摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 795 Real-time Internal Standardization for Inductively Coupled Plasma Atomic Emission Spectrometry Using a Custom Segmented-array Charge Coupled Device Detector* Jean-Michel Mermet Laboratoire des Sciences Analytiques University of L yon I 69622 Villeurbanne Cedex France Juan C. lvaldi The Perkin-Elmer Corporation 76 1 Main Avenue Norwalk CT 06859-0293 USA The use of real-time internal standardization for inductively coupled plasma atomic emission spectrometry (ICP- AES) was investigated using a new echelle spectrometer equipped with a custom segmented-array charge coupled device detector. The simultaneous data acquisition system is useful for the study of the degree of correlation in the line signals for a multi-element line set.Precision values below 0.1% RSD were readily obtained when the proper conditions were used. Shot noise which cannot be correlated degrades the correlation between the analyte and the reference signal. The instrumentation used permitted the examination of the same spectral line signal of an element in adjacent orders of the echelle grating thus providing a useful means of studying signal correlation and the effect of shot-noise. A signal which is generated from a linear combination of signals coming from a multi-element line set was used as the reference signal for internal standardization. Results with this synthetic signal were compared with the use of a Sc reference line. Precision improvements as large as a factor of three were observed with simultaneous internal standardization when the analytical signal was limited by flicker noise.Improvement factors of -50 were observed when the signal contained a significant drift component. The precision of shot-noise limited signals was not improved by applying simultaneous internal standardization. Methods are described for discriminating between flicker noise shot noise and drift in the analytical signal. Keywords Inductively coupled plasma atomic emission spectrometry; internal standard; segmented-array charge coupled device The precision of the analytical signal is an important figure of merit. Consequently the literature contains exten- sive studies of the precision and noise of the atomic emission signals from the inductively coupled plasma (ICP).1-9 The fluctuations in the analytical signal originating in the ICP and the hardware used to make the measure- ments contribute to the net precision of the signal. With the appropriate operating conditions the noise signatures of the analytical signals from different elements can be highly correlated as a function of time. Thus with simultaneous data acquisition internal standard signals can be used to significantly improve To make the best use of the signals for internal standardization and to obtain the best operating conditions an understanding of the sources of noise the nature of the elemental lines used and a well- characterized data acquisition methodology is necessary. In the work of Myers and Tracy,' the conditions required to obtain significant improvements in performance are described.Some of the conditions relate directly to the spectrometer hardware and detection system used to make the measurements. Three such conditions are the following the signals for both the analyte and the internal standard must be collected at nearly the same instant in time; the observation volume should be the same; and the RSD of the analyte signal should not be shot-noise limited. In this work ICP instrumentation is used which has simultaneous multi-element readout capability and allows easy examina- tion of the degree of shot noise present in the signal. The correlation between two different signals is described using correlation coefficients. * Presented at the 1993 European Winter Conference on Plasma Spectrochemistry Granada Spain January 10- 15 1993. Table 1 Instrumentation and operating conditions Spectrometer Perkin-Elmer Optima 3000 Spray chamber Demountable torch Perkin-Elmer type I1 Scott-type with pumped drain Conespray nebulizer R.f.powerfW Nebulizer argon gas flow rate/l min-I Intermediate argon gas flow rate/l min-l Plasma argon gas flow rate11 min-l Viewing height Solution flow rate/ml min- Conespray design with 150 pm diameter orifice 1100 0.4 0.8 15 12 mm above load coil 2 Cross-flow nebulizer Gem-tipped R.f. power/W 1000 Nebulizer argon gas Intermediate argon gas Solution flow rate/ml min-I 1 Nebulizer argon gas flow ratefl min-I 0.8 Intermediate argon gas flow rate/l min-I 1.2 Heater temperature/"C 140 Cooler temperaturePC 5 flow rate/l min-l 0.95 flow rate/l min-I I .o Ultrasonic nebulizer Cetac 5000 AT Experimental All measurements were made on a Perkin-Elmer Optima 3000 Cchelle spectrometer which has a custom segmented- array charge coupled device detector (SCD).This instru-796 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 mentation has been described in detail previously.lO,I I Operating conditions are summarized in Table 1. The instrument provides all the spectral data in any given measurement (including background signals) simultane- ously. Thus all the pixels on the detectors are commanded simultaneously to begin or end an exposure. Signals recorded on pixels within a subarray or between subarrays are acquired at the same time. The spectra were collected and stored on a magnetic disk for subsequent processing.The signals were produced by pneumatically or ultrasonically nebulizing multi-element standard solutions into the ICP. The concentrations of the elements were in a range such that spectral interferences were not significant at the wavelengths used and specta consisted mainly of simple peaks on a featureless back- ground. To obtain the analytical signal the spectra were background corrected using a single point off the peak. The area under the peak was calculated by summing the data from five pixels encompassing the peak which is equivalent to about four spectrometer bandpasses. This methodology avoids any loss of correlation that would result if any wavelength registration drift was present. In all cases except where specified precision values are based on 20 consecutive replicate readings. In this work the term ‘flicker noise’ is used to describe multiplicative noise in the analytical signal which has a noise magnitude proportional to the magnitude of the signal.Flicker noise originates in the ICP source and is largely the result of the sample introduction process which involves the production and transport of sample aerosol droplets and solvent vapour. For certain element lines two subarrays for the same spectral line but in adjacent orders of the echelle grating were available. These pairs of subarrays are referred to as ‘sister’ subarrays and were used to examine the noise components in the signal. Ideally the fluctuations in the signals from two such subarrays should be very highly correlated if the noise is dominated by flicker noise originating in the ICP source.The presence of shot noise will reduce the degree of correlation present in the signal. A good way to represent the degree of correlation between two time-resolved signals is by considering the two signals to be X- Y pairs applying linear regression and calculating a correlation coefficient. The correlation coefficients were calculated in this manner using commercial spreadsheet software. The concentrations of the elements used for the correlation studies are given in Table 2. A convenient reference signal for use as a compromise internal standard is the artificial construction of a signal X . The signal X is calculated as a linear combination of signals from various elements dominated by flicker noise.Each element signal to be used in the sum is first normalized to the average value of the intensity for that element. Then the signals for the various elements under consideration are summed. Eqn. ( 1) shows the construction of X from a set of n element lines which require a single compromise internal standard signal X = i where i is an index corresponding to the identity of