摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 199 1 VOL. 6 653 Direct Determination of Cadmium and Lead in Geological and Plant Materials by Electrothermal Atomic Absorption Spectrometry Franci DolinQek Janet Stupar and Vinko Vrlieaj Jozef Stefan Institute University of Ljubljana 61 11 1 Ljubljana Yugoslavia A laboratory-assembled atomic absorption spectrometer and graphite cup atomizer were evaluated for direct analysis of solid samples. A number of geological (soil and sediment) and botanical reference materials were analysed for lead and cadmium either by direct introduction (0.02-5 mg) or by a slurry technique (5-200 mg). Integrated absorbance was measured and aqueous standards were used for calibration. An (NH4)2HP04 chemical modifier was necessary in the analysis of botanical samples.The effect of particle size on the accuracy and precision of analytical results was studied. Atomization of small sample aliquots (20-500 pg) by direct ‘weighing in’ sample introduction technique requires fine grinding (1 pm) whereas slurry sample introduction is more tolerant to particle size effects. Keywords Cadmium; lead; geological and botanical reference materials; solid sample introduction; electrother- mal atomic absorption spectrometry Direct determination of trace and minor elements in solid samples of geological and biological origin has long been a challenge to analytical chemists employing atomic spectro- metric methods of analysis. Several advantages of these techniques have been pointed out in the Among the most important are the ability to analyse very small samples and to reduce sample contamination in comparison with conventional solution analysis.A variety of technical approaches have been investigated for the introduction of solid samples into flames plasmas and graphite furnaces for atomization. The most convenient and easiest way of achieving this was to prepare a slurry from the powdered sample. Nebulizing slurries for intro- duction into a flame or plasma however proved to be inefficient producing severe systematic errors in the analy- tical result^.^^^ On the other hand electrothermal atomiza- tion owing to its specific features proved to be the most suitable for direct analysis of solids. It has been extensively employed for the determination of elements having low and medium volatility in a variety of Cadmium and lead have been the most frequently determined elements because of their environmental importance.’ 1-14~19-22924+30933 Solid samples can be introduced into a graphite furnace by a direct ‘weighing in’ technique or by pipetting slurries.Although the latter is preferred there are two major problems involved in these sampling techniques which affect the accuracy and precision of analytical results. The first is sampling error the difference between the content of the analyte of the measured fraction and that of the bulk of the sample which depends on the sample homogeneity and which is further related to the mean particle size and distribution. The sampling error is likely to be critical when using the direct ‘weighing in’ approach where extremely small amounts (20-500 pg) of geological or botanical samples are analysed directly.When employing the slurry sample introduction technique this sampling error can be largely eliminated by generally taking a sample amount of 5-200 mg. The only prerequisite in such a situation is that the slurry should be homogeneous during pipetting. This can be accomplished either by adding a thickening agent or by employing mechanical or ultrasonic agitation. Majidi and H ~ l c o m b e ~ ~ ~ ~ * brought attention to sedimentation and to the volumetric errors both of which might have a significant influence on the accuracy of analytical results. The second major source of error could result from incomplete vaporization of the analyte element in the furnace.Matrices of low volatility and refractory elements are more susceptible to this source of error. As the vaporization of the analyte in the furnace might in some way be related to the particle size of the sample particle size can be assumed to be an important parameter govern- ing the precision and accuracy of analytical results. This has been realized by several workers who have investigated the effects of particle size and distribution on the results obtained by direct analysis of solids using eletrothermal atomic absorption spectrometry (ETAAS). Fuller et aZ.,23 carrying out direct determination of Cr and V in ore samples by slurry ETAAS concluded that acceptable precision and accuracy could be obtained pro- vided the samples are ground to at least a 25 pm size and aqueous standards are used for calibration.It was further shown that sampling error becomes significant when part- icle size increased above this value. Jackson and Newmanls and Hinds et aZ.16 have measured lead and cadmium in different types of soil samples by direct introduction of slurried samples into the electrother- mal atomizer. It has been shown that significantly more lead is contained in smaller size fractions of the soil in comparison with larger particles and fine grinding (9OOh of particles of less than 11 pm) was found to be essential for complete recovery of these elements from soild samples. However the major source of error was ascribed to incomplete atomization rather than sampling error. The efficiency and reproducibility of pipetting the slurries were affected as the particle size increased beyond 50 pm.Hinds and Jackson1* and Karwowska and Jackson17 have also studied the atomization characteristics of lead from alu- mina matrices and artificial soil samples. It was reported that alumina and clay minerals have the same effect of retarding the vaporization of lead from the soil sample in a comparison with aqueous solutions. Organic matter in the soil however has the opposite effect of shifting the absorbance peak of lead to a shorter appearance time. The use of chemical modifiers and isothermal atomization were strongly suggested for use with direct solid analysis. In addition by careful examination of the absorbance-time profiles the possibility of determing lead fractionation in the soil samples was shown.Olayinka et a1.22 recommended slurry sampling ETAAS for the determination of cadmium in foodstuff samples. The samples were ground to a 44 pm particle size and 0.04-0.2°/6 m/v slurries were prepared by magnetic stirring. The peak height absorbance was mea- sured and calibration was accomplished by a standard654 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 199 1 VOL. 6 additions technique. Vegetable plant and botanical samples were successfully analysed for cadmium lead copper and other elements by slurry sampling ETAAS. Stephen et af.,20 employing a thickening agent in order to stabilize the slurry found that the precision of measure- ment was not significantly affected by reducing the bulk of the particle size of the powdered samples from 54 to 2 1 pm.Similarly no matrix interference was associated with the change in particle size distribution for slurries of liver samples. Slurries containing 0.08- 10 g of sample per 100 ml were measured with an acceptable precision i.e. 4-8%. Ebdon and Lechotyckiz1 recommended platform atomiza- tion and the use of an (NH4)H2P04 chemical modifier to eliminate matrix interference in the determination of cadmium. Slurries were prepared from amounts of sample between 50 and 500 mg which had been ground to a particle size of less than 20 pm (90% of all particles) in a volume of 25 ml employing magnetic stirring for stabilization. Excellent agreement was obtained between direct analysis and conventional solution atomization employing inte- grated absorbance measurements and aqueous standards for calibration.In contrast to these workers Carri6n et af.24 used 1 mg ml-l slurries of National Institute of Standards and Technology (NIST) Standard Reference Material (SRM) 1575 Pine Needles with much coarser particles (less than 160 pm) and found good correlation with wet digestion solution analysis. A slurry sample of NIST SRM 1 575 Pine Needles prepared by ultrasonic homogenization served for calibration. Miller-Ihli25~29 carried out an extensive investi- gation of slurry preparation for direct multi-element analy- sis of a variety of biological and botanical RM samples. Slurries were prepared using 5-10 mg of sample in 5 ml of 5% HNOJ containing 0.04% Triton X-100. Relatively coarse particles lying in a narrow range (250-600 pm) yielded better precision than particles typically found in NIST SRMs ( t 2 5 0 or t 4 2 5 pm) which is in contrast to the observations of previous workers.15*19 Ultrasonic probe and vortex mixing have been critically compared. Although the precision of the measurements was found to be comparable the former technique was pre- ferred particularly for elements and samples in which settling of the heavy particles in the slurry might produce a systematic error in the analytical results. This paper deals with the evaluation of the performance of a laboratory- assembled atomic absorption spectrometer and a graphite cup atomizer designed specifically for direct analysis of solid samples. Special emphasis is given to the importance of the size of the sample particles and the distribution in affecting the accuracy and precision of the determination of lead and cadmium in a variety of NIST SRMs.The direct weighing in technique and a slurry sample introduction technique are critically compared. Experimental Apparatus A laboratory-assembled atomic absorption spectrometer and graphite furnace were employed in this work to achieve so-called stabilized temperature platform furnace condi- tions as suggested by Slavin et uf.26 The spectrometer was constructed from partly-modified commercial units and laboratory-made electronic parts. A schematic diagram of the instrument is presented in Fig. 1. The element hollow cathode lamp (1) and deuterium lamp (1) are positioned at right-angles to each other and the emission of both lamps passes through a semitransparent quartz mirror (2) in the same optical path.The lamps operate successively at a frequency of 1 kHz (see Fig. 2) and each lamp is switched on for 0.33 ms every 1 ms. At time intervals when neither lamp is activated the graphite furnace emission signal was measured and subtracted from the lamp signal. The high 1 2 J 11 12 13 . 18 - 17 16 15 :+ 1 1 %EL Y02H J Fig. 1 Block diagram of the instrument. I Hollow cathode lamps (element and deuterium); 2 semi-transparent quartz mirror; 3 graphite cup atomizer; 4 monochromator SPM-2 (Carl Zeiss); 5 photomultiplier tube R- 106 (Hamamatsu); 6 photomultiplier tube power supply; 7 modified power unit CRA-90 (Varian); 8 power supply feed-back system; 9 temperature measurement sensor; 10 hollow cathode lamp power supply; 1 1 frequency standard (clock); 12 chopper; 13 furnace emission eliminator; 14 electrical filters; 15 zero adjustment unit; 16 logarithmic amplifer; 17 system for signal subtraction; 18 integrator; 19 dual-channel storage oscillo- scope 564 B (Tektronix); and 20 recorder.EL ZH 0.33 ms 0.33 ms m w w H 1 ms ~ I I J I ' Time - Fig. 2 Lamps operation characteristics. EL element hollow cathode lamp and 2H deuterium lamp frequency signal obtained is split by means of precise timers into two separate signals. The first represents the element hollow cathode signal and the second the deuterium lamp signal. By using the two-channel storage oscilloscope ( 19) both signals can be seen at the same time and overlapping of atomic and background signals can easily be observed in transmission mode [see Fig.3(a)]. The transmission signals are converted into absorbance by means of a logarithmic amplifier (16). The deuterium lamp signal is subtracted from the element hollow cathode lamp signal and the difference represents the atomic absorbance signal. A laboratory-made integrator (1 8) is built into the electronic system and enables the absorbance signal to be integrated. Thus absorbance and integrated absorbance signals can be observed at the same time [see Fig. 3(b)]. The time constant of the instrument can be varied in the range 1.1-1 200 ms enabling fast electronic signals to be followed. The photo- multiplier tube ( 5 ) and lamps (1) are powered from locally- made sources (6) and (7) and run under the recommended conditions.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 199 1 VOL.6 655 1 1 TvDe B Type c .v- - I 1 I 0 10 20 Fig. 5 Graphite cups in cross-section I I I 0 5 10 Ti me/s Fig. 3 (a) Atomization of the lead in the presence of NaCl matrix. 1 Lead signal; and 2 background signal. (b) Absorbance and integrated absorbance signals. 1 Peak height absorbance; and 2 integrated absorbance 12 Fig. 4 Graphite cup atomizer. 