from the homogenization process. The difference between the experimental and certified concentrations of Mn in these homogenates was appreciably less (mean -6.S%), reflecting the fact that [Mn] in these tissues was appreciably greater than the 0.39 pg ml-I introduced by the homogenizer. The slurry preparation technique was then applied to four botanical CRMs and to five dried animal feeds. The processed tluids were then analysed for copper, iron and manganese since these metals are monitored routinely in animal feeds. In contrast to the uniform homogenates of zoological materials, the botanical materials resulted in suspensions containing some residues of fibres and particulate matter.Two other solubilizing procedures, prior to homogenization, did not change the characteristics of the homogenate appreciably and were not investigated in detail. In the first pre-homogenization treatment, diied plant tissue, suspended in TRIS buffer (pH S.O), was extracted, at 60-65 OC, with hexadecyltrimethylammoniuni bromide and EDTA for 30 min. Alternatively, dried material was incubated at 37 "C with 1% cellulase and 3% tnacerase- pectinase at pH 5.0 for 2 h. During the subsequent metal determinations by ETAAS, no attempt was made to resuspend solids; rather, only the supernatant fraction was sampled. The difference [(accepted - experimental)/accepted] between the experimental results and the certified concentrations of the CRMs or the FAAS results for the feed samples again was acceptably small.For [Cu], the mean difference between the experimental results and accepted values was 4.4 f I I .6% and for [Mn] it was - 1.7 k 1 1 5%. In contrast. the mean difference for [Fe} was somewhat greater, -11.6% k 8.7%, suggesting that all the analyte Fe might not have been extracted into the liquid phase. To test this hypothesis, a fresh aliquot (approximately 0 . 1 g) of each sample was homogenized in the presence of SO mg of' Na2EDTA. The resulting determinations of [Fe] in homogenates prepared in the presence of Na2EDTA are reported in Table 4.For [Fe] in homogenatec prepared with Na2EDTA-fortified solvent mix- ture, the mean difference was decreased to -6.0 ? 9.7%. Despite the apparent slight improvement (decreased difference) in estimates of [Fe] in homogenates prepared with Na*EDTA, the larger absolute standard deviations ( 1 0%) associated with these means precludes a conclusion that this additive had any significant influence on the extraction of iron into the homogeniLing solvent. To determine whether Na2EDTA had any influence on levels of contaminants tnobiljzed during the slurry preparation sequence, solvent blank fortified with Na2EDTA was homogenized then analysed for metal analytes. The results are given in Table 2. The concentrations of Al, Cr and Cu were increased appreciably whereas those of Cd, Fe, Mn, Pb and Ni were not changed appreciably when compared with EDTA-fortified solvent blank plus solvent blank homoge- nized in the presence of the ruby disc.The apparent stability of crude homogenates with respect to metal analytes was also monitored with time. The supernatant fractions of homogeriates were reanalysed for Cu, Fe and Mn after 1 0 days of contact, at room temperature in sealed Table 3 Chromium, manganese. nickel. and iron concentrations (pg g- I f I standard error of estimate.' expressed as a percenlage) i n zoological certified reference materials (CRMs) and cervine liver and kidney with calibration by standard additions. Reported concentrations have been corrected for analyte concentrat ions (Table 2) introduced