a particular element line r is an index corresponding to the replicate number in a set of measurements X is the compromise reference signal to be used for replicate r and Ii is the analyte signal for element line i and replicate r. Normally the element to be normalized should be excluded Table 2 Concentration of elements for correlation studies with sister subarrays Concentrationlpg ml-I Element Conespray Ultrasonic Ba 2 0.2 c o 10 1 Fe 2 2 Ni 10 1 from X.However for a great number of elements this does not make any significant difference. In the discussion a mathematical expression for the precision of a signal is used to aid in the interpretation of the sources of noise. The precision of the analytical signal can be modelled with a simple equation if a few assump- tions are made. If the signal-to-background ratio (S/B) is large such that the contribution of the background to the relative standard deviation (RSD) of the analyte signal is negligible and the total number of counts registered at the detector are large compared with the readout noise then eqn. (2) which is a simplified analogue of the RSDB (RSD of background) equation described by Boumans,12 can be used to describe the RSD of the analyte when no internal standardization is performed RSDA= J 2 (xA +- l3 (2) ZA7 where RSDA is the relative standard deviation of the analyte signal (referred to in the text as RSD) aA is the flicker noise coefficient which can be interpreted as the flicker noise limited RSD p is the shot noise coefficient zA is the magnitude of the analyte signal in counts s-’ and 7 is the integration time used.With ideal internal standard correction the best cor- rected RSD would correspond to the value from eqn. (2) when aA is assigned a value of 0. If the above assumptions are not true as in the instance of drift then a more complicated expression is required. Results and Discussion Sources of Noise Noise in ICP atomic emission signals can be very highly correlated as shown in Fig.1. In this example the correla- I 99.0 I I 0 10 Replicate No. 20 Fig. 1 Correlation between the time-resolved signals of two Cd lines and two Mg lines when simultaneous data acquisition is used; integration time 10 s. a Cd 214.439 nm; A Cd 226.502 nm; 0 Mg 279.553 nm; and a Mg 280.270 nmJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 106 E 104 .- 5 g 102 0 U .E 100 U 5 9 8 - 9 6 - 94 92 z 797 - - - - - - 108 I r l I 10 Replicate No. 20 Fig. 2 Normalized signal records from two sister subarrays (see text) for Fe at 273.955 nm. The signals show a lack of correlation in the fluctuations because they are shot noise limited at the short integration time of 300 ms and a signal level of around 3000 counts s-' 101.0 1 - 100.5 - E > fn U .- 100.0 - c.' .- '0 2 99.5 - i! b 99.0 - n I 98.5 0 10 20 Replicate No.Fig. 3 Normalized signal records for two sister subarrays (see text) of Ni at 22 I .647 nm. Correlation is high as the signals are flicker noise limited at an integration time of 30 s and a signal level of around 20000 counts s-l tion and the noise amplitude are very similar for the four signals. The signals shown in Fig. 1 were collected at an integration time of 10 s and the Mg and Cd concentrations were 1 and 10 ppm respectively. The RSD after ratioing any two signals is significantly improved compared with the RSDs of the uncorrected signals. The uncorrected RSDs are around 0.3% and the RSDs of the ratio of any two pairs of lines range from 0.05 to 0.09%.Therefore at this concen- tration and integration time enough photons are registered at the detector for these lines such that the magnitude of shot noise in the signal is much smaller than the flicker noise and a significant precision improvement is obtained by dividing by an internal standard signal. The influence of integration time on the degree of correlation in the analyte signals was examined. Integration time is a particularly important parameter because it influences the fraction of shot noise present in the RSD of the analyte signal. At longer integration times the shot noise contribution to the RSD is reduced. Increasing the magnitude of a particular signal at a fixed integration time has the same effect on the fraction of shot noise as an increase in integration time.If any two signals are ex- amined simultanously the fluctuations in these signals with respect t o each other can be classified as being between two 108 106 - 104 - g 102 ; 100 c 98 8 6 96 94 92 EIl .- U U 92 94 96 98 100 102 104 106 108 First Fe signal (%) Fig. 4 Correlation plot of the data for Fe in Fig. 2. The scatter indicates a total lack of correlation between the two Fe signals. The line shown is the ideal line of slope 1 and intercept 0 (r=0.01) 102 I 1 1 I 98 99 100 101 102 First Ni signal (%) Fig. 5 Correlation plot of the data for Ni in Fig. 3. The data points gather near the ideal line of slope 1 and intercept 0 indicating a high degree of correlation (r=0.95) extreme cases perfectly correlated (including anticorrelated cases) and totally uncorrelated signals.Examples approaching these extremes were examined using sister subarrays (see under Experimental for defini- tion). A special version of the conespray nebulizer13 was used with the conditions given in Table 1. The reason for the low nebulizer flow rate was to avoid nebulizer back- pressures higher than 5 bar. Figs. 2 and 3 contain time- resolved signals from two sister subarrays of Fe 273.955 nm and Ni 221.647 nm respectively. In Fig. 2 the data from the two Fe subarrays are normalized to a value of 100% and superimposed on the same time base. The same is done for the Ni data in Fig. 3. It is observed that for Fe at signal magnitudes of about 3000 counts s-l and a short integration time of 300 ms signals are almost completely uncorrelated (Fig.2). In contrast the Ni signals in Fig. 3 display very high correlation in the fluctuations as a function of time. The Ni signals were collected at a higher signal magnitude of 20000 counts s-l and a longer integration time of 30 s. An alternative representation of the same data is given in the form of correlation plots in Figs. 4 and 5 for Fe and Ni respectively. On these plots perfectly correlated signals would fall ideally on a straight line of slope 1 with a 0 intercept and uncorrelated signal data would appear ran- domly scattered. As expected the Fe data are scattered with a coefficient of correlation near 0 and the Ni data are gathered near the ideal line with a correlation coefficient798 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL.8 Table 3 Influence of the integration time on the correlation coefficient for a set of four pairs of sister subarrays covering four analytical lines Integration time/ms Ba 233 nm Co 230 nm Fe 273 nm Ni 221 nm 100 0.39 0.19 0.0 1 0.57 300 0.78 0.09 0.0 1 0.78 1000 0.62 0.13 0.44 0.89 3000 0.78 0.2 1 0.06 0.75 10000 0.79 0.30 0.62 0.88 30000 0.95 0.72 0.92 0.95 near 1. The use of sister subarrays and the expression of the coefficient of correlation are a useful combination to study the sources of noise. The observations for Fe and Ni can be explained as examples of a shot noise limited signal (Fe) and a flicker noise limited signal (Ni). Even though the data collection is simultaneous across the different pixels of the detector the nature of photon shot noise is such that each pixel records a different uncorrelated pattern of noise.The magnitude of shot noise present in any signal increases with signal magnitude and integration time since more photons are registered at the detector. However signals in which the noise is shot-limited have RSDs which are reduced as a function of increasing signal magnitude and integration time as indicated in eqn. (2). In contrast the flicker noise pattern as a function of time should be identical for the same analytical line falling on two sister subarrays. Signals which are flicker noise limited tend to have RSDs that are not very dependent on integration time. Visual inspection of Fig. 3 indicates not only a high correlation between the two Ni signals but also the presence of drift.The RSD of the analyte signal must include the contribution of drift Use of sister subarrays Ni No.1 and Ni No.2 allows the determination of the RSD of shot noise as both flicker noise and drift noise are correlated RSDNi No.lMi NO.