1 Beam collimator; 2 quartz cover; 3 graphite cup; 4 graphite electrode; 5 quartz chimney; 6 watercooled brass block; 7 assembly screw; 8 power supply connection; 9 aluminium base plate; 1 0 poly(tetrafluoroethy1ene) electrical insulator; 1 1 holder and 12 furnace tension adjustment Graphite cup atomizer The graphite atomizer used was laboratory-made and is similar to a mini-Massmann as produced by Varian Techtron.It is shown in schematic form in Fig. 4. The graphite cup (3) is fixed by means of two support electrodes (4) which are clamped between two large water-cooled brass blocks (6) that serve as electrical terminals (8). The distance between the blocks can be varied by means of a screw (12) which enables the graphite cups to be ex- changed. Spring-loaded support electrodes ensure good and constant electrical contact. The chimney ( 5 ) and its cover (2) are made of quartz. Argon flowing through the chimney is used as the protective gas in order to minimize the oxidation of the carbon. The reconstructed Varian CRA-90 unit serves as the power supply to the graphite cup. The atomization cycle was extended up to 20 s the heating rate was made more rapid (maximum heating rate is more than 2000 K s-l depending on the graphite mass) and a constant temperature was achieved for the graphite cup through a built-in temperature feed-back system equipped with fibre-optic cable.Positioned near the graphite cup the fibre-optic cable is fixed in a small stainless-steel tube and connected to a phototransistor. The whole system faces the incandescent graphite cup wall. Another phototransistor fixed just below the stainless-steel tube serves for tempera- ture control of the graphite cup and the signal can be observed by using the storage oscilloscope. The atomization cycle is also partly reconstructed; a cooling step (of 12 s duration) can be applied prior to final atomization.The known temperature overshoot phenomenon normally ob- served when rapid heating rates are used was also largely eliminated. Graphite cups The graphite cups used in this work consisted of two parts an outer cup of 8 mm 0. d. 1 mm wall thickness and 9-12 mm in height; and a small inner cup of 5.5 mm 0. d. 0.5 mm wall thickness and 5-8 mm in height. The latter can either be placed on the bottom of the outer cup (type A Fig. 5 ) or hung on the edge of the hole made through the bottom of the outer cup (type B and type C Fig. 5). The inner cups are only heated by thermal conductivity and radiation from the outer cup and hence act as a platform. Hanging cups are particularly suitable for larger sample amounts and for solid samples that form carbonaceous residues or increase in volume during atomization (blood etc.).All inner cups with the exception of type C are easily inserted and removed from the outer cup by means of specially designed tweezers. Inner cup type C is similar to type B and has a graphite ring placed in the lower part which is fixed by self- friction (a push-fit) hence a much slower heating rate is obtained. The heating rate can also be varied by the height of the inner cup and thus a number of different cups similar to type B can be made for any particular analytical application. Up to 400 measurements can be carried out using a single graphite cup for the determination of cadmium and lead. Graphite cup temperature. The outer cup temperature control was performed by means of a phototransistor connected via an electronic circuit to the storage oscillos- cope or to the plotter.The corresponding voltages obtained were calibrated for temperature (in K) using an optical pyrometer (Pyrowerke). The pyrometer was focused at the inner wall surface of the outer cup. Slight variations of this system caused by unstable electrical contacts between the656 3000 4 I JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 1991 VOL. 4 ) Fig. 6 Heating characteristics of the outer cup and different inner cups. 1 Cup type A; 2 cup type B; 3 cup type C; and 4 outer cup 1 I J C J 0 2.5 5 2.5 5 1 . 0 L l Q ;;OECl; n & 0.5 1500 K 0 2 0 0 .o Fj 1500 :Fi 0.4 0.5 0.2 0 0 5 10 0 5 10 Time/s Fig. 7 Absorption profiles of cadmium in different graphite cups (a) and (b) type A; (c) and (d) type B (5 mm); and (e) and 0 type B (8 mm).Variation of absorbance and integrated absorbance with the heating rate. Sample 5 pl 0.025 pg ~ m - ~ . T=temperature profile support electrodes and cup can occur and periodic calibra- tion is recommended. As the rate of entry of the sample into the analytical volume of the cup depends on the heating rate of the inner cup the temperature-time functions for the inner cups presented in Fig. 6 were measured at certain selected outer cup temperatures. It can be seen clearly from this figure that the temperature-time functions of the inner cups investigated are phase delayed which is a character- istic feature of a platform. Moreover? at equilibrium? only the cup type A temperature reaches the outer cup tempera- ture whereas the equilibrium temperatures of cup types B and C are much lower (the temperature difference being approximately 600 K for these two types). Graphite cup characteristics.In order to investigate the performance of the graphite cup atomizer? cadmium and lead solutions were atomized using graphite cups of types A and B ( 5 and 8 mm in height). Aqueous solutions were used instead of solid samples so as to achieve better reproducibil- ity of results. Measurements were performed at a constant outer cup temperature while the heating rate was varied. Results presented in Figs. 7 and 8 quantitatively illustrate the extent of the delay in appearance and broadening of the I I I I I 0 2.5 5 0 2.5 5 s 0 5 10 1900 K e 0.5 0.5 0 % - 0 5 10 j C 1 .o El LoL]- 1900 K * 0.5 0.5 0 0 0 5 10 Time/$ 0 5 10 Fig.8 Absorption profiles of lead in different graphite cups (a) and (b) type A; (c) and (d) type B (5 mm); and (e) and u> type B (8 mm). Variation of absorbance and integrated absorbance with the heating rate. Sample 5 pl 1 pg ~ m - ~ . T=temperature profile 0 10 Time/s 20 Fig. 9 Absorption profiles of cadmium in various graphite cups. Sample NBS SRM 1645 River Sediment; 0.02 mg. 1 Cup type A wall atomization (without inner cup); 2 cup type B; 3 cup type C; and 4 outside cup temperature 2100 K absorbance-time profiles with decreasing heating rates. All of these phenomena were even more pronounced when cup type C was used for atomization. Despite these differences in peak shapes the integrated absorbances were almost constant when the same masses of cadmium or lead were atomized.The absorption peaks always appeared in the interval where the outside cup temperature was constant. Varying the heating rate of the outer cup and/or changing the type of inner cup influenced the rate of entry of cadmium or lead atoms into the analytical volume signifi- cantly? and hence extended the useful range for the determination of these elements over two orders of magni- tude without changing the analytical line. As an example determination of cadmium in National Bureau of Stan- dards (NBS now NIST) SRM 1645 River Sediment containing 10.2 pgg-l of cadmium is shown in Fig. 9. A 0.02 mg portion of the sample was atomized using an inner cup of type C without obtaining an excessively high peak absorbance value.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 199 1 VOL.6 657 Procedures Reagents and standard solutions Doubly distilled de-ionized water was used for preparing the slurries and for diluting stock standard solutions ( 1000 mg dm-3). Standard stock solutions were prepared by dissolving the appropriate salts or metals (all of analytical- reagent grade). An aqueous solution of (NH4)2HP04 (25 g dm-3) was used for chemical modification. Solid sample and slurry preparation Several powdered solid samples obtained from NIST and from the International Atomic Energy Agency (IAEA Vienna) were used in this work. Prior to analysis all of the samples were dried for 4 h at 80 "C. A part of each sample was ground in a vibrational mill (Retsch Model MM-2) equipped with either a stainless-steel or zirconia grinding jar and balls.In contrast to previous work,29 no significant lead or cadmium contamination was found after grinding silica. The samples were stored in polyethylene containers. After each grinding operation the device was carefully cleaned in order to prevent contamination of the following sample. Slurries were prepared by diluting 5-200 mg of powdered sample in 1-25 ml of doubly distilled water containing 0.04% Triton X- 100. If necessary a correspond- ing amount of (NHJ2HP04 was added directly to the slurry. Analytical procedure The samples were weighed either directly into the inner graphite cup or into special glass containers using Mettler Model AE- 1 63 and UM-3 microbalances. When weighed directly the inner cups were easily inserted into the outer cup and removed after completion of the analysis.Samples from the containers were introduced into the inner graphite cups using a small glass funnel. By using this sample introduction technique with samples weighing less than 30 pg the results showed that 98-102% of the sample entered the cup. Typically 0.02-5 mg of sample were taken for a single measurement. Slurries were placed directly into the inner graphite cups using a micropipette (1 0-20 pl). Prior to sampling the slurry was homogenized using an ultrasonic device with a 3 mm diameter titanium probe (Cole-Par- mer ultrasonic homogenizer Model 47 10). The instrumen- tal parameters employed in the determination of cadmium and lead are given in Table 1. The analytical signals were observed using the dual-channel storage oscilloscope (Tek- tronix Model 564B) and integrated absorbance was mea- sured.The mean (3 and standard deviation (a) were calculated from at least 11 independent measurements of each sample using either the weighing in technique or the slurry introduction procedure. Results and Discussion Particle Size Distribution Measurements In direct analysis of solids by ETAAS the particle size distribution of the sample has been shown to be an important parameter influencing the quality of the re- sults. 15~16s20f3*27*29 The accuracy and precision of measure- ments can be affected either by inadequate homogeneity or poor sample representation and/or incomplete recovery of the analyte from the matrix during the atomization process.An investigation was therefore conducted to assess the magnitude of these effects in the determination of cadmium and lead in a variety of samples employing both the direct weighing in and the slurry sample introduction techniques. For that purpose original and ground RMs were taken and particle size distribution measurements performed on both Table 1 Instrument settings and graphite cup parameters used in the determination of cadmium and lead in aqueous solutions and different solid materials Parameter Cd Pb Wavelengthhm Spectral band pass/nm Lamp current/mA Graphite cup type Argon flow rate/ml min-' Drying- TemperatureK Time/s Ashing- TernperatureK Time/s Cooling- TemperatureK Time/s TemperatureK Time/s Lamp current/mA Atomization- Hydrogen background corrector 228.8 0.6 5 A,B,C 3000 380-420 10-90 600-900 5-60 295 10 1400-2300 2-10 15-25 * Modifier added 125 pg of (NH4)2HP04.283.3 0.5 6 A,B,C 3000 380-420 10-90 600- 1 OOO* 5-60 295 10 1600-2300 2-12 15-25 sets of samples using a low-power helium-neon laser diffractional type granulometer Model HR 850 Cilas Alcatel. Some characteristic values such as the median (&) 90% diameter (Ow) and 10% diameter (Dlo) were taken from the measured distribution graphs and are summarized in Table 2. It is evident from these results that some of the original RMs in particular those of plant tissue were relatively coarse having broad particle size distribution. It should however be borne in mind that these materials were made for conventional analytical procedures where a minimum of 100-250 mg of sample are required for the analysis.In general additional grinding substantially reduced the mean particle size of all of these materials with the exception of the IAEA Soil 7. (This soil already existed in fine powder with a median diameter of 9 pm.) In addition to this the distribution of particle sizes obtained after the grinding stage was much narrower. Analysis of Solids by Direct Weighing In Sample Introduc- tion Lead and cadmium were determined by the introduction of weighed amounts of solid samples directly into the graphite Table 2 Particle size distribution data of original and ground reference material samples Sample Diameterlpm NBS 1633a IAEA Soil 7 NBS 1645 NBS 1571 NIST 1572 NBS 1573 Coal Fly Ash River Sediment Orchard Leaves Citrus Leaves Tomato Leaves Original sample Finely ground sample 14 46 2 9 31 2 13 55 1 38 189 6 78 265 7 88 350 6 2 8 0.6 7 29 0.7 4 24 1 13 38 3 13 37 3 11 40 3658 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 199 1 VOL.6 Table 3 Determination of cadmium in reference materials by ETMS using the direct weighing in method of solid sample introduction Result found (Xk o)/pg g-l Sample NBS 1633a IAEA Soil 7 NBS 1645 NBS 1571 NIST 1572 NBS 1573 Coal Fly Ash River Sediment Orchard Leaves Citrus Leaves Tomato Leaves * Value not certified. Certified value Amount of (Xf aypg 8-1 sample/mg 1 .Of 0. 15 0.6-1.3 - 0.3-1.5 10.2k 1.5 0.05-0.15 0.1 1 kO.01 1-3 0.03 f 0.0 1 2-5 3* 0.8- 1.2 Original sample Finely ground sample 0.82 k 0.2 1 1 .oo k0. 