by the ruby-capped homogenizer C h roni 1 u in Iron M angancse Nickel Matrix Experimental Reference Experimental Reference Experimental Referencet Experimental Reference DORM-I DORM-2 Bovine muscle Oy5ter ti\sue Hcpatopancrea4 B4-liverk B4-kidney B I 0-liver 1 B 1 &kidney I 3.50 f 0.37 33.53 f 3.75 0.081 f 2 2 1.24f2.7 3.0 f 0.5 0.126f 1.8 0.1 I4 f 0.6 0.095 f 2.7 0.1 1 f l x .l 3.6-t 11.1 34.7 f 15.9 0.07 I f 54 I .43 f 32.2 2.4 k 25.0 0.10 0.1 I 0.08 0.10 59.0 f 4.9 63.6 -I 8.3 I . I4 f 5.97 130.7 f 3.5 142k7.0 3.26 f 4.34 63.1 f 5 . 1 71.2f 12.9 0.32f4.58 540.2 f 5.8 539.1 f 2.8 I 1.78 f 5.1 172.5 f 6.6 186.0 f 5.9 24.0 f 0.53 4.1 I f0.84 1 .57 f 3.37 2.30 f 2.78 3.39 f 0.29 1.32 f 19.7 I .33 f 8.80 1.20 f 25.0 3.66 f 9.29 17.2 f 9.8 19.4f 16.0 0.37 k 24.3 n.d.0.05 f 80.0 12.3 f 12.2 2.19 f 10.2 2.25 f 19.6 23.4 f 4.27 2.29 f 4.89 2.3 f 3.01 4.40 1.62 2.73 3.4 I * Standard crror of estimate: SE,,, = (SE, + SE,lopd2)1/2. pg g-' (f I RSD, expressed as a percentage). 1 Cervine tissue. Table 4 Copper. iron and manganese concentrations (pg g- I k 1 standard error of estimate, expressed as ;I percentage) in botanical certified reference materials and in animal feeds as determined in the supernatant fraction of the slurry suspension immediately after or 10 d after preparation. Reported concentrations have been corrected for analyte concentrations (Table 2) introduced by the homogenization process Copper Iron Manganese Matrix Pine needles Corn \talk Apple leave5 Corn bran 303-5 299-5 158-5 3 14-5 307-5 Exptl.' Expt1.I Reference j 3.47 f 9.1 3.2 1 f 8.I 3.0 f 10.0 6.79 f 4.5 7.25 f X.2 X.0 k 12.5 5.99 f 6.3 5. I4 f 6.2 5.64 f 4.3 2.78 f 8.2 2.19 f 7.3 2.47 f 16.2 3 . 3 f X . 4 3.2 f 3 . 3 3.5 19.5 k 2.6 18.0 +_ 8.2 17.5 2 1 .0 -t 2.4 20.0 f 5.6 18.5 15.7f3.5 15.9f3.4 14.1 79.4f 14.4 84.2 f 7.2 88.4 Exptl. 170.2 f 10.5 113.8 f 4 . 3 78.94 f 6.4 I1.29f9.7 68.62 f 5.9 364.7 f 8.8 14.66 f 9.3 56.05 f 6.5 92 I .3 k 3.6 Exptl. Exptl.4 176.0f 13.0 184.3k5.2 I11.4f2.4 124.1 k3.6 75.9 f 4.3 85.9 f x.O 12.1 f 8 . 3 13.4f9.0 65.2 f 6. I 69.4 f 8.4 379.1 f 12.0 371.5 f 10.1 11.Sf4.2 14.2f 11.0 53.9 f 7.2 57.0 f 5.3 942.4 f 14.5 925.1 f 8.1 Reference r 200 k 5.0 139 +_ 10.8 83 -t 6.4 14.8 f 12.2 78.1 425 12.2 64.7 1027 Exptl . % 648.0 f 1.3 15.6 f 2.8 5 1 .8 f 2.2 2.12 f 5.0 14.05 k 5.6 51.3 k 5.5 11 1.5 f 3.8 40.6f 4.7 204.5 f 4.6 Exptl.i- 62 1.4 k 8.7 15.9f 13.3 50.9 f 4.0 2.1 1 f6.2 14.2 k 6.2 49.8 -t 12.0 117.0f9.2 34.25 f 7.1 201.2f7.6 Reference4 675 k 2.2 15.0f2.5 54 f 5.6 2.6 f 1 1.4 13.9 66.3 120.0 37.6 220. I Supernatant fraction from the slurry suspension was analysed immediately after preparation. Supernatant fraction from the slurry suspension was sampled I 0 d after preparation. -1 For CRMs, pg g-l f 1 KSD (expressed as a percentage); feed samples were analysed by FAAS following digestion in HNO.+ 9 Samples were slurried in 20 nil of ethanol-water (1 + 9 v/v) containing 0.25% m/m of TMAH and 50 mg of Na,EDTA.Analyst, October 1996, Vol. 121 1423 Table 5 Means of slopes* (k 1 RSD) of the best-fit linear regression models for standard additions of Cu, Fe or Mn to solvent blank or to slurries of zoological or botanical CRMs or of animal feeds Zoological Botanical Animal feeds Solvent blank’ Mean (blanks + Analyte CRMS (n = 5 ) CRMs ( n = 4) (17 = 5 ) (n = 3) feeds + CRMs) Cu (244.2 nm) 2.60 k 1.3% 3.22 f 3.8% 3.05 rfr 1.3% 3.06 k 0.3% 3.11 f 3.4% Fe (386.0 nm) 6.74 +_ 2.5% 6.77 f 2.7% 6.82 f 1 .1 % 6.87 f 1 .0% 6.75 f 1.9% Mn (403.1 nm) 8.38 f 2.2% 8.57 * 4.6% 8.46 f 3.2% Mn (279.5 nm) 1.28 f 1.2% 1.30 f I .O% 1.30 f 2.3% 1.30 f 1.4% (I? = 12) (12 = 17) ( n = 8) ( n = 12) * Slopes were determined from linear regression of the mean peak area (absorbance s) for three replicate determinations of each of four or more additions of aqueous standard to the homogenate or to the solvent.