^^ = RSDshot(Ni No.*?+ RSDshot(Ni N0.2 (4) As shot noise for Ni No.1 and Ni No.2 are of the same magnitude then Therefore RSDshot(Ni)=( lIv 2 RSDNi No.l/Ni No.2 (6) Results in Fig. 3 lead to the values of 0.54 0.53 and 0.12% for RSDNi No.l RSDNi No,2 and RSDNi No.lmi No.2 respectively. The value of RSDshot~Ni) is deduced from eqn. (6) to be The contribution of drift to the RSD can be eliminated from eqn. (3) by calculating the standard deviation of successive data (SD,,,) RSDshot(Ni) = 0. O 8 6%. SDsuc= l/ i+' 2(n- 1) (7) 10 - A 5 Q 1 - a v) 0.1 - 0.01 1 1 1 1 1 I I 1x10~ 1x10~ 1x10' 1x10~ I X I O ~ 1x10~ 1x10' Total counts Fig.6 RSD of the line signal plotted as a function of the total number of counts registered at the detectors for 0 conespray nebulizer and . ultrasonic nebulizer. The total number of counts is calculated as the product of integration time and intensity in counts s-l. A group of four pairs of lines from sister subarrays were used to generate the plot (see text) 0.49%. The value of RSDflickeflNi) is in the range of observed values for RSDflicker. The major contribution of the noise results from drift. As the same element was used for this experiment high correlation is expected for drift correc- tion. It will be seen later that this is not necessarily true when comparing the behaviour of different elements.The correlation coefficient as a function of integration time was examined for a set of four pairs of sister subarrays covering four analytical lines. The lines used were Ba 233.527 nm Co 230.786 nm Ni 221.647 nm and Fe 273.955 nm. Table 3 contains the correlation coefficients for each pair of sister subarrays at integration times ranging from 0.1 to 30 s. There is a trend in the data showing correlation coefficient values closer to 1 at longer integra- tion times and values closer to 0 at shorter integration times. This trend is expected and is predictable from eqn. (2). Longer integration times increase the total number of photoelectrons registered at the detector thus reducing the relative fraction of shot noise in the RSD and leaving flicker noise as the dominant form of noise. The contributions of flicker and shot noise were exam- ined further using the same group of four pairs of sister subarrays with both the conespray pneumatic nebulizer and the ultrasonic nebulizer.The RSD of the analyte signals was computed and plotted as a function of the total number of counts registered at the detector. Different integration times ranging from 0.1 to 30 s were used to collect the data. The data are shown for both nebulizers in Fig. 6. The trend as a function of the total signal counts is similar for both nebulizers but the flicker noise-limited precision at very high signal counts is about 1% for the ultrasonic nebulizer compared with 0.1-0.2% for the conespray nebulizer. This difference is partly due to low frequently drift in the signal that was observed for the ultrasonic data.The drift might have resulted from insufficient warm-up time of the ultrasonic nebulizer. It can be observed in Fig. 6 that the RSD at smaller values of total counts does not closely follow a shot noise limited trend. This may be due to the fact that the graph is a composite of RSDs taken at different integration times. As reported by Winefordner and Ward I 5 changing the integra- tion time changes the frequencies at which the signal is sampled. Thus a predetermined integration time and number of replicates effectively sets up a bandpass filter through which the signal is processed. The deviation from the expected shot noise trend in Fig. 6 can be explained as the consequence of altering the bandpass filter.The netJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 799 n c! cn U UJ r" Fig. 7 Digitizing interval Plot of a computer simulation of the calculated SD of a synthetic continuous signal with an SD of 100 units after passing through an A/D converter. The calculated or 'measured' SD is plotted as a function of the digitizing interval A SD from a sample of 20 data points; and B SD from a sample of 300 data points effect can be described as the violation of the assumption made under Experimental that the flicker noise limited RSD is independent of integration time. This can occur if the flicker noise is being processed differently through the different bandpass filters. Therefore if the integration time is widely varied then aA in eqn.(2) might not be constant. Flicker noise and shot noise can be described as aperiodic noise. Other sources of aperiodic noise are detector read- out noise dark current shot noise and digitizing noise from the analogue-to-digital (ND) converter. From previous investigations,' it can be concluded that detector readout and dark current noises are negligible since signal levels are far above the detection limit. Digitizing noise is important only when strong and weak signals are measured simultane- ously and the full range of the 16 bit A/D converter is used. For the weakest signals and short integration times SDs of the order of the digitizing interval of the A/D converter might be observed. In order to estimate the influence of digitizing noise on the RSD a simulation was constructed in which a continuous signal with a SD of 100 arbitary units was digitized at various digitizing intervals. The SDs of the digitized signals were then calculated to determine if there was any significant difference from the assigned value of 100.The noise in the signal was simulated to have a Guassian probability distribution. The results of the simu- lation are shown graphically in Fig. 7. It can be observed that at n=20 replicates the calculated SD is not signifi- cantly affected as long as the digitizing interval is less than about 3 of the assigned SD. The SD with n=300 replicates provides a smoother expression of the error function resulting from digitizing error. This information provides a guide for judging when digitizing error significantly influ- ences the precision.Periodic noise is also present in the analytical signal. Line frequency and harmonics (50/60 Hz) along with peristaltic pump noise (typically 2-10 Hz) can be observed in the analytical signal. Usually the integration times used in the measurements are sufficiently long so that influences from periodic noise are not observed. Peristaltic pump pulsations can begin to contribute at integration times shorter than about 500 ms.* Simultaneous Internal Standardization for Precision Improvement A single internal standard signal cannot provide the best possible correction for all element lines being used in an 1 5 n o to K .oo P A 0.01 1x10' 1x10' 1x10' 1x10' 1x10' Signal magnitude/counts s-' Fig.8 RSD of a group of element line signals plotted as a function of signal magnitude 0 uncorrected RSD; . RSD after internal standard correction using the signal X and A RSD after internal standard correction using the Sc 36 1.384 nm signal. A theoretical approximation to the uncorrected RSD; and B theoretical approxi- mation to the shot noise limited RSD. Theoretical curves are based on eqn. (2). analy~is,~,~ but a best compromise can be selected for a given set of analyte lines. For precision improvements it is necessary to use an internal standard signal that has a negligible proportion of shot noise in its RSD and a dominant component of flicker noise.'J Thus the signal should be a relatively strong one with high total signal counts. This is consistent with the data presented under Sources of Noise which indicates that only flicker noise limited signals show correlation in the fluctuations as a function of time.Similarly it has been shownI6 that simultaneous background correction provides improve- ments in detection limits by filtering out correlated flicker noise in the background when the background measure- ments are flicker noise limited. No improvement was obtained with shot noise limited background signals. The element Sc has often been used as an internal standard because it can be obtained in a high purity form has bright lines and is not commonly a major con~tituent.