12 5.9k 1.5 6.70k0.87 0.13 k 0.04 0.73 k 0.07 0.97 k 0.12 0.09 k 0.03 0.02 1 k 0.004 0.033 & 0.007 2.0 k 0.2 2.3 k O .l Table 4 Determination of lead in reference materials by ETAAS using direct weighing in method of solid sample introduction Result found Q k a)/pg g-I Sample NBS 1633a IAEA Soil 7 NBS 1645 NBS 1571 NIST 1572 NBS 1573 Coal Fly Ash River Sediment Orchard Leaves Citrus Leaves Tomato Leaves Certified value (Xf a)/= g-1 72.4 f 0.4 55-7 1 714k28 4 5 f 3 13.3 f 2.4 6.3 f 0.3 Amount of sampleimg 0.1-0.3 0.1-0.5 0.02-0.03 0.25-0.5 0.5-1.2 1-1.5 Original sample Finely ground sample 56.2 f 2.2 69.4 f 8.3 67.3 k 4.5 63.9 k 7.0 800 f 280 806 k 201 32k5 36k2 10.3 f 2.2 12.1 k0.8 4.2 k 0.8 6.5k 1 cups. Different inner cups (see Fig. 5 ) were employed for particular samples depending on the type of sample and concentration of the analyte present.For example the analysis of NBS SRM 1645 River Sediment of relatively high lead and cadmium contents required the use of inner cup type C facilitating a slow release of analyte elements during the atomization process. Although the amount of sample used for a single measurement was in the range 20-150 pg. Analysis of biological and plant materials produced a high background coinciding with the atomic absorption peak. Overlapping of these signals was largely reduced by the use of the (NH4)2HP04 chemical modser which is illustrated in Fig. 10. Therefore the use of a chemical modifier was obligatory for the determination of cadmium and lead in these samples if accurate results are to be obtained employing aqueous standards for calibration.Foaming of some of the biological and plant samples was observed during the atomization process which might result in sample loss and poor precision of measurement. This problem was solved by directing a small stream of argon in a direction perpendicular to the bottom of the inner cup during the drying and ashing step. After each measurement the residue if any was also blown-out of the cup by means of the stream of argon. The results of the determination of cadmium and lead for some of the selected RMs by direct introduction of weighed amounts of sample into the graphite atomizer are summarized in Tables 3 and 4. Sample aliquots used for the determination of these elements were in the milligram range with the exception of NBS SRM 1645 River Sediment IAEA Soil 7 and NBS SRM 1633a Coal Fly Ash in the determination of lead where sub-milligram amounts were analysed.The original materials produced results that were pre- dominantly lower than the certified values. Additional grinding of these samples improved the accuracy of results with the exception of the soil sample where particle size distribution was not substantially changed. Plant material samples ground to a particle size of t 4 4 pm should yield acceptable results provided sample aliquots of 0.5-2 mg are atomized. The same would probably apply for soil and geochemical samples ground to particle sizes of less than 10 pm in diameter.16 Sample homogeneity can definitely be a dominant factor in the production of erroneous results when very small sample aliquots (20-500 pg) are atomized.As can be deduced from the data for NBS SRM 1633a Coal Fly Ash (see Table 2) although reducing the particle size (Ow) from 46 to 8 pm did not totally eliminate error in the result for lead. It would therefore appear that sub-micro- metre particle size is required for geological and related samples if accurate results are to be obtained by direct weighing in sample introduction. Analysis of Solids by Slurry Sample Introduction The convenience of sampling slurries in the direct analysis of powdered samples was first pointed out by Fuller et uLZ3 Sample amounts in the range 5-10 000 mg were typically ~ s e d ~ ~ ~ ~ ~ to prepare slurries which largely eliminated the problem of how well the specimen represents the sample.The homogeneity of the slurry during pipetting of aliquots of 10-50 pl for atomization was maintained by the addition of a thickening agent,*’ magnetic stirring16 or ultrasonic homogenizati~n.~~*~~-~~ The last was found to be superior particularly for slurries containing large heavy particles.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 199 1 VOL. 6 659 Table 5 Determination of cadmium in reference materials by ETAAS using slurry sample introduction Result found (Xf a)lpg g-I Sample (Xf a)/* g-1 mg ml-l sample Slurry Certified value concentration/ Original sample Finely ground NBS 1633a Coal Fly Ash IAEA Soil 7 NBS 1645 River Sediment NBS 1571 NIST 1572 NBS 1573 Orchard Leaves Citrus Leaves Tomato Leaves *Value not certified.1 . O f 0.15 2013 10013 - 2012 10013 10.2f 1.5 1015 100125 0.1 1 fO.01 20012 0.03 f 0.0 1 9512 3* 5/ 1 2515 10015 1.1 f O . 1 1.0 f 0.1 1.3 f 0.1 1.2 f 0.1 7.9 f 0.5 8 . 0 ~ 1.0 0.08 f 0.06 - 2.7 f 0.6 3.2 f 1.5 2.5 kO.5 1 .Of 0.1 1 .Of 0.1 1.2 f 0.1 1.1 20.1 8.9 -t 0.6 9.6 f 0.8 0.12 -t 0.01 0.029 f 0.003 3.0f 0.3 2.7 f 0.4 2.6 -t 0.2 Table 6 Determination of lead in reference materials by ETAAS using slurry sample introduction Result found (X-t a)lpg g-' Sample ( 2 a)/* g-' mg ml-l sample Slurry Certified value concentration/ Original sample Finely ground NBS 1633a Coal Fly Ash IAEA Soil 7 NBS 1645 River Sediment NBS 1571 Orchard Leaves NIST 1572 Citrus Leaves NBS 1573 Tomato Leaves 72.4 f 0.4 20/2 55-7 1 2015 200/15 100110 714f28 512 80120 4 5 f 3 1012 100110 13.3 k 2.4 2012 200/2 6.3 f 0.3 2012 10015 69.0 f 7.6 7 1.8 f 6.4 59.8 f 7.2 52.6 f 5.8 704f 134 7 4 t h 112 54.1 f 11.0 47.7 f 7.6 15.9 f 4.5 6.2 k 1.3 6.3 f 2.5 14.0f 1.3 67.4 f 6.0 71.3k4.3 57.6 f 5.2 51.8f 1.0 774-t 108 748 -t 60 52.7 k 5.3 48.2 f 1.9 13.6-tO.5 13.6 f 0.5 6.5 f 1.1 6.5 5 0.4 The effect of particle size on the quality of the analytical results has been discussed by several workers but opinions were fairly different.Fuller e? ~ 1 1 . ~ ~ analysing ore and silicate rock samples set a limit of 25 pm above which sampling was the major problem. Hinds e? a1.,16 on the other hand reported poor recovery of cadmium and lead from larger particles in the atomization of soil surries. In order to achieve good results these samples should be ground until 90% of particles are less than 1 1 pm in diameter.Ebdon and LechotyckP accurately determined cadmium in biological and botanical reference materials which had been ground until 90% of the particles were less than 20 pm. Stephen e? all9 reported that stable slumes of vegetable and protein samples can be obtained if the majority of the particles are less than 50 pm in size. Contrary to these workers Miller- Ihliz5 found better precision for coarse (250-600 pm) but more uniform NBS SRMs of mixed diet when measuring eight different elements. In the present study a variety of geological and plant material samples were slurried for the determination of cadmium and lead. The mass of the sample and the particle size distribution were the variable parameters. Nine consecutive measurements of 10 pl aliquots were made. The results are summarized in Tables 5 and 6.In comparison with the direct weighing in sample introduction 1-3 orders of magnitude greater sample amounts were used in the preparation of slurries which should almost eliminate the problem of sample representa- tion. Indeed all of the values but one obtained for the six different RMs were closely matched with the certified values. The variation of sample mass in the range 10-200 mg did not influence the accuracy of the results even when plant material powders of fairly non-uniform particle size distribution (see Table 2) were atomized. Fine grinding of these samples which significantly reduces both the mean particle size and distribution did improve the precision of measurement.The same effect was observed for the determination of lead in the soil sediment and coal fly ash samples. However the particle size in these samples with the exception of the coal fly ash was only slightly affected by additional grinding. It could therefore be anticipated that particle size distribution would have little effect on the accuracy and precision of results when slurries of biological botanical or food samples were atomized. Although a slight improvement in either accuracy or precision for slurries of geological soil or sediment samples was obtained grinding to a particle size less than 30 pm15 is recommended for these660 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 199 1 VOL. 6 100 50 - E D C .- 0 2.5 Time/s 5.0 Fig. 10 Effect of chemical modifier on cadmium absorption profile.Graphite cup type B (8 mm). Sample IAEA A2 Dried Animal Blood (2.16 mg). (a) Without chemical modifier and (b) chemical modifier added. 1 Cd absorbance; and 2 background absorbance materials in order to achieve a more uniform evaporation of elements from the matrix. Conclusions An atomic absorpition spectrometer capable of measuring transient signals was assembled and a graphite cup atom- izer suitable for direct atomization of solid samples was designed. A variety of inner cups acting as a platform could be incorporated in the graphite atomizer thus enabling the rate of entry of the analyte into the analytical volume to be varied substantially. This facilitates the extension of the useful analytical range over two orders of magnitude without changing the analytical line.No matrix condensa- tion problems often experienced with tube atomizers were observed. Different geochemical (soil and sediment) and biological (plant) RMs were analysed and direct weighing in and slurry sample introduction techniques compared. Aqueous standards were used for calibration and the addition of an (NHJ2HP04 chemical modifier was found to be necessary in the analysis of botanical samples. Slurry sample introduction was found to be superior to the direct weighing in technique for the following reasons (i) better accuracy and precision were obtained enabling easier and more rapid sample handling and offering the possibility of dilution at high analyte concentration; and (ii) the particle size and the distribution of the sample was found to be extremely important in the atomization of small sample aliquots (20-500 pg) when using direct weighing in sample introduction.A sub-micrometre particle size would prob- ably be required for geological samples in order to achieve acceptable accuracy and precision. In contrast the results obtained by slurry sample introduction using sample amounts that were 1-3 orders of magnitude larger were less susceptible to particle size effects but the effects of particle size should not be totally ignored particularly when geochemical samples are analysed. Professor S. A. Katz of the Rutgers University New Jersey USA and Dr. W. Frech of the Department of Analytical Chemistry University of Umei Sweden are thanked for the fruitful discussion and valuable suggestions in preparing this manuscript.The supply of graphite material by B. Hutsch of Ringsdorff-Werke is also gratefully acknow- ledged. References 1 Langmyhr F. J. Talanta 1977 24 277. 2 Slavin W. and Manning D. C. Prog. Anal. At. Spectrosc. 1982 5 243. 3 Langmyhr F. J. and Wibetoe G. Prog. Anal. At. Sprectrosc. 1983 8 193. 4 Kurfurst U. Fresenius Z. Anal. Chem. 1987 328 3 16. 5 Ajlec R. and Stupar J. Spectrochim. Acta Part B 1983,38,7. 6 Stupar J. and Ajlec R. paper presented at The Third International Colloquium on Solid Sampling with Spectros- copic Methods Wetzlar October 10th- 12th 1988. 7 Chakrabarti C. L. Wan C. C. and Li W. C. Spectrochim. Acta Part B 1980 35 93. 8 Rygh G. and Jackson K. W. J. Anal. At. Spectrom.1987,2 397. 9 Atsuya I. Itoh K. and Akatsuka K. Fresenius Z. Anal. Chem. 1987,328 338. 10 Fleckenstein J. Fresenius Z. Anal. Chem. 1987 328 396. 11 Vollkopf U. Grobenski U. Tamm R. and Welz B. Analyst 1985 110 573. 12 Strubel G. Rzepka-Glinder V. Grobecker K. H. Fresenius 2. Anal. Chem. 1987 328 382. 13 Lucker E. Rosopulo A. and Kreuzer W. Fresenius Z. Anal. Chem. 1987,328 370. 14 Homer E. and Kurfurst U. Fresenius 2. Anal. Chem. 1987 328 386. 15 Jackson K. W. and Newman A. P. Analyst 1983 108 261. 16 Hinds M. W. Jackson K. W. and Newman A. P. Analyst 1985 110 947. 17 Karwowska R. and Jackson K. W. J. Anal. At. Spectrom. 1987 2 125. 18 Hinds M. W. and Jackson K. W. J. Anal. At. Spectrom. 1987 2 441. 19 Stephen S. C. Littlejohn D. and Ottaway J. M. Analyst 1985,110 1147. 20 Stephen S. C. Ottaway J. M. and Littlejohn D. Fresenius Z. Anal. Chem. 1987 328 346. 21 Ebdon L. and Lechotycki A. Microchem. J. 1987 36 207. 22 Olayinka K. H. Haswell S. J. and Grzeskowiak R. J. Anal. At. Spectrom. 1986 1 297. 23 Fuller C. W. Hutton R. C. and Preston B. Analyst 1981 106 913. 24 Carridn N. de Benzo Z. A. Moreno B. Fernhndez A. Eljuri E. J. and Flores D. J. Anal. At. Spectrom. 1988 3 479. 25 Miller-Ihli N. J. J. Anal. At. Spectrom. 1988 3 73. 26 Slavin W. Manning D. C. and Carnrick G. R. At. Spectrosc. 1981 2(5) 137. 27 Majidi V. and Holcombe J. A. Spectrochim. Acta Part By 1990 45 753. 28 Holcombe J. A. and Majidi V. J. Anal. At. Spectrom. 1989 4 423. 29 Miller-Ihli N. J. Fresenius J. Anal. Chem. 1990 337 271. 30 Epstein M. S. Carnrick G. W. Slavin W. and Miller-Ihli N. J. Anal. Chem. 1989 61 1414. 31 Miller-Ihli N. J. J. Anal. At. Spectrorn. 1989 4 295. 32 Hoenig M. and Van Hoeyweghen P. Anal. Chem. 1986,58 26 14. 33 Baxter D. C. and Frech W. Fresenius J. Anal. Chem. 1990 337 253. Paper I /0025 7K Received January 18th 1991 Accepted July Ist 1991