-1 Aqueous analyte standard added to solvent blank or to solvent blank homogenir.ed in the presence or absence of the ruby disc. s: Mean of slopes for botanical CRMs + feeds + solvent blanks. [ Cul in zoological CRMS has been determined about 45 d previously and were not included in the calculation. containers, with the particulate fraction. The results are given in Table 4. The differences for [Cu], [Fe] and [Mn] determinations in the homogenates after 10 d were -2.4 * 9.1% (\wsus +4.4 k 11.6% at day 0), -1 1.1 k 8.6 (~w.sus -1 1.6 f 8.7% at day 0) and -7.7 f 9.6% (\,el-sus - 1.7 k 1 1 .S% at day 0), respectively. Although the individual estimates changed slightly with time, there was no appreciable evidence for any consistent change in analyte concentrations in the supernatant fractions.These results indicate that homogenates can be stored, at room temperature, for up to 10 d without any Cu, Fe or Mn segregation. Determinations of selenium in the homogenates were also attempted. However, for both zoological and botanical CRMs the estimated [Se] was appreciably less than the certified values even for calibrations by standard additions. For wheat-based feed samples (containing approximately 10 pg g-1 of Se), previous studies7’ had suggested that virtually all of the Se was protein bound and that neither selenate, selenite, selenomethio- nine nor selenocystine was present in detectable amounts. Presumably, protein-bound Se in the botanical CRMs is not liberated efficiently by the ashing cycle of the electrothermal programme.Although the analyte determinations reported in Tables 2-4 were calibrated by standard additions, there were no significant differences among the slopes of the calibration plots for the regressions of peak area on the amount of analyte standard added back into each of the matrices (Table 5). Moreover, the slopes of the calibration plots for standard additions to solvent blank that had not been homogenized or had been homogenized in the presence or absence of the ruby cap were not significantly different from each other. Hence, there was no evidence for any matrix effect in any of the feed samples, the cervine tissues or the reference materials. Calibrations performed with standards added to solvent blank were used to determine the levels of contamination due to processing.These values proved to be repeatable, indicating that the external aqueous standards could have been used equally well for calibration. Conclusions The results indicate that high-pressure homogenization is capable of generating emulsions/dispersions of soft organ tissues, dried animal feeds and zoological and botanical CRMs that can be reliably sub-sampled during 10 d of storage (botanical CRMs, feeds). The elements that can be determined in these slurries include Cu, Cr, Mn, Ni and Fe. In contrast, estimates of [Se] in biological CRMs were consistently lower than their certified values, suggesting that the release of this analyte from the protein matrix during the pyrolysis and ashing sequence was inefficient. The addition of EDTA to the solvent prior to processing did not perceptibly improve the recoveries of analyte metals from either the botanical CRMs or the feeds but did increase the levels of contamination.The presence of the ruby cap on the valve head