~ The use of Sc as a simultaneous internal standard has been demonstrated to yield considerable improvements in preci- sion.6 A synthetic signal called X (see under Experimental) was generated to investigate the corrections possible with this signal.Since X is a sum of line intensities the proportion of shot noise in X is significantly reduced compared with any one of the component signals. Thus the noise in X represents the average flicker noise behaviour in the group of element lines used in the linear sum. This can be considered to be the best compromise signal for the group of lines. Analytical signals were collected for a group of elements and simultaneous internal standardization was applied to the signals by simply ratioing each analyte line to the internal standard signal. The cross-flow nebulizer was used with conditions specified in Table 1. For this data the integration time z was fixed at 10 s.Both Sc 361 nm and X were used as the internal standard in order to compare the results. Table 4 and Fig. 8 contain the results. Included in Fig. 8 are the approximate theoretical values for the uncorrected and perfectly corrected RSD based on eqn. (2). Values of 0.0016 and 0.067 were used for aA and p respectively. These values were determined using similar procedures to those outlined in previous work with the SIB-RSD detection limit theory.'* It is observed that for most elements there is an improvement in precision with the use of either Xor Sc as800 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 Table 4 Comparison of the experimental %RSD (RSD,,,) the %RSD after internal standardization using Sc RSDk the %RSD after internal standardization using the synthetic element X RSDx the theoretical %RSD deduced from eqn.(2) RSDlheO and the theoretical %RSD deduced from eqn. (2) assuming only shot noise RSDsho1 for several elements; 2 is the magnitude of the signal in counts s-I. Elements are classified by increasing values of 2 Element Wavelengthhm Concentration/ ZAI RSD, RSDk RSDx RSDtheo RSDshot mg I-' counts s-I ( O W (Oh) (W (Oh) (O/O) Pb Pb Pb Mo Mo Ba No. 1 Mg Ca Ba No.2 Cd Mg Zn Zn Cd Cd Ca Mg Zn s c s c s c 216 220 26 1 204 203 233 285 317 233 228 280 206 202 214 226 422 2 79 213 357 36 I 424 50 50 50 50 50 1 1 1 1 50 1 10 10 50 50 1 1 10 1 1 1 52 1 1079 1364 2140 2400 3083 309 1 3925 583 1 6583 7520 791 1 10400 13131 15795 21397 229 I0 472 15 73990 107800 180105 0.66 0.35 0.35 0.22 0.29 0.2 1 0.24 0.26 0.19 0.13 0.17 0.26 0.23 0.18 0.23 0.12 0.17 0.15 0.13 0.18 0.2 1 0.70 0.30 0.42 0.26 0.25 0.24 0.22 0.28 0.16 0.15 0.13 0.16 0.15 0.10 0.08 0.10 0.07 0.08 0.06 0.04 - 0.68 0.27 0.36 0.23 0.18 0.16 0.16 0.19 0.09 0.12 0.08 0.20 0.07 0.07 0.12 0.09 0.08 0.07 0.09 0.1 1 0.15 0.39 0.30 0.27 0.24 0.23 0.22 0.22 0.2 1 0.19 0.19 0.19 0.18 0.18 0.18 0.17 0.17 0.17 0.16 0.16 0.16 0.16 0.36 0.25 0.22 0.18 0.17 0.15 0.15 0.13 0.1 1 0.10 0.09 0.09 0.08 0.07 0.06 0.06 0.05 0.04 0.03 0.02 0.02 Table 5 Comparison of the contribution to signal changes caused by el'fects such as instrument warm-up and drift; energy defect (eV) is the difference between the energy sum of the element and that of Sc 361 nm (9.99 eV) Element Wavelengthhm Energy defectlev RSD RSDSUC RSDk RSDx Zn Zn Pb Cd Cd Be Mo Mg Mg Ba sc Ba 202 206 220 2 14 226 313 204 279 280 233 36 1 455 - 5.52 5.4 1 4.80 4.78 4.48 3.29 3.17 2.09 2.07 1.23 0 -2.06 7.14 7.10 6.35 6.90 6.55 6.14 5.99 6.04 5.94 5.69 5.33 4.35 1.12 1.12 1.1 1 1.10 1.07 1.08 1.1 1 1.04 1.03 1 .oo 0.98 0.83 I .93 1.89 1.1 I 1.68 1.31 0.86 0.75 0.77 0.66 0.45 1.04 - 1.05 1 .oo 0.30 0.79 0.42 0.14 0.46 0.14 0.24 0.5 1 0.90 1.92 the internal standard but the precision after correction does not reach the shot noise limit for most elements.The precision improvement after correction results largely from the reduction of correlated flicker noise which is common to both the analyte and the internal standard. The full potential benefit of internal standardization is not achieved because of loss of correlation and differences in the relative magnitude of the noise amplitude caused by the individual behaviour of the different line signals.As seen in Fig. 8 there is a marginal improvement for many elements with the use of X compared with Sc as the internal standard. Major exceptions to this are the three points at highest signal magnitudes in Fig. 8. These three points are precision data for three Sc lines and it is expected that Sc 36 1 would be the best internal standard for the other two ionic lines of Sc. By using Xas the internal standard precision values of about 0.07°/o were obtained for the best cases. Another example of the usefulness of X is during signal changes caused by effects such as instrument warm-up or a power change. This results in drift and high values of the analyte RSDs (Table 5).The contribution of both the flicker noise and the shot noise is evaluated by calculating the S D of successive data (SD,,,,). The value of RSD, is far below that of RSDA (Table 5) which confirms that the major contribution in the RSD is drift. In the instance of power variation it has been reported14J7-19 that to a first approxi- mation the change in the line intensity is related to the sum of the ionization energy and excitation energy E,, for ionic lines in the 8-15.5 eV energy sum range. This behaviour is summarized in Fig. 9 for four elements that cover the range of energy sum. Therefore a single internal standard will fully compensate only for lines with a similar energy sum. Use of the Sc I1 361 nm line (Es,,=9.99 eV) is not necessarily adequate for an efficient Compensation as its energy sum is in the lower part of the range.It can be expected that the use of X permits a better compensation than the use of Sc and suggests a better compromise element for internal standardization. The improvement in RSD using either Sc or X is given in Fig. 10 as a function of the energy defect i.e. the difference in energy sum between the analyte line and the Sc line (Table 5). It has been verified that the best Sc compensation is observed for lines with similar energy sums. The compensation is rather poor for lines with a high energy sum such as Zn I1 206 nm. Use of X generally provides a better compensation. A large scatter in the improvement is observed resulting from the low RSD values obtained after compensation (Table 5).Inspection ofJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 80 1 1.09 - 1.04 0 13 v) .- u 0.99 .- w - 0 0.94 z 0.89 0.84' 1 I 1 I 1 1 1 I I 1 3 5 7 9 11 13 15 17 19 Replicate No. Fig. 9 Variation of line intensity as a function of replicate No. when a drift resulting from warm-up i.e. a change in the energy transfer is observed average intensity is normalized to 1. Four elements with different sums of ionization energy and excitation energy (see Table 5) are shown A Zn 202 nm; B Pb 220 nm; C Sc 361 nm; and D Ba 455 nm 50 I 1 40 L 30 > g ~ P .E 20 n a v) 10 I I I I I I -2 -1 0 1 2 3 4 5 Energy defectlev Fig. 10 Improvement in the RSD after internal standardization when a drift resulting from a change in the energy transfer is observed A use of Sc as an internal standard; and B use of a synthetic element X as an internal standard Fig.10 suggests that a good compromise could be found for an internal standard with an energy sum of about 12- 1 3 eV. The use of X provides improvement in precision but cannot be used directly to improve accuracy because the magnitude of X is dependent on the component signals present in the sample. Therefore X represents a model signal which can be used to select an optimum compromise internal standard for a sample. Future work will involve continued studies of the performance improvements pos- sible using the simultaneous information which is produced from this instrumentation. Conclusion Simultaneous internal standardization is a useful means for improving precision with ICP-AES.When signals are well correlated precision values of (0.1 O/o can be obtained. The improvements possible with this technique are fundamen- tally bounded by the level of shot noise present in the signals used. The instrumentation used in this work permits easy examination of the level of shot noise in the signals by making use of the simultaneous spectral registration on the CCD detectors and the fact that the same analytical line can appear in different orders of the echelle grating. Multi- element line signals can be linearly combined to produce a signal for use as the reference internal standard signal. This signal has been used to investigate the precision improve- ments which can be obtained in comparison to Sc at the 36 1.384 nm line. On average the artificial signal provides a better compromise improvement compared with the Sc signal. This was also true when there was drift present in the analytical signal. References I Myers S.A. and Tracy D. H. Spectrochim. Acta Part B 1983 38 1227. 