ISSN:0267-9477    

DOI:10.1039/JA9910600653

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 199 1 VOL. 6 66 1 Determination of Selenium by Electrothermal Atomic Absorption Spectrometry Part 1. Chemical Modifiers* Hana D&ekalova Veterinary Research Institute Hudco wa 70 CS-62 1 32 Brno Czechoslovakia Bohumil Ddekal Institute of Analytical Chemistry Czechoslovakia Academy of Sciences Vevefi 97 CS-611 42 Brno Czechoslo wakia Josef Komhrek and Ivan Novotny Department of Analytical Chemistry Faculty of Science Masaryk University Kotlakka 2 CS-611 37 Brno Czechoslovakia The effect of various chemical modifiers including nitrates of palladium nickel magnesium calcium lanthanum europium and aluminium on the analytical signal of selenium in a graphite furnace was studied. The signals of various selenium compounds such as selenite selenate and organic compounds representing different types of selenium forms in body fluids (selenomethionine and trimethylselenium iodide) were evaluated.The shape of the transient signal appears to be influenced not only by the chemical reactions in the graphite-selenium-modi- fier system but also very strongly by vaporization effects connected with the physical character of the charred residue. It follows that successful chemical modification involves the application of a considerable excess (higher than 1000-fold) of some metal nitrates which produce refractory oxides and no thermally stable carbides and are at the same time capable of quantitative conversion of the analyte into a single form. An integral part of the modifier action is trapping of the resulting compound by the modifier residue.Keywords Chemical modification; modifier effect; selenium determination; electrothermal atomic absorption spectrometry The determination of selenium by electrothermal atomic absorption spectrometry (ETAAS) has been the subject of many studies in recent years. The elucidation of the processes taking place in the electrothermal atomizer is still a great challenge for many atomic spectroscopists. A comparison of the results presented in many papers over the last decade indicated that for selenium these processes are complicated and variable. 1-9 The following factors play a decisive role the type of selenium cornpound1v2 and the type of graphite used for the construction of a tube or a platform (ordinary electr~graphite,~~~*~ pyrolytic graphite coated ele~trographite,~~ totally pyrolytic glassy carbon7 and graphite cloth3) the composition of the modifier the procedure for its preparation* and the compo- sition of the sheath g a ~ e s .~ - ~ Some processes taking place during the charring and atomization steps have been studied in detai11*2s4-10 but general comparison of most other results is confusing. This is obviously caused by the application of different experimental conditions. In order to study atomization and other processes in the electrothermal atomizer several workers have employed devices of different designs,'-'* in many instances of unusual c~nstruction,~*~*~ commercial and non-commerical tubes andor platforms. Experiments have often been performed under complicated non-stabilized atomization temperature condition~.~*~-~*~ During thermal pre-treatment the behaviour of various selenium compounds especially those containing (Se-") or those metabolized in biological samples and traced by in vivo incorporation of 75Se has seldom been e~amined.l-~.~*~J 1-14 Additionally modifiers of various compositions have often been utilized in different *Presented in part at the 3rd Conference on ICP and Develop- ment Trends in Atomic Spectroscopy held on the occasion of Professor E.PlSko's 60th binhday at Smolenice Czechoslovakia March 26th-30th 1990. mass ratios of modifier to analyte.14-16 Therefore the results cannot be compared directly in many instances. In general modifiers can only be applied with some limitations to the analysis of real samples by conventional ETAAS.Originally the ultimate aim of this work was to examine the stabilization and modification effects of various com- pounds on different forms of selenium in a commercially available graphite tube atomizer equipped with a L'vov platform. The work was undertaken with a view to choosing a suitable and efficient modifier for the direct analysis of body fluids.17 However owing to the complexity of the problem studied it proved to be necessary to examine the behaviour of the modifier for various pure selenium compounds first. This topic is dealt with in this initial paper. Experimental Apparatus A Perkin-Elmer Model 3030 atomic absorption spectro- meter with deuterium background correction equipped with an HGA-500 graphite furnace an AS-40 autosampler and a PR-100 printer was used for the atomic absorption measurements.The instrumental parameters used are summarized in Table 1. Unless otherwise mentioned Perkin-Elmer pyrolytic graphite coated graphite tubes (Part No. 121 092) with L'vov platforms of a 'new' type (Part No. 121 091) with cavity dimensions of 3 x 13 mm were used for the experk ments. Also two additional platform types were used for comparative measurements. A platform of the older type (marketed recently by Perkin-Elmer under the same part number) with approximate cavity dimensions of 1.8 x 13 mm and platforms of the design resembling the 'new' type manufactured from Slovak commercial spectroscopic elec- trographite SU and SG (Elektrokarbon TopolEany Czechoslovakia) (these having different density and poro-662 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 1991 VOL.6 Table 1 Instrumental parameters and settings Temperature programme TemperatureV'C Drying stage 90 140 Chamng stage Variable Atomization stage 2400 Cleaning stage 2650 Cooling stage 300 Cooling stage 20 Ramp/s Holds 10 30 10 10 20 20 1 20 ot 3 1 3 1 20 Element Internal argon flow/ ml min-I 300 300 300 300 ol 300 300 Parameter se Pd Ni Wavelengtldnm Spectral bandwidthlnm Radiation source 196.0 204.0 247.6 303.8 2.0 0.7 0.2 0.2 EDL (5 W) HCL (30 mA) HCL (1 5 mA) * Nominal values. t 'Maximum power heating' with optical feedback control (-2 K ms-l). $ Gas stop; read command selected. sity) and from RW 1 spectroscopic graphite and graphite cloth (Ringsdorffwerke Bad Godesberg Germany) were used.For measurements of the timedependent temperature of the tube or platform during the atomization step a TMR 4854 optical pyrometer (Dr. Georg Maurer GmbH Kohl- berg Germany) equipped with a VL 100 optical adaptor and an ETSZL detector-amplifier unit was used. Either a defined segment (an area with diameter of 1.5 mm) of the tube wall or the platform in the central part of the tube were imaged and focused through the sampling hole on to the pyrometer sensor. Scanning electron microscopy (SEM) was used to obtain micrographs of the platform surface. A Tesla BS 300 microscope (Tesla Czechoslovakia) employing an acceler- ating voltage of 25 kV with magnification ranging from 200x to ~ O O O O X and a Philips SEM 505 microscope employing an accelerating voltage of 20 kV equipped with an EDAX PW 9900 microprobe system for surface analysis were used.These imaging techniques enabled the detection of the modifier as a separate phase on the graphite surface if it was present in amounts of greater than 1 x lo2 ng. Reagents Stock solutions of selenium were prepared from analytical- reagent grade selenium oxide sodium selenate (Lachema Bmo Czechoslovakia) selenomethionine (Sigma) and tri- methylselenium iodide (Institute of Nuclear Biology and Radiochemistry Prague Czechoslovakia). Stock solutions of the modifier were prepared from analytical-reagent grade crystalline palladium nickel magnesium calcium lan- thanum europium and aluminium nitrates (Lachema Brno) and acidified with analytical-reagent grade HN03.The substances used for the preparation of concentrated solu- tions were checked for selenium contamination and yielded a negative result. Spectroscopic-reagent grade argon was used as a sheath gas. Procedure Typically 20 pl of a solution containing either 2 ng of selenium in the selected chemical form or the same amount of selenium modified with various amounts of the selected nitrate were dispensed into the platform cavity. The acidity of the final solution was adjusted with HN03 to 0.003 mol 1-l in the acid. An independent measurement verified that this concentration had no influence on either the magnitude of the signal or its shape. Absorbance signals for selenium were measured and recorded following the charr- ing procedure at various temperatures (see Table 1).In order to eliminate possible influences of cross-memory effects between several different modifiers on the selenium signal a new untreated tube equipped with a new platform was usually used for each nitrate tested. Measurement of Analyte Distribution If double or multiple peaks were observed selenium distribution between the platform and tube wall was checked. After the charring step the platform was removed from the tube and the fraction of selenium retained on the tube wall was determined using peak area evaluation. The selenium remaining on the platform surface was measured following the re-insertion of the platform into the grooves of the tube. Measurements performed under the same conditions but without the removal and re-insertion proce- dure showed that the above technique is justified i.e.that the peak areas from the wall and platform are additive. Results and Discussion Selenium in the Absence of a Modifier Selenium signals of various separately sampled selenium compounds were measured under a variety of charring conditions with different graphite platform materials (see Figs. 1-3). Commercially available ('new' type) pyrolytic graphite platforms did not show a significant amount of reactivity with the different selenium compounds as can be seen in Fig. 1. It appears that the extent of selenium stabilization in the atomizer corresponds to the thermal behaviour of the particular selenium compounds. While selenate and selenite are volatile above 300 0C,497 seleno- methionine decomposes below 200 "C and trimethyleselen- ium iodide dissociates into its fairly volatile constituents dimethyl selenide and methyl iodide without distinct phase changes at even lower temperatures.'* As has been ob- served the signals corresponding to the remaining selenium show a similar shape (appearance time time of peak maximum etc.) for all the selenium compounds selected.Charring temperature had no influence on the signal shape. Finally variation of the volume of solution dispensed between 5 and 80 pl (while keeping the analyte massJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 199 1 VOL. 6 v) s f Y Q L t A 0 Tube temperaturePC Fig. 1 Chamng curves for various selenium compounds. Sele- nium (2 ng) was dispensed as 20 pl solutions of A selenate; B selenite; C selenomethionine; and D trimethylselenium iodide onto the 'new' ordinary pyrolytic graphite platform 0 500 lo00 1500 Tube temperaturePC Fig.2 Chamng curves for platforms of different graphite ma- terials. Selenium (2 ng) as selenite (20 pl) was placed onto the platform. P commercially available 'new' pyrolytic graphite; Po 'older' SU SG and RW home made from the corresponding electrographite materials; and GT graphite cloth. Charring tem- perature 500 "C constant) showed no influence on the magnitude and shape of the signal. Measurements with selenium using amounts differing by two orders of magnitude (2 and 200 ng) showed no changes in signal shape with concentration in contrast to the observations made by Di5dina et al.5*6 who used a platform made from polycrystalline graphite. In contrast to totally pyrolytic graphite a pronounced reactivity of the platform graphite with inorganic selenium compounds was observed when the platform was made of polycrystalline electrographite material.In analogy with the observations of Chung et al.3 and Di5dina et al.,596 the graphite substrate stabilized selenate and selenite (see Fig. 2). Time-resolved signals displayed double peaks (see Fig. 3) the first of which decreased with increasing charring temperature and the second of which remained constant up to 1300 "C. When the commercially available 'new' type of platform was used the latter peak showed a slight signal tailing ( ( 5 % of the integrated absorbance signal) (see Fig. 3). Additional measurements at a charring temperature of 500 "C confirmed that the double peaks are not due to selenium deposition on the tube wall.In all instances less than 2Oh of the selenium applied was transferred to the tube wall. Unlike the inorganically bound selenium organic sele- SG p 0.5 2 0 P" RW 2 Time/s 4 663 Fig. 3 Time-resolved signals for selenium using platforms made from different graphite materials (for specification see caption to Fig. 2) nium compounds particularly trimethylselenium iodide did not show any interaction with graphite. The first peak on the signal trace was considerably larger than the second and disappeared at lower temperatures analogous to the situation shown in Fig. 1. Stabilization of organic selenium compounds therefore seems to be considerably less pro- nounced in contrast to selenite as shown in Fig.2. Selenium in the Presence of a Modifier Selenium signals were measured in the presence of various amounts of selected modifying compounds at various charring temperatures and for different chemical forms of selenium. The signal traces for selenate are summarized in Fig. 4 however the same trends were observed for selenite selenomethionine and trimethylselenium iodide. Charring curves in the presence of palladium and magnesium nitrates as chemical modifiers are shown in Figs. 5 and 6. Modifica- tion of all the selenium compounds studied with the remaining (nickel aluminium calcium europium and lanthanum) nitrates led to charring curves with the same characteristic form as with magnesium nitrate as shown in Fig. 6. It was noted that in several instances (e.g.for calcium and lanthanum nitrates) peak shape and peak area values changed dramatically with the amount of modifier present (see Fig. 4). In comparison with commonly used palladium and nickel compounds the application of the remaining nitrates produces surprisingly similar results but for a greater molar excess of the modifier. It can generally be stated that the shape and the position of the signals varies little with the form of the selenium compound and with the charring temperature. It follows that the type of atomization reaction does not appear to be the limiting factor influencing the shape of the signal.664 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 199 1 VOL. 6 1.4 1 (a) 0 1 100 10 A i o o o 1 1 I a 1.3 - ( C ) 300 3000 300 000 s 0 1.2 -(el O r 0 1 2 1 .o 0 30 000 1.0 - ( f ) 0 - 0 1 2 3 0 P E! 0 w x $ I- 0 1 2 3 Time/s Fig.4 Absorbance versus time profiles for 2 ng of selenium as selenate in the presence of the selected modifier (a) palladium nitrate (b) nickel nitrate (c) magnesium nitrate (d) aluminium nitrate (e) europium nitrate u> calcium nitrate and (g) lanthanum nitrate. The modifier to selenium molar ratio is indicated on the corresponding traces. A charring temperature of 600 "C was chosen except for 300 "C for aluminium nitrate. The atomization temperature was 2400 "C except for nickel nitrate where 2100 "C was used. The broken line represents the signal of the modifier element (3 ng of palladium and 100 ng of nickel) and the dotted line the background signal of the aluminium matrix.A is the temperature profile of the tube; and B the platform An increasing amount of the modifier shifts the time of appearance of the signal to a greater value (see Fig. 4) so that the atomization of selenium takes place under the stabilized temperature conditions in the atomizer. For small amounts of some of the modifiers used splitting of selenium signals was observed (see Fig. 4). The first part of the split peak disappeared above a chamng temperature of 600 "C while the second peak remained up to 900-1 200 "C. For organic selenium compounds the first peak was larger than that for the inorganic form. The signal splitting disappeared as the amount of modifier increased 5 0 Q confirming the critical effect of the amount of modifier on b 0.5 the stabilization efficiency.0 The extent of selenium stabilization in its various n chemical forms in the atomizer as a function of the amount of the modifier and of the charring temperature applied is summarized in Figs. 5 and 6. It was found that the selenium signals for the four selenium compounds examined pos- sessed similar shape and magnitude if the molar excess of the modifier was approximately 300-fold for palladium (1 pg) 30 000-fold for magnesium nitrate (200 pg) 2000-fold for aluminium nitrate (20 pg) and 2000-fold for europium nitrate (20 pg). For nickel calcium and lanthanum nitrates at their maximum applicable modifier concentration ( 100 0 500 1000 1500 10 and 0.5 mg ml-l) the signals of trimethyselenium iodide were only about 50,90 and 80% respectively of the selenite 0.5 v) Y Tube temperature/"C selenium molar ratio A 0; B 1; C,- 10; D 100; E 1000; F 10 000; and G 40000 on the physico-chemical parameters of compounds likely to be produced in the atomizer e.g.selenate and selenite. ForJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 199 1 VOL. 6 665 0 I -.