of the homogeniLer appreciably attenuated but did not eliminate contamination introduced by the processing. Sample preparation proved to be rapid (approx- imately 3 min) and the homogenizer was readily cleaned between samples by processing fresh solvent. In total, the technique presents a rapid means of sample preparation but, at present, can only be applied to samples that contain analyte levels in excess of the levels of contaminants introduced by processing . Feed samples and analyses of their contents of Cu, Fe and Mn, as determined by FAAS, were generously provided by E.R. Chavez, Department of Animal Science, McCill University. Financial support from the Natural Science and Engineering Kesearch Council (NSERC) of Canada is gratefully acknow- ledged. References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 I S 19 20 Novozamski. I., van der Lee, H. J., and Houba. V. J. G., Micn,r.him. Acta, 1995, 119, 1x3. Sansoni, B., Panday, V. K., in Analytical Techniques ,for Heuvy Metals in Biological Fluids, cd. Fachetti, S., Elsevier. Amsterdam, 1983, pp. 91-131. Mincwewski, J., Chwastowska, J., and Dybczynski. R., Scpat-ation and Prec,onc,cntration Methods in Inoi.ganic Trace Analysis, Ellis Horwood. Chichcster, 1982. Matusiewicz, J., and Sturgeon, R. E., PI-og. Anal. Spcc.tro.sc., 1989, 12,21.Reid, H. J., Greenfield, S., arid Edmonds T.E.. Analyst, 1995, 120, 1543. Langmyhr, F. J., Analyst, 1979: 104, 993. Langmyhr, F. J., Prog. Anal. At. Sper.trosc., 1985, 8, 193. Miller-Ihni, N. J., Ai?ul. Chem., 1992, 64, 965A. de Benzo, Z. A., Velosa, M., Ceccarelli, C., de la Guardia, M., and Salvador, A., Frcseniirs’ .I. Anal. Cl~enz., 199 1, 339, 235. Bendicho. C., and de Loos-Vollebregt, M. T. C., .I. AtzaI. At. Specmmi., 1991, 6, 353. Miller-lhli, N. J., Fresenius’ .I. Anal. Chern., 1993 345, 482. Stephen, S. C., Littlejohn, D., and Ottaway, J. M.. Analyst, 1985, 110, 1147. Thompson, D. D.. and Allen, K. J., At. Spectrosc., 1981, 2, 53. Madrid, Y., Bonilla, M., and Camara, C., .I. Anul. At. Sp~~tiwiz., 1989, 4, 167. L6pez Garcia, I., Ortiz Sobejano, I;.. and HernBndez Cordoba, M., Analyst, 1991, 116, 517. Hoenig, M., and Hoeyweghen, P. V., Anul. Chrm., 1986, 58, 2614. Albers, D., and Sacks, R., Anrrl. Cheni., 1987, 59, 593. Tsalev, D. L., Slaveykova. V. I., and Mandjunkov, P. B., Spec- trociiini. Acta, Rev., 1990, 13, 225. Miller-Ihli, N. J., J. A m l . At. Spectl-om., 1989, 4, 295. Bendicho, C., and de Loos-Vollebregt, M. T. C., Spectr-ochinz. Acta, Part B, 1990, 45, 679.1424 Aiiulyst, Octohei- 1996, Vol. 121 21 22 23 24 25 26 27 Miller-Ihli, N. J., .I. Anal. At. Spectt-mi., 1988, 3. 73. Lynch, S., and Litllejohn, D., .1. Anul. Ar. Spectrum., 1989, 4, 157. Fagioli, F., Landi. S.. Locatelli, C., Kighini, F., Settimo, K., arid Magarini, K.. ./. A w l . At. Spc~-tron?., 1990, 5, 5 19. Hansen, D. L., and Bush, E. T., Anul. Rior.hem., 1967, 18, 320. Jackson, A. J., Michael, L. M., and Schumacher, H. J.. Anul. Chern., 1972, 44. 1064. Murthy, L.. Menden, E. E., Eller, P. M., and Petering, H. G., A ~ u l . Riochctii., 1973, 53, 365. Uchida, T.. Isoyanna, H., Yamada, K., Oguchi, K.. Nakagawa, G., Sugie. H., and lida. C., Atrul. Chin?. Act(7, 1992. 256, 277-284. 28 29 30 31 Lei, T.. and Mar\hall, W. D., Appl Orgunonlet Cheni.. 1995, 9, Dion, B., Ruzbie, M., van de Voort, F. K., lsmail, A. A.. and Blais, J. S., Anulvst, 1994, 119. 1765. Tan, Y., Marshall, W. D., and Blais, J-S., Anu/vs/, 1996, 121, 48.1. Millen-Ihli, N. J., At Sprctio\c . 1992. 13, I . 149.
ISSN:0003-2654
DOI:10.1039/AN9962101419
出版商:RSC
年代:1996
数据来源: RSC
|
|