2 Belchamber R. M. and Horlick G. Spectrochim. Acta Part B 1982,37 1037. 3 Schmidt G. J. and Slavin W. Anal. Chem. 1982 54 249 1. 4 Lorber A. and Goldbart Z. Anal. Chem. 1984 56 37. 5 Ramsey M. H. and Thompson M. Analyst 1984 109 1625. 6 Marshall J. Rodgers G. and Campbell W.C. J. Anal. At. Spectrom. 1988 3 24 1. 7 Wallace G. F. At. Spectrosc. 1984 5 5. 8 Goudzwaard M. P. and de Loos-Vollebregt M. T. C. Spectrochim. Acta Part B 1990 45 887. 9 Tiggelman J. J. Oukes F. J. and de Loos-Vollebregt M. T. C. Spectrochim. Acta Part B 1990 45 927. 10 Barnard T. W. Crockett M. I. Ivaldi J. C. and Lundberg P. L. Anal. Chem. 1993 65 1225. 1 1 Barnard T. W. Crockett M. I. Ivaldi J. C. Lundberg P. L. Yates D. A. Levine P. A. and Sauer D. J. Anal. Chem. 1993,65 1231. 12 Boumans P. W. J. M. Spectrochim. Acta Part B 1991 46 431. 13 Ivaldi J. C. Vollmer J. and Slavin W. Spectrochim. Acta Part B 1991 46 1063 14 CarrC M. Poussel E. and Mermet J. M. J. Anal. At. Spectrom. 1992 7 79 1. 15 Winefordner J. D. and Ward J. L. Anal. Lett. 1980 13 1293. 16 Ivaldi J. C. and Barnard T. W. Spectrochim. Acta Part B in the press. 17 Mermet J. M. C. R. Acad. Sci. Ser. B 1977 284 319. 18 Ramsey M.H. and Thompson M. Analyst 1985 110 519. 19 Marichy M. Mermet M. and Mermet J.-M. J. Anal. At. Spectrom. 1987 2 56 1. Paper 3/0 1 33 IF Received March 8 1993 Accepted May 12 I993

ISSN:0267-9477    

DOI:10.1039/JA9930800795

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 803 Comparison of Optical Emission Spectrometric Measurements of the Concentration and Energy of Species in Low-pressure Microwave and Radiofrequency Plasma Sources* Jiirgen Ropcke Andreas Oh1 and Martin Schmidt Institute for Low- Temperature Plasma Physics Robert-Blum-Str. 8- 10 D- 17489 Greifswald Germany Understanding of specific plasmachemical reactions occurring in plasma sources with the addition of complicated molecules requires knowledge of the particular plasma conditions. The subject of the present paper is a comparison of the spectrometrically measured relative concentration distributions and energies of species in low-pressure argon discharges containing an organosilicon compound hexamethyldisiloxane (HMDSO) which has a large molecular size.The optical diagnostics were performed with three different plasma sources a special planar microwave plasma source (v=2.45 GHz) an r.f. planar reactor (v=13.56 MHz) and a capacitively coupled r.f. model discharge tube (v=460 kHz). The energy distribution functions of each plasma are not the same but their general forms are similar to a Maxwell-Boltzmann distribution. The results of comprehensive spatially resolved measurements of the relative concentrations of atoms and radicals (H Si CH and C2) the neutral species gas temperature rotational temperature and optical excitation temperature are reported. The use of the same gas mixture in three plasma sources each of distinct construction excited by different frequencies once again clearly indicates that the comparison and interpretation of optical diagnostic results has to be done very carefully taking into consideration the discharge conditions.Results obtained by actinometry show considerably different particle density gradients in the plasmas. Gradients of the excitation temperatures ( Te,,=0.45-0.7 eV) are less but the neutral gas temperatures also exhibit large spatial gradients (T,=ambient temperature about 2000 K). This clearly indicates the absolute necessity of spatially resolved optical emission measurements for the purpose of comparisons between different plasma sources. Keywords Optical emission spectrometry; plasma diagnostics and characterization; low-pressure microwave and r. f. discharges; species concentration and energy The parameter set needed to describe the properties of non- equilibrium low-pressure gas discharge plasmas is very large.Owing to specific excitation mechanisms many different particle ensembles exist. Since energy is trans- ferred step by step from the external electric field to plasma electrons and plasma ions and then to activated neutral species and as the transfer efficiencies of these processes differ by orders of magnitude particle ensembles with different thermodynamic properties exist simultaneously. The energy of single particles of each of these ensembles is distributed within a certain characteristic range. Normally the energy distribution functions are not equal to each other but their general form is similar to a Maxwell-Boltzmann distribution.Therefore following thermodynamic methods each of these particle ensembles can be character- ized by generalized thermodynamic variables such as temperature and the number densities of the particles. In this way it is possible as a first approach to describe the properties of low-pressure non-equilibrium plasmas by a set of temperatures and number densities. When consider- ing systematic investigations of the fundamental properties and application of plasmas knowledge of this set of parameters must be available. As the largest differences exist between the electron ensemble on one side and all other heavier particle ensembles on the other side the parameter set can be reduced to two ensembles in the simplest case. The heavy particle ensemble includes ions as well as radicals and non-activated neutral species.If as is usual the degree of ionization is low this heavy particle ensemble can be described by the properties of the neutral species. This two-temperature model is the simplest ap- proach to characterize low-pressure non-equilibrium plas- mas and is well established in the characterization of pure rare gas discharges. * Presented in part at the 1993 Winter Conference on Plasma Spectrochemistry Granada Spain January 10- 15 1993. The addition of molecules especially of complex mole- cules to rare gas discharges leads to an enormous increase of the number of species ensembles. Nevertheless the differences between electrons and neutral species remain high. Hence the characterization of these plasmas can also begin with the determination of electron temperature neutral species gas temperature and densities of the most relevant species.However many of these molecular gas plasmas which are very relevant in practical applications are chemically active. Under these circumstances a complex investigation of plasmas including probe measurements mass spectrometry and optical methods is difficult and measuring methods which do not disturb the plasma become very important. Optical emission spectrometry (OES) is one of these methods of measurement. It can detect a certain number of different species and the determination of characteristic temperatures is also possible. The subject of the present paper is OES measurements in low-pressure argon discharges containing small additions of an organosilicon compound hexamethyldisiloxane (HMDSO).In some cases the carrier gas also contains a few per cent of helium or nitrogen for measuring purposes. Three different plasma sources are compared a special planar microwave