-I '1 0 500 1000 1500 \ X Tube temperaturePC Fig. 6 Chamng curves for 2 ng of selenium as (a) selenate and selenite (b) selenomethionine and (c) trimethylselenium iodide in 20 pl of solution for various amounts of magnesium nitrate. The broken lines were obtained for the modifier mixture of magnesium and palladium nitrates in the molar ratio of 300000 (magne- sium) 4000 (palladium) 1 (selenium) i.e. 2 mg of magnesium nitrate with 10 pg of palladium.Modifier to selenium molar ratio A 0 B 300; C 3 000; D 30 000; and E 300 000 example the affinity of some metal ions such as NiZ+ CuZ+ and especially PdZ+ for the functional groups containing selenium is well known.19 Chemisorption of Se02(g) and Se(g) by MgO at 500 "C by CdO at 700 "C and by La203 at 1100 "C has also been d e s ~ r i b e d . ~ ~ - ~ ~ Therefore the transformation of all the selected selenium chemical forms into SeO by the excess of nitrate present can be expected in addition to sorption of SeO by the metal oxides formed. It is appropriate to mention in this context that there are no references in the literature suggesting that selenium can be stabilized by the formation of metal selenites which are usually thermally stable up to 600-700 "C and in some instances up to 1000 "C (for calcium and lanthan~m).~O-,~ The formation of such thermally stable selenium com- pounds during the analysis of biological samples e.g.body fluids could be advantageous as it could provide a common oxidative ashing procedure unlike the method of selenide formation which usually requires different chemical condi- tions. The modifiers tested except for calcium and lanthanum nitrates and large amounts of palladium nitrate did not give rise to any selenium or modifier memory effects. After 0 1 2 3 4 5 Tirnels Fig. 7 Influence of platform contamination by lanthanum on the selenium signal for 2 ng of selenium as selenite in 20 pl of solution containing A and B lanthanum nitrate (10 pg 900-fold molar excess); and C magnesium nitrate (2 mg 300000-fold molar excess).Traces A and B were recorded before and after contamina- tion caused by 100 pg of lanthanum nitrate (9000-fold molar excess) respectively. Trace C is not affected by the contamination the cleaning step the platform behaved in a similar manner to the untreated one based on measurements with pure selenium solutions. Such measurements showed peak char- acteristics closely resembling those obtained with a new platform. The lifetime of the platforms was 300-400 firing cycles without any perceptible change in the analytical properties. Calcium and lanthanum nitrates and large amounts of palladium nitrate ( 100 ,ug of palladium) caused considerable corrosion of the platform graphite owing to carbide formation and intercalation.The peak height and the peak area values first increased with an increasing amount of the modifier (see Figs. 5 and 6). This corresponds proportionally to the stabilization effects. However signal depression and broadening were caused by a much larger amount of a modifier (see Fig. 4). Remarkable differences were observed between the effects caused by the following nitrates calcium versus magnesium lanthanum versus europium and palladium versus magne- sium. A larger amount of a refractory matrix probably influences the atomization processes and thereby reduces the selenium signal. During the atomization step at temper- atures above 1600 "C calcium and lanthanum oxides produce thermally stable ionic carbide~,~O-~~ in contrast to magnesium and europium ~ x i d e s .~ l J ~ These carbides obvi- ously prevent the release of trapped selenium from the modifier residue (see Fig. 4). Similar effects of slow vaporization from a compact molten residue enhanced by intercalationZS might be expected with the use of a larger amount of palladium modifier. The influence of the platform memory effect on the selenium signal caused by lanthanum nitrate is shown in Fig. 7. This effect cannot be eliminated by repeating the cleaning step. The use of magnesium nitrate modifier prevented this problem. It can therefore be concluded that the magnesium nitrate matrix traps selenium and minimizes the contact between sele- nium compounds and the lanthanum oxide residue. These results also suggest that there is no evidence for a gas-phase interference caused by the decomposition products of lanthanum oxide (oxygen releasing effect) during the atomization step.A similar effect caused by a calcium nitrate residue which is finally converted into the less thermally stable carbide,2z was found to be relatively smaller. Exact interpretation of most of the results mentioned above characterized by general and common trends of changes in the shape of the signal could not be based solely on chemical interactions chemical modification or stabili- zation and on the formation of a new chemical b ~ n d * ~ * ~ ' (see also refs. 28-31 for comparison of common trends). The physical and physico-chemical properties of dried and charred residues such as the grain size of particles heat of vaporization melting- and boiling-points of the compounds generated also seem to play an important role by critically determining the shape and position of the signal.Vaporiza-666 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 199 1 VOL. 6 tion effects might therefore act as a limiting step that influences the form of the final transient signal. Vaporization of a relatively refractory oxide matrix which is produced from the nitrate modifier during the charring step and traps selenium limits the vaporization and consequently also the atomization of selenium. It could be expected that the modifier allows selenium to vaporize and atomize only under such conditions where the matrix of the refractory modifier itself is volatilized. This is demonstrated for example by signal traces in Fig.8. A comparison of the signals for selenium and some modifier elements (see Fig. 4) also confirmed this conclusion. The matrix of the modifier traps the selenium in its bulk not only by chemical reaction but also by physical interactions such as sorption occlusion dissolution formation of solid solutions and melts depending on the pre-treatment condi- tions. This agrees very well with the situation where the tested mixtures of the modifier and trimethylselenium iodide were only dried at 140 "C (see Fig. 9). Chemically stable trimethylselenium iodide could not react with nitrate at this temperature and under these chemical conditions and thus it could not be converted into the selenite form. However the same signal shift and the signal increase were observed for this highly volatile selenium compound1* as for the selenite form.The position of the signal was not varied by higher temperature pre-treatment in comparative measurements using a charring step. For all selenium compounds tested the signal splitting disappeared if the modifier was used in a larger excess (see Fig. 4). It follows that this effect can only be caused by physical trapping in the matrix of the modifier. The SEM micrographs of the charred modifier residue on the platform surface showed that there were differences in the grain size and the spatial distribution depending on the amount of modifier. For example with a relatively small amount of modifier i.e. 300 ng of palladium well-separated particles with an approximate diameter of 1 x 1 0-1 pm were Fig.10 Scanning electron micrograph of 300 ng of palladium deposited as nitrate solution on the platform surface. Charring temperature of 1100 "C was applied. Magnification 5000x was used on an SEM Philips 505 microscope I I I J G 1 2 3 Time/s Fig. 8 Influence of modifier composition on the signal shape for 2 ng of selenium in 20 p1 of solution containing A palladium nitrate ( 10 pg of palladium 4000-fold molar excess); B magnesium nitrate (2 mg 300 000-fold molar excess); and C mixture of the same amounts of palladium and magnesium nitrates as for A and B I I I 0 1 2 3 Time/s Fig. 9 Time-resolved signals for dried non-charred residue containing 2 ng of selenium as trimethylselenium iodide and also with 200-fold molar excess of aluminium calcium europium lanthanum and magnesium nitrates.A drying and charring temper- ature of 140 "C was applied before atomization. The dotted line represents the signal of the non-modified solution 10 pm Fig. 11 Scanning electron micrograph of 100 pg of palladium deposited as nitrate solution on platform surface. Charring temper- ature 1 100 "C was applied. Magnification 5000x was used on a BS 300 Tesla microscope observed on the graphite surface (see Fig. 10). On the contrary with a larger amount of modifier e.g. 100 pg of palladium a significant agglomeration of particles was noticed (see Fig. 11). Similar micrograph patterns were observed for the remaining nitrates tested. The differences in the shape of the signal (see Fig. 4) such as changes in width and broadening could be influenced by the physical conditions of the residue as it vaporizes more easily from small separated particles or drops than from either robust agglomerates or from a compact film of the melt or alloy.Effect of the Modifier An attempt has been made to formulate a general hypothe- sis about the effects of modifying agents on the selenium signals based on a comparison of the experimental results discussed above with literature data.I-l0J5 It has been noted that the amount of modifying compound critically affectsJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 199 1 VOL. 6 667 the data observed and the signal characteristics i.e. appearance time peak area and peak shape. The following conclusions were therefore drawn as a function of the amount of modifier present. In the absence of a modifier the processes are determined by mutual affinity of ‘active sites’ on the graphite plat- for the chemical forms of selenium being used and/or by their decomposition and vaporization character- istics as mentioned above.The extent of graphite activa- tion and interaction with selenium is very difficult to define as the amount of selenium is very low (1 x 10-lo-1 x lo-” mol). Therefore different stabilization efficiences distorted and split signals have often been observed. Stabilization of selenium by graphite is therefore of no practical impor- tance as concluded by D&dina et U Z . ~ ~ ~ particularly for the analysis of biological samples. In the presence of a ‘small’ amount of modifier i.e. with an equimolar ratio or with a molar excess of one or two orders of magnitude over selenium and probably over active graphite sites the processes in a modifier-analyte- graphite three-component system are more complicated.Depending on the affinity the modifier either reacts only with the selenium compound to form a new thermally stable precursor such as selenide’ or ~elenite,~,~ or graphite also affects the reaction producing ternary compo~nds.~J~ A slight excess of the chemical modifier might trap the new species in the matrix and thereby also affect the atomiza- tion. Therefore it appears that some confusing data have been presented in the literature as unfortunately the majority of work has been oriented towards the study of a ‘small’ amount of modifier. In the presence of ‘large’ amounts of modifier i.e. with the molar excess over selenium of greater than 1000-fold the incidental interactions of analyte with graphite are minimized and the reaction with the modifier prevails.Analyte compounds can be transformed into thermally stable compounds or can be trap~ed~O9~j by adsorption chemisorption and other physico-chemical processes in the crystal cage (lattice34) of the matrix of the modifier. In line with these physico-chemical phenomena the extent of analyte trapping controls the processes of atom formation to a considerable extent. Thus the vaporization of the carrier matrix and its interaction with graphite influences the release of the analyte. For example carbide formation or intercalation of the modifier residue during the atomiza- tion step has a negative affect on the release of analyte.Therefore a larger amount of a suitable modifying agent except for compounds forming thermally stable carbides should be applied for practical use. Conclusion According to observations the resulting transient signal for selenium appears to be determined not only by various solid-phase chemical reactions which have usually been accepted and discussed as critical phenomena but also strongly by the effects of physical and physico-chemical processes e.g. by occlusion and sorption in the matrix of the modifier and by vaporization of the residue of the modifier. Some refractory oxides and metals are efficient modifying carriers. Corresponding nitrates are the most suitable precursors of these carriers as the other salts such as chlorides and sulphates are potential interferents* during the charring and atomization procedures and they do not easily or completely form oxides.Nitrates are also the most suitable oxidizing agents to facilitate the mineralization of an organic matrix. The limitation that the chemical modifier should not form a thermally stable carbide during the atomization restricts the choice of a suitable agent from a group of metal nitrates. With regard to all the experimental limitations and other considerations a mixture of palladium and magnesium nitrates with a considerably larger amount of the latter than usual is proposed as the most efficient modifier for the determination of selenium particularly in biological fluids as both compounds react with selenium compounds at low and high temperatures and cause negligible corrosion of the graphite platform.At the same time the large amount of magnesium nitrate is expected to eliminate negative com- petitive effects of the sample matrix during the ashing procedure as will be reported in future work dealing with this topic. The results mentioned above clarify why the magnesium nitrate based modifiers have been popular for a wide range of application^.^^ The authors thank Dr. J. Kozel (Institute of Nuclear Biology and Radiochemistry Prague) for a donation of a pure trimethylselenium iodide standard Dr. M. Svoboda (Institute of Physical Metallurgy Czechoslovak Academy of Sciences Brno) and J. Kudrna (Veterinary Research Insti- tute Brno) for the SEM micrographs Dr. Z. Slovak Institute of Fine Chemicals Lachema Brno) and Dr.M. 5 ucmanova (Perkin-Elmer Austria) for instrumental sup- port of this work and Professor J. P. MatouSek (University of New South Wales Australia) and the referees of this paper for helpful comments and criticism during the preparation of the manuscript. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 References Welz B. Schlemmer G. and Vollkopf U. Spectrochim. Acta Part B 1984,39 501. Cedergren A. Lindberg I. Lundberg E. Baxter D. C. and Frech W. Anal. Chim. Acta 1986 180 373. Chung C.-H. Iwamoto E. Yamamoto M. Yamamoto Y. and Ikeda M. Anal. Chem. 1984 56 829. Droessler M. S. and Holcombe J. A. Spectrochim. Acta Part B 1987 42 981. DEdina J. Frech W. Lindberg I. Lundberg E. and Cedergren A. J Anal. At. Spectrom.1987 2 287. D5dina J. Frech W. Cedergren A. Lindberg I. and Lundberg E. J. Anal. At. Spectrom. 1987 2 435. Styris D. L. Fresenius 2. Anal. Chem. 1986 323 710. Voth-Beach L. M. and Shrader D. E. J. Anal. At. Spectrom. 1987 2 45. Droessler M. S. and Holcombe J. A. J. Anal. At. Spectrom. 1987 2 785. Styris D. L. Prell L. J. Redfield D. A. Holcombe J. A. Bass D. A. and Majidi V. Anal. Chem. 1991 63 508. Saeed K. Thomassen Y. and Langmyhr F. J. Anal. Chim. Acta 1979 110 285. Aadland E. Aaseth J. Radziuk B. Saeed K. and Thomas- sen Y. Fresenius 2. Anal. Chem. 1987 328 362. Lindberg I. Lundberg E. Arkhammar P. and Berggren P.-O. J. Anal. At. Spectrom. 1988 3 497. Alexander J. Saeed K. and Thomassen Y. Anal. Chim. Acta 1980 120 377. Schlemmer G. and Welz B. Spectrochim.Acta Part B 1986 41 1157. Kirkbright G. F. Hsiao-Chuan S. and Snook R. D. At. Spectrosc. 1980 1 85. DoEekalovB H. C.Sc. Thesis Masaryk University Bmo 1990. Rheinboldt H. in Houben- Weyl Methods of Organic Chemis- try S- Se- Te- Verbindungen ed. Muller E. Georg Thieme Stuttgart 1955 band IX pp. 974 979 980 1046 and 1187. Livingstone S. E. Comprehensive Inorganic Chemistry Per- gammon Press Oxford 1975 ch. 13. Gmelin s Handbook of Inorganic Chemistry Selenium Syst. No. 10 Supplment BI Compounds Springer Berlin 198 1 pp. 96 and 138. Gmelin ’s Handbook of Inorganic Chemistry Magnesium Syst. No. 27 Part B Verlag Chemie Weinheim 1939 p. 299. Gmelin s Handbook of Inorganic Chemistry Calcium Syst. No. 28 Part B Verlag Chemie Weinheim 196 1 pp. 822 and 835. Gmelin 3 Handbook of Inorganic Chemistry Lanthanides Syst. No. 39 Part CI Springer Berlin 1974 pp. 166 173 212 and 374.668 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 1991 VOL. 6 24 25 26 27 28 29 30 Duval C. Inorganic Thermogravimetric Analysis Elsevier Amsterdam 1963. Selig H. and Ebert L. B. Advances in Inorganic Chemistry and Radiochemistry Academic Press New York 1980 ch. 23 Teague-Nishimura J. E. Tominaga T. Katsuva T. and Sakurada O. Takahashi H. and Tap M. Bunseki Kagaku 1989,38,411. Hinds M. W. Katyal M. and Jackson K. W. J. Anal. At. Spectrom. 1988 3 83. Rettberg T. M. and Beach L. M. J. Anal. At. Spectrom. 1989 4 427. Slavin W. Carnrick G. R. and Manning D. C. Anal. Chem. 1982,54 621. 3 1 32 p. 281. 33 Matsumoto K. Anal. Chem. 1987 59 1647. 34 35 Hinds M. W. and Jackson K. W. J. Anal. At. Spectrom. 1990 5 199. Heuttner W. and Busche C. Fresenius 2. Anal. Chem. 1986 323 674. Tsalev D. L. Slaveykova V. I. and Mandjukov P. B. Spectrochim. Acta Rev. 1990 13 225. Tsalev D. L. Slaveykova V. I. and Mandjukov P. B. lecture presented at the VII Polish Spectroanalytical Conference and X CANAS Torun September 5th-9th 1988. Schlemmer G. and Welz B. Spectrochim. Acta Part By 1986 41 1157. Paper 0/03541F Received August lst 1990 Accepted July 23rd 1991