plasma source (v=2.45 GHz) an r.f. planar reactor (v= 13.56 MHz) and a capacitively coupled r.f. model discharge tube (v=460 kHz). The results con- cerning the spatial distribution of hydrogen and silicon atoms and of CH and C2 radicals neutral species gas temperature rotational temperature and optical excita- tion temperature which is closely related to the energy of the electron gas are reported. Knowledge of these para- meters is vital for the detailed analysis of the plasmachemi- cal activity of these plasmas which results in the growth of so called plasma polymer films on the plasma vessel walls and on the electrodes.Although it is known especially with plasmapolymerization of HMDSO that large differences between processes in r.f. excited plasmas and microwave804 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 plasmas exist,' the microphysical explanation of these differences is still a matter of discussion. The latter is a result of the complex character of these plasmas. Hence a complex comparison of these would be interesting. The analysis of the emission measurements includes a number of assumptions which are described as follows. As a matter of principle it is assumed that the Corona model is valid describing the emission of spectral lines or molecular bands corresponding to the population of excited states as determined by the balance between collisional excitation and spontaneous radiative decay only.The validity of this assumption is mainly based on the low degree of ionization in these plasmas which is a maximum of 1 x If the population of the excited atomic states corresponds to a Boltzmann distribution the so called electronic excitation temperature or excitation temperature T, of the emitting atoms can be derived by means of relative line intensity measurements. In this case the value log(Iil/gA) is a linear function of the excitation energy. Using the Boltzmann plot method and plotting log(Iil/gA) as a dependent of the excitation energy the slope of the plot is related to the excitation temperature. It is equal to -0.625/Te when E is in cm-I I is the line intensity ;I the line wavelength A the transition probability and g=gk/gj is the quotient of the statistical weights of the two atomic levels (k>j].In the present paper the excitation tempera- ture is measured by using the argon line intensities. The reason for the use of argon lines is the existence of a large number of good detectable lines in the spectral range available. The transition probabilities were taken from the literature.2 In this instance the argon lines with wavelengths 603.2 nm (transition probability A=0.246 x lo8 s-l) 614.5 nm (A=0.079x lo8 s-l) 667.7 nm (A=0.0241 x lo8 S - I ) and 696.5 nm (A =0.674 x lo8 s-I) were used. The spatially resolved relative concentrations of atoms and molecules are determined by actinometry.This method compares the emission of the species of interest with the emission of for example an inert gas the so called actinomer which is added in small amounts to the discharge gas m i x t ~ r e . ~ The preconditions ensuring validity of actinometric results are as follows. The addition of the actinomer should not essentially influence the discharge characteristics the electron energy distribution has to be spatially constant and the emitting states are solely excited by electron impact upon the ground state. The last condition means that dissociative excitation should be negligible. Often these preconditions are not completely fulfilled therefore the plasmas under investigation are always checked for changes of electron energy distribution.For this purpose the intensity ratio of the H a to the HP line is assumed to be a useful measure of relative changes as has been done previ~usly.~ For the determination of translational or kinetic temper- ature of heavy particles from Doppler broadening of spectral lines the velocity distribution of these particles is assumed to be Maxwellian. This is self-evident since spectral linewidth measurements are only performed for neutral species and since the degree of ionization of the plasmas is low the kinetic temperature is assumed to agree with the neutral gas temperature. The influence of mi- crofield Stark effect is neglected owing to low absolute electron densities not exceeding 1 x 1 0-12 ~ m - ~ . Because in most cases natural linewidths also can be negligible the line profile is pure Gaussian.The full width of half maximum (FWHM) of the Gaussian profile directly represents the Doppler broadening Ad = (8 x ln2 x k/mc2)'12 x Tg112 x Tg112 x Am where T,=neutral gas temperature (K); m=particle mass in all cases. (kg); A,=wavelength (nm) in the middle of the profile; Ic=the Boltzmann constant; and c=the velocity of light (m s - ' ) . ~ The H a line is an extensively studied spectral line. Owing to the fine structure splitting it consists of seven closely related components. The relative intensities are lfound to depend on the discharge conditions. At neutral species gas temperatures lower than about 600 K this fine structure line splitting must be taken into account by deconvolution. In this temperature region the H a line ,appears as a doublet each component being a blend of different fine structure lines.An experimental separation of the two main components of 0.014 nm which only requires a relatively low instrumental error of about 1 x nm is sufficient for determination of the neutral species gas temperature. Another method for the determination of the gas temper- ature used in this work is the measurement of the rotational intensity distribution of excited molecular bands. In order to use this method close correspondence of the rotational state population to the gas temperature must be assumed. The energy separation between the rotational levels of the measured molecules is very small. This makes the rota- tional excitation via interaction with the translation degrees of freedom of the molecules very efficient.Excitation by electron collisions or other processes which can result in non-Boltzmann distributions of the excited rotational levels is therefore neglected. The rotational temperature is derived by measuring the relative intensities of rotational lines within a vibrational band. It is assumed that all rotational lines in the vibra- tional band have the same transition probability. The sensitivity of the apparatus is taken into account. Under these conditions the following relationship can be written Log(I/p)=const- [p(pk 1)B,hc]/2.3026kTro + = R-branch - = P-branch where T, is the rotational temperature B the rotational constant of the upper electronic level I the rotational line intensity p is the rotational quantum number and k the Boltzmann constant.In a Boltzmann plot diagram log(l/p)=flpO,-t l)] the slope of the straight line gives the rotational temperature. In this paper results of rotational temperature measure- ments made by means of nitrogen molecule emission are reported. A small amount of nitrogen is added to the plasma. When considering the nitrogen emission the molecular band of the second positive system C311,(0)-B311,(O) with its bandhead at 337.1 nm is used. To minimize the influence of branch overlapping the R-branch rotational line distribution from quantum numbers greater than 20 is measured. For these large quantum numbers the rotational lines were clearly identified. In this case the deviations from the linear relationship in the Boltzmann plot were less than 5%.Experimental The measurements were performed using a special planar microwave plasma source (v= 2.45 GHz) a capacitively coupled r.f. model discharge tube (v=460 kHz) and an r.f. planar reactor (v= 13.56 MHz). In each case argon was used as the carrier gas. Hexamethyldisiloxane was added in concentrations of 0.2-10%. The optical diagnostic system a schematic diagram of which is shown in Fig. 1 in combination with the microwave plasma source consists of a computer con- trolled 2 m grating spectrometer and a high sensitivity optical multichannel analyser system OMA-VISION (PAR).JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 Ar -+Q I-+} 7 HMDSO {-G > 805 Plasma source Microwave supply Gas handling unit c Mechanical Vacuum vessel translation system I I ' Fig. 1 Schematic diagram of the experimental set-up including the planar microwave plasma source El The 2 m grating spectrometer with quartz collimation optics and a grating with 1302 grooves mm-' achieves a wavelength resolution of about 1 x nm.This allows Doppler line broadening measurements of the red Ha line for determination of neutral species gas temperature in the plasma down to ambient temperature. With respect to the plasma conditions which are characterized by a variety of low intensity atomic lines and molecular bands and by relatively high noise levels lock-in signal detection is used. The computer controlled data collection system used allows further treatment of the stored spectra. Fit and deconvolution procedures are very important but smooth- ing spreading out or averaging procedures were also used.The OMA-VISION system is equipped with a 0.5 m grating monochromator and a Peltier cooled highly sensitive UV intensified charge coupled device (CCD) matrix detector. The maximum sensitivity of this detector is 4-10 photons per count. This system is Molybder Telescope I Opticall window X Vessel wall I used for im boat molecular band analysis for determination of the rotational temperature and for atomic line intensity measurements. Molecular rotational band analysis is performed using a grating with 2400 grooves mrn-l. This provides a wave- length resolution of about 0.1 nm which is sufficient for most cases. Atomic line intensities were measured using a grating with 600 grooves mrn-l.For spatially resolved measurements a quartz telescope and a rectangular slit- like aperture are mounted on a mechanical translation system combined with a quartz light guide cable used for light transfer to the monochromators. The planar microwave plasma source investigated works on the principle of distributed coupling of microwave power into the vaccum vessel. In this way the difficulties concerning the generation of large area microwave plasmas are overcome which arise mainly because of the short wavelength of microwaves and their low penetration depth into the plasma. For a more detailed description of this type of microwave plasma source see ref. 6. Substrate Microwave window ry n\\\\\U - 7 Substrate II Thermocouples holder Fig. 2 Details of the plasma region and substrate arrangement (x distance to the microwave window)806 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL.8 Electrode 4; kHz \$; Ar+He HMDSO Observation Pump - Observation ,-+-+ Fig. 3 Schematic diagram of the r.f. model discharge tube ( I side-on observation position) The test source used generates a long extended laterally homogeneous planar plasma for low-pressure plasma pro- cessing with lateral dimensions of 14 x 4 cm. Sources with up to 50 x 10 cm lateral dimensions have been achieved. In Fig. 2 details of the microwave plasma region and substrate arrangement are shown. The microwaves which penetrate into the reaction vessel through the quartz microwave window are strongly absorbed. Therefore the plasma is strongly decaying in the direction normal to the window.The thickness of the active microwave power absorbing plasma region with electron densities above the so called cut-off density of about 7 x 1 0lo ~ m - ~ does not exceed a few centimeters at low pressures and increasing the pressure compresses this region. The active plasma region is fol- lowed by another decaying diffusive afterglow plasma region. Usually the extended planar substrates for plasma polymerization are placed in this afterglow region. The optical emission of optically thin plasmas was spatially resolved being observed in an end-on configura- tion by means of a special quartz telescope resulting in laterally integrated information. Plasma homogeneity along the line of sight was ensured. Therefore margin effects of the plasma edge regions which significantly influence linewidth measurements can be neglected.Fig. 3 shows a schematic diagram of the r.f. model discharge tube used. To ensure well defined stable dis- charge conditions a gas flow regime with a separate supply 1.6 A (a) I 0 D 1.4 - 1.2 - A 1.0 - 0.8 - 0.6 - .- C '5 0.2 C 1 I - of the monomer downstream of the argon supply is necessary. The r.f. energy (v=460 kHz) with maximum 40 'W power is capacitively coupled to the plasma. The over-all (discharge volume is about 50 cm3. In addition to the end-on observation side-on observation of the spectral emission is also possible. The 13.56 MHz r.f. planar reactor is of the usual type. It consists of two circular electrodes one of which is grounded and the other powered by capacitive coupling.Both electrodes have the same diameter of 13 cm. The distance between the electrodes is 4 cm. The r.f. power is supplied via a matching network. The optical emission is observed side on. Additionally Langmuir probe and mass spectrome- try measurements are possible. The gas supply is organized in a flow regime using a flow controller gas inlet system. Results and Discussion In Fig. 4(a) measurements of the relative spatial concentra- tion distributions of Si atoms H atoms CH radicals and C2 radicals between the microwave quartz window and the substrate in the microwave plasma source are shown with addition of 0.2% HMDSO. The species exhibit pronounced local concentration maxima which are shifted toward the active plasma region near the window.Under these dis- charge conditions atomic line emission of the helium actinomer is undetectable hence an argon line (750.4 nm) is used for actinometric estimation. The low helium emission is believed to be due to the relatively low electron temperature. In contrast between the electrodes of the r.f. 25 (a) 1 1/11 15 l o t 10 0 5 10 15 20 25 xlmm Fig. 4 (a) Dependence of the intensity ratios A C2 516 nm system:Ar 750.4 nm; B CH 430 nm system:Ar 750.4 nm; C Si 288.2 nm:Ar 750.4 nm; and D HmAr 750.4 nm ( x 10) on the distance to the microwave window. (b) Dependence of the intensity ratio Ha:HP on the distance to the microwave window. Ar+0.2% HMDSO+2.5Oh He; p=230 Pa; and substrate position s=25 mm I I 0 10 20 30 40 dlm m Fig. 5 (a) Dependence of the intensity ratio of A C2 516 nm system:He 587.6 nm and B CH 430 nm system:He 587.6 nm on the position between the electrodes d of the r.f.planar reactor. (b) Dependence of the intensity ratio Har:H/? on the position between the electrodes d of the r.f. planar reactor. Ar+ 10% HMDSO+ 5% He p = 70 Pa d=O mm grounded electrodeJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 807 100 50 0 .- c. h r fn c. .- ; o - l l b i l 5 0 100 200 300 400 Ilm m Fig. 6 (a) Dependence of the intensity ratio of A C2 516 nm system:He 587.6 nm ( x loT2); B CH 430 nm system:He 587.6 nm ( x and C Si 288.2 nm He 587.6 nm on the side-on observation position 1 in the r.f. model discharge tube. (b) Dependence of the intensity ratio Ha:HP on the side-on observa- tion position 1 in the r.f.model discharge tube Ar+lO% HMDSO+2% He p=70 Pa planar reactor such a concentration maximum of the CH and C2 molecules is not observable [Fig. 5(a)]. In the r.f. model discharge tube maxima of the relative Si CH and C2 concentrations appear directly downstream from the position where the HMDSO feed gas comes into the tube. For this discharge spatial distributions of the polymer deposition rate were observed. The optical emis- sion maxima [Fig. 6(a)] correlate with the spatial depen- dence of the deposition rate. In all of the plasmas the validity of actinometry was ensured. The ratios of the hydrogen line intensities indicate an almost constant electron temperature over the plasma region investigated [Figs. 4(b) 5(b) and 6(b)]. No order of magnitude changes were observed. One possible explana- tion for this effect is the relatively low level of HMDSO resulting in an argon-hydrogen gas mixture of constant composition as the dominant plasma gas component over the entire discharge volume.The HMDSO is known to produce large amounts of atomic hydrogen under the influence of gas discharges. The ratio of Ha to HP in the microwave plasma source [Fig. 4(b)] is essentially higher 0.6 I-.