ISSN:0267-9477    

DOI:10.1039/JA9910600661

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 1 99 1 VOL. 6 669 Combination of Chemical Modifiers and Graphite Tube Pre-treatment to Determine Boron by Electrothermal Atomic Absorption Spectrometry Milagros Luguera Yolanda Madrid and Carmen Camara* Departamento de Quimica Analitica Facultad de Quimicas Universidad Complutense 28040 Madrid Spain Several ways of increasing the atomization efficiency in the determination of B by electrothermal atomic absorption spectrometry were evaluated the use of a chemical modifier (Ni Pd and ascorbic acid); pre- treatment of the graphite tubes with carbide forming elements (Zr Ta and V); and addition of NaF as a cleaning solution between measurements. The tubes were pre-treated by soaking in the carbide forming metal solution under vacuum drying for 4 h at 150 "C and finally subjecting them to slow electrothermal heating to at least 1600 "C in order to form the metal carbide on the inner side of the tube.The best results were obtained using Zr- treated tubes in conjunction with Ni as chemical modifier. This resulted in high sensitivity (80 pg) a negligible memory effect and an increased tube lifetime (200 atomization cycles of 3 s at 2650 "C). The method was successfully applied to determine B in waste and river waters. Keywords Pre-treatment of graphite tube; chemical modification; determination of boron; electrothermal atomic absorption spectrometry Conventional electrothermal atomic absorption spectrome- try (ETAAS) has a limited application in the determination of B because of the tendency of this element to form refractory carbides with the carbon of the graphite tube before atomization occurs.Among the problems associated with the formation of refractory carbides are loss of sensitivity and the occurrence of 'memory peaks' due to the atomization of residual boron carbides formed during previous determinations. This reduces the reproducibility of the analytical measurements. Furthermore the high vaporization temperature of boron carbide (about 2700 "C) reduces the tube lifetime.' Several methods have been reported in the literature to remove the problems mentioned above. Most of them attempt to prevent the formation of refractory carbides by means of a chemical modifier that inhibits this reaction such as Ca Mg Sr Ba Ni and La2-4 or a Ti and ascorbic acid m i x t ~ r e .~ Some methods involve pre-treating the graphite tube with elements that form carbides that are more thermodynamically stable than boron carbides. Others seek to destroy residual boron carbide by flushing the instrument with cleaning solutions between measure- ments such as NaF solution or CH30H-H2S04.1 The purpose of the present work was to evaluate the ability of the following procedures to increase the efficiency of B vaporization the use of chemical modifiers; pre- treatment of the graphite tube; flushing with cleaning solutions; and a combination of chemical modification and pre-treatment of the tube. Finally the method was applied to the determination of B in waste and river waters by ETAAS. Experimental Apparatus A Perkin-Elmer Model 1 1 OOB atomic absorption spectro- meter equipped with an HGA-400 graphite furnace atom- izer was used.A hollow cathode boron lamp was operated at 34 mA and the absorption was measured at 249.7 nm with a 0.7 nm slit-width. A deuterium lamp was used to correct for background absorption. * To whom correspondence should be addressed. Reagents All chemicals were of analytical-reagent grade or higher purity and de-ionized water from a Milli-Q system (Milli- pore) was used throughout. A 1000 mg 1-l stock standard solution of B was prepared by dissolving 2.86 g of H3B03 in 500 ml of de-ionized water and storing in a polyethylene bottle. Working solutions were prepared each day by diluting appropriate aliquots of the stock solution. Metal solutions used as chemical modifiers or as graphite tube pre- treatment agents were prepared from the salts or oxides of the pure metals depending on the species of interest.Tube Pre-treatment Procedure Both pyrolytic graphite and non-pyrolytic graphite tubes were immersed in 10 ml of a carbide forming metal solution (1 g 1-l for Zr and V and 10 g 1-l for Ta) in a bottle. A vacuum pump was connected for a few minutes to drive bubbles from the tubes and then atmospheric pressure was restored. The process was repeated until no more bubbles were formed. This procedure forces the carbide forming metal solution to flow into the pores of the tubes. The tubes were removed from the bottle and dried first at room temperature for 1 h and then in an oven at 150 "C for 24 h. The ends of the tubes were carefully polished with a soft laboratory tissue to remove any remaining salts that might alter the electrical conductivity between the graphite tubes and the graphite contact cones. The tubes were further dried by subjecting them to slow electrothermal heating to at least 1600 "C in order to form the metal carbide in the graphite tube. A similar procedure was followed with the L'vov platform.Sample Analysis Procedure River and waste water samples were collected in polyeth- elene bottles filtered and acidified with HN03 to pH<2. Nickel (to a final concentration of 1000 ppm) as a chemical modifier was added to the samples and the B was deter- mined by injecting 20 pl of the sample solution into a pyrolytic graphite coated graphite tube treated with Zr as described under Tube Pre-treatment Procedure. The graph- ite furnace temperature programme used is summarized in Table 1.The analytical peaks were recorded as integrated670 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 199 1 VOL. 6 Table 1 Graphite furnace temperature programme using Ni as a chemical modifier and a Zr-treated pyrolytic graphite coated graphite tube Air flow Step ture/"C time/s time/s ml min-l Tempera- Ramp Hold rate/ Drying 120 20 1 300 Ashing 900 20 20 300 Atomization 2700 0 3 0 Cleaning 2650 1 3 300 Cool down 30 1 5 300 absorbance. The concentration of B in the sample was obtained by extrapolation from the calibration graph. Results and Discussion Effect of Graphite Tube Type on Boron Atomization In order to study how the type of graphite tube employed affects the determination of B by ETAAS several types of tube (without pre-treatment) were tested non-pyrolytic graphite tubes; pyrolytic graphite coated graphite tubes; and L'vov platforms with pyrolytic graphite coated graphite tubes.Pyrolytic graphite coated graphite tubes gave a 75% higher sensitivity than L'vov platforms with pyrolytic graphite coated graphite tubes. Non-pyrolytic graphite tubes were totally unsuitable for the determination of low levels of B as they gave a poor sensitivity which could be due to the higher porosity that facilitates the formation of boron carbides and thus increases the memory effect. Although pyrolytic graphite coated graphite tubes were chosen for all further experiments the determination of B without a chemical modifier became very imprecise because of the poor sensitivity and the too wide analytical peaks that were obtained. Furthermore a large memory effect was obtained which made it necessary to apply 3-5 cleaning cycles between successive atomizations.This increased the analysis time and decreased the lifetime of the tubes (20 firings for a pyrolytic graphite coated graphite tube). Addition of NaF Cleaning Solution In order to prevent the memory effect 10 pl of 4% m/v NaF solution were injected between atomization cycles to transform the residual B that remained in the tube as B4C into the volatile species BF3. This decreased the memory effect so that it was only necessary to perform one cleaning cycle between measurements. Furthermore the sensitivity was significantly higher than for the conventional method. This could be explained by an interaction between the B and F remaining in the tube which facilitates B vaporiza- tion by the formation of the volatile species BF3.However this procedure was not sufficiently sensitive to determine B at low concentrations and the long analysis time and the short lifetime of the graphite tubes could not be overcome. Use of Ni Pd and Ascorbic Acid as Chemical Modifiers It is well established that elements such as Ca Ba Mg and Sr increase the efficiency of B vaporization owing to the formation of the respective borides which are more volatile than boron carbides. In this work the influence on the atomization signal of Ni and Pd (which are well known as universal modifiers) was studied both separately and in combination with ascorbic acid. The results displayed in Table 2 show that the use of these ~ Table 2 Relative sensitivity obtained with 20 pl of different chemical modifier solutions Relative Chemical modifier sensitivity B (1.0 pg ml-I) 1 Ascorbic acid (0.25% m/v) 45 Pd (1000 pg ml-I) 72 Ni (1000 pg ml-I) 170 Ni (1 000 pg ml-I) + ascorbic Pd ( 1000 pg ml-l) + ascorbic acid (0.25% m/v) 45 acid (0.25% m/v) 35 elements as chemical modifiers significantly enhances the signal Ni being the more effective. Other advantages are a lower background and a smaller memory effect i.e. it is not necessary to apply cleaning cycles between the measure- ments.The lifetime of the tubes increased by up to 100 firings without any significant loss in sensitivity in the determination of the B.The combination of Ni and Pd with ascorbic acid applied successfully in the determination of Se,6 gave worse results (Table 2) than those obtained using each element alone. Tube Pre-treatment Combined With the Use of Ni as Chemical Modifier In order to increase the efficiency of B atomization with respect to that obtained in the section above the pyrolytic graphite coated graphite tubes were pre-treated with car- bide forming elements. This pre-treatment has been suc- cessfully applied in the determination of other refractory metals such as Si and Al.7-9 Of the three tube pre-treatments tested (Zr Ta and V) the Zr-treated tube gave the best results. Although the characteristic mass obtained using Zr- treated tubes is higher than that obtained using Ni as a chemical modifier (Table 3) the former method increases the lifetime of the tube by up to 200 firings without deterioration. On considering these results both effects were combined for the proposed procedure.The results obtained using pyrolytic graphite coated graphite tubes pre- treated with several metals and Ni as chemical modifier are compared in Table 3 expressed in terms of characteristic mass and memory effect. The improvement obtained by the combination of Ni as chemical modifier and a Zr-treated pyrolytic graphite Table 3 Comparative results obtained in terms of sensitivity and memory effect using several procedures to determine B by ETAAS Type of pre-treatment of pyrolytic graphite Memory Characteristic Modifier coated graphite tube effect mass*/pg - - ++++t 24295 Ni - + +$ 143 ++ 338 Pd - Zr +§ 7 90 Ni Zr -7 88 Ni Ta 216 Ni V 126 - *The characteristic mass is defined as the mass of analyte in picograms required to give an integrated absorbance signal of 0.0044 absorbance seconds.t Very significant memory effect. # Significant memory effect. 0 Low memory effect. 7 No memory effet.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 199 1 VOL. 6 67 1 0.200 I 0.175 0) 0.150 x $ 0.125 n 0 0.100 0) U 0.075 - 0.050 0) U 0.025 1 1 &50 2600 2650 2700 Temperatu rePC Fig. 1 Effect of atomization temperature on the signal for 20 pl of 0.1 pg ml-I of B using Ni as chemical modifier. A Untreated pyrolytic graphite tube; B Ta; C V; and D Zr-treated pyrolytic graphite coated graphite tubes coated graphite tube could be explained by the conjunction of two factors (z] B is stabilized as nickel boride thus reducing the tendency of B to react with the graphite; and (ii) Zr forms a stable carbide on the carbon surface and reduces the boron-carbon interaction.On the other hand the use of pyrolytic graphite coated graphite tubes pre-treated with carbide forming elements provides higher sensitivity than that obtained with un- treated pyrolytic graphite coated graphite tubes at the optimum vaporization temperature. Fig. 1 shows the different integrated absorbance values obtained using Ta- V- and Zr-treated pyrolytic graphite coated graphite tubes and Ni as chemical modifier at several atomization temperatures. The use of a Zr-treated L'vov platform gives lower sensitivity than that with treated pyrolytic graphite coated graphite tubes.Determination of Boron in River and Waste Water Samples The combination of a Zr-treated pyrolytic graphite coated graphite tube and Ni as chemical modifier was applied to the determination of B in river and waste water samples. The analytical conditions were those indicated in the procedure and the results obtained are summarized in Table 4. In no instance was a matrix effect observed and hence the standard additions method was not required. The procedure gives a high recovery of B as measured by adding known amounts of B to the real samples. In order to confirm the accuracy of the proposed method the samples were analysed by inductively coupled plasma atomic emission spectrometry (ICP-AES) as an alternative tech- nique.The results show good agreement with those ob- tained by ETAAS (Table 4). The precision of the method (mean of ten values between batches expressed as relative standard deviation) was 6.1 ~ ~ Table 4 Determination of B in river and waste waters Concentration*/ Recovery ICP-AESI Waste water No. 1 1.26k0.02 9 6 k 4 1.18k0.06 Sample ml-I (%) pg ml-I No. 2 0.42 k 0.02 0.38 f 0.03 No. 3 0.25 k 0.03 0.3 1 k 0.02 River water No. 1 1.09k0.04 9 4 k 3 1.10k0.06 No. 2 0.63 k 0.04 0.70 k 0.06 No. 3 0.22 k 0.02 - * Mean value of six determinations k standard deviation. and 3.0% for 0.4 and 1.0 pg ml-l of B respectively. The detection limit calculated using the IUPAC recommenda- tion (30) was 0.07 pg ml-l and the determination limit (1 00) was 0.25 pg m1-I.Conclusion The proposed method for the determination of B by the combination of a chemical modifier Ni and a Zr-treated pyrolytic graphite coated graphite tube provides several advantages over the methods reported previously in the literature increased sensitivity no memory effect shorter analysis time and longer lifetime of the graphite tube. The method allows the determination of B in waste water samples. It was shown that it is not necessary to use the standard additions method which could be of interest for the routine analysis of samples of this kind. The authors thank M. Gormann for the revision of the manuscript and the Comisi6n Interministerial de Ciencia y Technologia (CICYT) (Project No. PB 880094) for financial support. References 1 2 3 4 5 6 7 8 9 Barnett N. W. Ebdon L. Evans E. H. and Ollivier P. Anal. Proc. 1988 25 233. Jyan Y. Yao J. and Huang B. FenxiHuuxue 1989,17,456. Norval E. Anal. Chim. Acta 1986 181 169. Van der Geugten R. P. Fresenius Z. Anal. Chem. 198 1,306 13. Goyal A. Patel D. and Sastry M. D. Anal. Chim. Acta 1986,182,225. Kwoles M. B. Broodie K. G. J. Anal. At. Spectrom. 1989,4 305. Runnels J. H. Merryfield R. and Fisher H. B. Anal. Chem. 1975,47 1258. Nater E. A. and Burau R. G. Anal. Chim. Acta 1989 220 83. Ng K. C. and Caruso J. A. Anal. Chim. Acta 1982,143,209. Paper I /0133 7H Received March 20th I991 Accepted August 14th I991

ISSN:0267-9477    

DOI:10.1039/JA9910600669

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 1991 VOL. 6 673 Effect of the Matrix on the Determination of Some Impurities in Europium(ii1) Oxide by Flame and Electrothermal Atomic Absorption Spectrometry Vera Spevackova Karel Kratzer and Ma’ja Cejchanova Department of Nuclear Chemistry Faculty of Nuclear Science and Engineering Czech Technical University 115 19 Prague 1 Brehova 7 Czechoslovakia The possibility of the direct determination of Co Cr Cu Fe Mn Ni Pb and Zn in europium oxide by flame and electrothermal atomic absorption spectrometry was studied. The purification of the europium solution used for the experiments was necessary in order to eliminate possible contaminants. Copper was measured at 327.4 nm because of spectral interference from europium at 324.75 nm.The results show that the direct determination of the elements under study is possible. Keywords Europium; impurity; flame atomic absorption spectrometry; electrothermal atomic absorption spectrometry; spectral interference The determination of Co Cr Cu Fe Mn Ni Pb and Zn in pure europium oxide is necessary in the electronics indus- try where the highest content of these elements has to be lower than 0.005%. Atomic absorption spectrometry (AAS) flame and electrothermal atomization was chosen for the analysis because of its selectivity and sensitivity. The verification of the method and the study of the possible matrix effect on the determination of each element under study was the aim of this work. In preliminary experiments on synthetic solutions an increase in the absorbance signal for Pb Zn and especially Cu after the addition of a europium solution was observed.This might be caused by impurities of Pb Zn and Cu being present in the europium solution used and/or by a matrix effect. In order to eliminate the possible contaminants purification of the europium solution was carried out. Copper Pb and Zn can be extracted from europium solution using diethyldithiocarbamate (DDC). From the published values of extraction constants1** for all the elements under study the optimum experimental condi- tions can be established. The extraction order is Cu>Pb>Zn; the pH value of the extraction should be greater than 4.’ Experimental Apparatus A Varian Techtron 875 spectrometer equipped with a burner or a GTA-95 graphite furnace were used.Measure- ments of pH were made with a Radiometer PHM 62 pH meter. Radioactivity measurements were made with a TESLA NZQ 7 17T and a NaI(T1) wall type crystal. Reagents Standard metal solutions (1 mg ml-l in 2% HN03) were obtained from Aldrich and DDC (0.02 mol dm-3) in butyl acetate from Merck. The following reagents were also used hydrochloric acid (pro analysi); acetate buffer solution (pH 5); europium oxide (99.9%) final solution 100 mg ml-l of Eu in 0.1 mol dm-3 HNO,; and carrier-free radionuclides lszEu and 6sZn. Purification of Europium Oxide In order to check the proposed extraction procedure labelling of the solution by radioisotopes was performed. Owing to the higher values of extraction constants of Cu and Pb compared with Zn all three elements were quantita- tively extracted under the same experimental conditions as for the quantitative extraction of Zn.The following proce- dure was used the pH of the solution under study was established at pH 5 with acetate buffer and 5 ml aliquots containing 0.5 pg ml-l of Zn and 10 mg ml-l of Eu labelled with 6sZn and lS2Eu were shaken with 5 ml of 0.02 mol dm-3 DDC for 3 min. After separation of the phases the measurements of the activities were performed. The extraction yield of Zn was 99.9% and the extraction of Eu was lower than 0.5%. Influence of the Europium Matrix The following model solutions corresponding to the highest limit of impurities in europium oxide (0.005% for each element) were prepared in approximately 0.1 mol dm- HN03 solution A element under study 0.5 pg ml-l; Table 1 Conditions of measurement by flame AAS Wavelength/ Slit-width/ Element nm nm c o Cr c u Fe Mn Ni Pb Zn 240.7 357.9 324.8t 248.3 279.5 232.0 217.0 213.9 0.2 0.2 0.5 0.2 0.2 0.2 1 .o 1 .o Flame Air-C2H2 Air-C2H2 Air-C2H2 Air-C2H2 Air-C2H2 Air-C2H2 Air-C2H2 N20-CZH2 Background corrected Yes No No Yes Yes Yes Yes Yes Detection limit*/ pg ml-* 0.3 0.2 0.1 0.2 0.07 0.3 0.3 0.02 *Detection limit = 30 standard deviations of the blank.tPreliminary experiment.674 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 199 1 VOL. 6 Table 2 Conditions of measurement by electrothermal atomiza- tion Ash Atomization Detection Element temperature/"C temperature/"C limiting co 900 2300 40 Cr 1 000 2600 25 c u 900 2300 20 Fe 800 2300 60 Mn 800 2300 20 Ni 900 2400 50 Pb 350 2000 40 Zn 3 50 1900 5 Table 3 Results of the flame AAS measurement (concentration of each element 0.5 pg ml-l) n=3 FouncUpg ml-l Element Solution A Solution B Solution C Absorbance c o Cr Cut Fe Mn Ni Pb Zn 0.49 f 0.02 0.49 k 0.02 0.5 1 + 0.02 0.49 f 0.02 0.51 f 0.02 0.49 f 0.02 0.50 f 0.02 0.49 k 0.02 0.48 k 0.01 0.49 f 0.02 0.48 f 0.02 0.49 f 0.03 0.5 1 f 0.03 0.5 1 k 0.03 0.5OkO.01 -# ND* ND ND ND ND ND§ NDll -# 0.017 0.014 0.052 0.035 0.060 0.023 0.0 14 0.085 *ND= not detectable.tMeasured at 324.8 nm in preliminary experiments. #Value much higher than expected. §Without purification of Eu solution a value of 0.04 pg ml-l of VWithout purification of Eu solution a value of 0.13 pg ml-l of Pb was obtained. Zn was obtained. Table 4 Influence of Eu on the determination of Cu at other wavelengths Wavelength/ Solution A/ Solution B/ Solution C/ nm pg ml-' pg ml-l pg ml-l 327.4 0.5 1 f 0.02 0.48 k 0.02 0.02 f 0.02 2 17.9 0.52 f 0.03 0.47 f 0.03 0.05 f 0.02 ties of 0.005%.For lower concentrations a furnace tech- nique is necessary. When measuring Cu a spectral interfer- ence of Eu appeared (&,=324.754 nm &,=324.753 nm).6 This interference is not mentioned in frequently used manuals because the ratio of intensities (ICU:IEu= 500 1) is favourable for Cu which in the majority of samples is present in higher concentrations than Eu and therefore in such samples is not of great importance. For this reason other resonance lines for Cu were studied and the line at 327.4 nm was found to be unaffected by Eu and therefore could be used for direct determination.The results are given in Table 4. solution B 0.5 pg ml-l of the element under study + 10 mg ml-1 of Eu; and solution C 10 mg ml-l of Eu. All solutions were measured by flame and electrothermal AAS at the appropriate The measurement conditions are given in Tables 1 and 2. Results and Discussion The results of the measurements using flame AAS are given in the Table 3. From this table it follows that with the exception of Cu all the elements under study could be measured directly in a Eu matrix. The detection limit in flame AAS is sufficient for concentration limits of impuri- Conclusion The possibility of the direct determination of Co Cr Cu Fe Mn Ni Pb and Zn was verified. Preliminary observa- tions of increases in the absorption signals of Pb and Zn were caused by impurities in the synthetic Eu solution and after its purification they disappeared. For Cu there is a spectral interference of Eu at the wavelength 324.8 nm and therefore the line at 327.4 nm was used. References Stary J. and Freiser H. IUPAC-Chelating Agents Part IV Pergamon Press Oxford 1978. Kratzer K. Ph.D. Thesis University of Prague 1967. Koch 0. G. and Dedic G. A. Handbuch der Spurenanalyse Springer Berlin 1974. Manual of Flame Techniques Varian Techtron Springfield IL 1979. Manual of Flameless Techniques Varian Techtron Springfield IL 1982. Welz B. Atomabsorptionspektrometrie Verlag Chemie Wein- heim 1983 p. 312. Paper I I002 74K Received January 21st 1991 Accepted July 3rd 1991

ISSN:0267-9477    

DOI:10.1039/JA9910600673

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 199 1 VOL. 6 675 INTER-LABORATORY NOTE Sulphur Contamination From Synthetic Materials in the Argon Gas Supply in Inductively Coupled Plasma Atomic Emission Spectrometry A. P. M. de Win CFT Chemical Analysis SAC? Nederlandse Philips Bedruven 6. V. Postbus 218 5600 MD Eindhoven The Netherlands In trace analysis the blank values contribute additional 'background' and consequently adversely affect the detec- tion limits. In order to attain adequate suppression of blank signals it is important to know the origin of the contamina- tion. Some workers1v2 have already reported the presence of a blank sulphur signal in their instruments. Different reasons for the cause of the blank signal were suggested but satisfactory proof was not given.A similar situation has been encountered in the inductively coupled plasma atomic emission spectrometry equipment (JY 38 plus) used in this laboratory. The magnitude of the persistent blank signal represents a very large value (0.9 f 0.1 mg 1-l of S). In a systematic approach using subboiled distilled water the blank appeared to originate for the main part from the nylon gas tubing. The nylon resin is plasticized with a sulphur-containing additive probably an organic sulphon- amide. The total sulphur content of the tubing material was 1.7&0.2%. All of the nylon gas tubes were changed to Teflon tubes of the same dimensions. Despite rigorous cleaning of the vital parts in the argon gas line (flow meters argon humidifier nebulizer spray chamber torch) and flushing with argon overnight the blank sulphur signals remained at a low but easily detect- able level of about 0.1 mg 1-' of s.Only after removal of the argon humidifier did the blank drop to the detection limit level. Again some synthetic parts of the humidifier contain- ing 0.7% s were shown to be the cause of the contamina- tion. The signal-to-background ratio of 1 mg 1-l of S at a wavelength of 180.73 nm was 6.0. With a relative standard deviation for the background at this wavelength of 296 the detection limit (30) was 10 pg 1-l of S. The investigations have shown that all of the synthetic materials in the argon gas line which contain appreciable amounts of sulphur can be the potential source of a contamination error. References 1 2 Morita M. Uehiro T. and Fuwa K. Anal. Chirn. Actu 1984 166,283. Attar K. Appl. Spectrosc. 1988 42 1493. Paper 1/04919D Received September 23 I991 Accepted September 30 1991

ISSN:0267-9477    

DOI:10.1039/JA9910600675

出版商:RSC    

年代:1991

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 1991 VOL. 6 677 Abell Ian 145 Abollino O. 119 Alary Josette 647 Ali Abdalla H. 21 1 Andersen Knut-Jan 277 Apte S. C. 169 Arnaud Josiane 647 Arpadjan Sonja 487 Baeten Wilhelmina L. M. 609 Barnes Ramon M. 57 Barry Eugene F. 545 Baxter Douglas C. 109 Beinrohr Ernest 33,307 Belitz Ronald K. 393 Bendicho Carlos 353 Berglund Ingemar 109 Berman Shier S. 19,283 Blades Michael W. 215 Blais Jean-Simon 225 Blue James L. 261 Branch Simon 151,155 Bridenne Martine 49 Brindle Ian D. 129 Brindle Mary E. 129 Brown James A. 393 Butcher David J. 9 Butler L. R. P. 329 Bye Ragnar 389 C h a r a Carmen 669 Canals Antonio 139,573 Carbonell Vincente 233,58 1 Carre Martine 49 Caruso Joseph A 605 Cejchanova Ma’ja 673 Cervera Maria Luisa 379,477 Chen Hengwu 129 Chou Lei 273 Collins C.S. 329 Colbn Luis A. 545 Comber S. D. W. 169 Corns Warren T. 155 Crain J. S. 601 CsCmi Pavol 307 Dawson John B. 93 de Laat Wim J. M. 609 de la Guardia Miguel 233 379 de Loos-Vollebregt 477,581 Margaretha T. C. 165 323 353 de Win A. P. M. 675 Dim de Rodriguez Olga 49 Dittrich Klaus 3 13,465 Mekal Bohumil 661 DoCekalovA Hana 661 DoleZal Jifi 521 Dolinkk Franci 653 D’Ulivo Alessandro 565 Ebdon Les 151 155,421 Elmahadi Hayat A. M. 643 Evans E. Hywel 421,605 Fairman Ben 397 Falk Heinz 631 Fang Zhaolun 179,301 CUMULATIVE AUTHOR INDEX FEBRUARY-DECEMBER 1991 Favier Alain 647 Fell Gordon S. 559 Fernandez Sanchez Maria Luisa Forbes Kimberely A. 57 Ford Mick 15 1 Foulkes Mike 15 1 Franks Jeff 145,591 Frech Wolfgang 109 Fuchs Holger 3 13 Fuge Ronald 445 Furata N.199 Garden Louise M. 159 Gardener M. J. 169 Gervais Lyne S. 41,493 Gilmutdinov Albert Kh. 505 Greenway Gillian M. 643 GUM A. M. 169 Hang Heng-bin 385 Hassell D. Christian 105 Haswell S. J. 339 He Bin 385 Hecking Le T. 637 Hernhndez Cdrdoba Manuel 627 Hemandis Vincente 139,573 Hieftje G. M. 191 Hill Steve 155 Hinds Michael H. 473 HlavAt Robert 52 1 HlavAtek Ivan 535 HlavA&ovA Irena 535 Hoenig Michel 273 Holcombe James A. 105 Houk R. S. 601 Hu Bin 623 Huang Degui 2 15 Huang Min 221 Huyghues-Despointes Alexis 225 Igarashi Yasuhito 205,335 Iida Yasuo 541 Irwin Richard L. 9 Ishii Izumi 3 17 Ishizuka Toshio 541 Ivanov V. P. 505 Jackson Kenneth W. 473 Jiang Zucheng 22 1,623 Julshamm Kaare 277 Karanassios Vassili 457,527 Ketterer Michael E.439 Kibble Helen A. B. 133 Kim Chang-Kyu 205 Kluckner Paul D. 37 Kolihovh Dana 521 Komhrek Josef 661 Koons Robert D. 45 1 Kratzer Karel 673 Kunz Frank W. 393 Lampugnani Leonardo 565 Larsen Erik H. 375 Latimer Kathryn E. 473 Le Xia-chun 129 Ledingham Kenneth W. D. 73 Lee Kwang W. 431 Li Ang 385 397 Li F. H. 457 Littlejohn David 159 Liu B. 457 L6pez Garcia Ignacio 627 Luguera Milagros 669 Lund Walter 389 Luong Van T. 19 L’vov Boris 191 Lyon Thomas D. B. 559 Maage Amund 277 Madrid Yolanda 669 Majidi Vahid 105 Marot Yves 49 Marshall John 145 159,591 Marshall William D. 225 Masuda Kimihiko 335 Matusiewicz Henryk 283 Mauri A. R. 581 McInroy James F. 335 McKay Keith 559 Mentasti E. 119 Mermet Jean-Michel 49 3 13 Michel Robert G. 9 Millward Christopher G. 37 Miyazaki Akira 173 Momplaisir Georges Marie 225 Montaser Akbar 3 17 Montoro Rosa 379,477 Mora Juan 139,573 Morikawa Hisashi 541 Morita Shigemitsu 205 Mukhtar S.339 Navarro Ascensio 477 Ng Kin C. 21 1 Ni Zhe-ming 385,483 Novotni Ivan 66 1 Offley Stephen G. 133 O’Neill Peter 151 155 Park Chang J. 43 1 Parsley David H. 289 Pearce Nicholas J. G. 445 Peng Runzhong 165 Perkins William T. 445 Peters Charles A. 45 1 Peters Michael J. 439 PEtroS Libor 521 Poluzzi Vanes 33 Porta V. 1 19 Prell Laurie J. 25 Piischel Petr 521 Rademeyer Cor J. 329 Radziuk Bernard 465 Rapta Miroslav 33 Rebbert Pamela S. 45 1 Redfield David A. 25 Regnier Pierre 273 Ren J. M. 527 Reszke Edward E. 57 Rivikre Brigitte 3 13 Rowbottom William H. 123 Salin Eric D. 41,457,493,527 Salvador Amparo 233,477 581 Sampson Barry 1 15 Sanz Angel 233 Sanz-Medel Alfmdo 397 Sarzanini C.119 Scheeline Alexander 553 Schwartz Robert S. 637 Scott Roger D. 559 Seare Nichola J. 133 Seki Riki 205 Shiraishi Kunio. 335 Singhal Ravi P. 73 Slavin Walter 191 Sperling Michael 179 295 301 Spevackova Vera 673 Stratis John A. 239 Stupar Janez 653 Sturgeon Ralph E. 19,283 Styris David L. 25 Sychra VAclav 521 Taddia Marco 33 Takahashi Junichi 9 Takaku Yuichi 205 335 Tan Hsiaoming 3 17 Tanaka Gi-ichiro 335 Tao Hiroaki 173 Tiggelman Johan J. 165,323 Tisdale Preston J. 439 Travis John C. 261 Tserovsky Emil 487 Tsuge Akira 541 Tsumura Akito 205 Turk Gregory C. 261 Tye Chris 145 Tyson Julian F. 133,307 Uden Peter C. 57 Uwamino Yoshinori 54 1 van de Weijer Peter 609 Vanhoutte C. N. 323 Vizcaino Martinez Mm’a Jeslis Voloshin A. V. 505 Vrzaj Vinko 653 Vullings Peter J. M. G. 609 Vyskdilovh Olga 52 1 Wang Jiansheng 605 Wang Zuwei 553 Watters Robert L. Jr. 261 Welz Bernhard 179,295,301 Wickstrom Torild 389 Willie Scott N. 19 Winefordner James D. 2 1 1 Winge Royce K. 601 Xu Fu-zheng 385 Yamamoto Masayoshi 205 Yamasaki Shin-ichi 205 Yan Xiu-ping 483 Ybhiiez N. 379 Yin Xuefeng 295,615 Yoshimizu Katsumi 335 Yu Li-Jian 261 Zachariadis George A. 239 Zakharov Yu. A. 505 Zamboni Roberto 565 Zeng Yun’e 221,623 615 627 465,615

ISSN:0267-9477    

DOI:10.1039/JA9910600677

出版商:RSC    

年代:1991

数据来源: RSC