- 0.5 0.4 I I I I 0 5 10 15 20 25 x/mm Fig. 7 Dependence of the excitation temperature T, on the distance to the microwave window x (Ar+0.2% HMDSO; p=230 Pa; substrate position s=25 mm derived from argon line intensities) I 1 I I I 0 0.25 0.50 0.75 Relative wavelengthIA Fig. 8 Experimental H a line profile fitted by Gaussian profiles than in the r.f.plasma sources [Figs. 5(b) and 6(b)]. This indicates the expected lower electron energy of microwave plasmas. A certain flat maximum between the microwave window and the substrate must be recognized. In accor- dance to the Ha to HP intensity rdtio the spatial distribu- tion of the excitation temperature T,, determined by argon line intensity relationships shows a flat minimum between the microwave window and the substrate (Fig. 7). The tendency to decrease with increasing distance from the microwave window agrees well with former Langmuir probe measurements.' In the vicinity of the substrate the excitation temperature seems to be influenced by this additional wall. In contrast to the excitation temperature the neutral species gas temperature exhibits larger spatial variations.In the r.f. model discharge tube the neutral species gas temperature is derived from Doppler line broadening of the red Ha line (656.3 nm) and rotational band analysis of the nitrogen molecule. The influence of fine structure line splitting on Doppler broadening is taken into account by using numerical fit procedures. This is recommended particularly in the medium temperature range exhibiting incomplete line splitting. So the error in temperature determination is reduced to 5-10%. Fig. 8 shows an example of such an experimental Ha line profile fitted by Gaussian profiles. In Fig. 9 the dependence of the neutral species gas temperature Tg on the supplied power P is shown. The value of Tg increases slightly with increasing power.In contrast to this the analysis of the nitrogen molecular emission resulted in rotational temperatures of 600 550 Y ' 500 A / 450 1 I 0 Ann 1 1 I 1 I 7"" 0 10 20 30 40 50 PMI Fig. 9 Dependence of the neutral gas temperature Tg on the power P in the r.f. model discharge tube (0 Ar 5 5 Pa A Ar 87 Pa; and . Ar+ 10% HMDSO 60 Pa derived from Ha Doppler line broadening)808 iaoo 5 1600 G 1400 JOURNAL OF ANALYTIICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 - - - 2ooo 2000 1500 Y Za 1000 500 - - - - 0 1200 I I 1 1 0 100 200 300 400 Mm m Fig. 10 Dependence of the rotational temperature T, on the distance 1 along the tube (Ar+ 5% N2 p = 54 Pa P=45 W derived from nitrogen band analysis) more than 1000 K. Fig. 10 shows the distribution of these rotational temperatures along the tube.A minimum value of the rotational temperature in the region between the two electrodes can be observed. The phenomenon that nitrogen band analysis leads to too high rotational temperatures has also been reported by other worker^.^.^ In ref. 8 gas temperatures measured by interferometry were compared with rotational temperatures. The latter were considerably higher although rotational Boltzmann distribution was ensured. Deviations from a linear relationship were less than 5%. However it has also to be taken into account that addition of nitrogen can influence the discharge character- istics. Obviously longer relaxation times of the excited nitrogen molecules disturb the local equilibrium between molecular vibration and translation. An accumulation of vibrational energy in the discharge could be caused by differences between excitation and relaxation times.lo In ref. 11 correspondence between rotational and gas tempera- ture was predicted for higher pressures. In Fig. 11 the position dependence of the neutral species gas temperature Tg between the electrodes of the r.f. planar reactor is reported. The value of Tg is derived from Ha line Doppler broadening. A pronounced maximum of the neutral species gas temperature of about 2000 K near the powered electrode is detectable. This temperature distribu- tion is similar to the spatial distribution of the neutral gas temperature in the microwave discharge. The latter also exhibits a local maximum in the region near the microwave window.I2 This similarity is explained by similarities in the absorption energy.In both plasmas most of the energy transfer from the electric field to the plasma is localized near one of the walls and the electrodes respectively. Only the conditions of localization and the mechanisms of energy I 0 1 0 10 20 30 40 dlm rn Fig. 11 Dependence of the neutral gas temperature T on the position d between the electrodes of the r.f. planar reactor (Ar+lO% HMDSO+5% He p = 7 0 Pa d=O mm grounded electrode derived from H a line Doppler broadening) transfer are different. In the r.f. plasma source the plasma sheath which transfers most of the energy is localized at the powered electrode. This is due to capacitative coupling of this electrode. In the microwave discharge the localiza- tion of power absorption is due to the low penetration depth of microwaves into the plasma.A comparison of all measurements especially the re- ported considerable spatial gradients of neutral species gas temperature measurements as well as particle density measurements clearly demonstrates the complexity of the problem under investigation. If different optical tempera- ture measuring methods are used to obtain information concerning the energy of species in the plasmas possible error sources have to be taken into account. The validity of the measured excitation temperatures depends on the uncertainty of the transition probabilities used which can be estimated to be typically 20-30O/0.~~ In contrast the error in neutral gas temperature determination by Ha line broadening is about 5- 1 O% and dissociative excitation can be neglected.The error in the rotational temperature measurements is also about this value. The rotational temperatures of nitrogen molecules exhibit strong deviations in comparison with the neutral species gas temperature. The addition of nitrogen disturbs the plasmas being considered. Therefore measurements of the rotational temperature of hydrogen molecules will be more useful for comparison purposes in further investigations. It is clear that comprehensive complex emission spectro- metric measurements are not sufficient to compare r.f. and microwave plasma sources satisfactorily. In particular the absolute necessity of spatially resolved measurements has been demonstrated i.e. at present only a comparison of locally and energetically well defined plasma source volume elements seems to be promising.The authors would like to thank Urte Kellner and Dietmar Gott for carrying out measurements and for technical assistance and Mathias Schaller for software development. This work was supported by BMFT (Federal Ministry of Research and Technology) of Germany Contract No. 13N5939. 1 2 3 4 5 6 7 8 9 10 1 1 12 13 References Wertheimer M. R. and Moisan M. J. Vac. Sci. Technol. 1985 A3 2643. Mermet J.-M. in Inductively Coupled Plasma Emission Spectroscopy Part 11 Applications and Fundamentals ed. Boumans P. W. J. M. Wiley-Interscience New York 1987. Coburn J. W. and Chen M. J. Vac. Sci. Technol. 198 1,18,353. Lopata E. and Countrywood J. J. Vac. Sci. Technol. 1988 A6 2949. Griem H. R. Plasma Spectroscopy McGraw-Hill New York 1964. Ohl A. in Microwave Discharges Fundamentals and Applica- tions ed. Ferreira C. M. NATO AS1 Series Plenum New York in the press. Ropcke J. and Ohl A. Contrib. Plasma Phys. submitted for publication. Krysmanski K. H. and Walter W. Beitr. Plasmaphys. 1978 18 367. Marawi I. Bielski B. A. Caruso J. A. and Meeks F. R. J. Anal. At. Spectrom. 1992 7 899. Wojaczek K. Beitr. Plasmaphys. 1984 24 551. Neumann W. Ergebnisse der Plasmaphysik und Gaselektro- nik eds. Rompe R. and Steenbeck M. Akademie-Verlag Berlin 1970. Ropcke J. and Ohl A. Contrib. Plasma Phys. 199 1 31 669. Herzberg R. G. Jaffe S. M. Larjo J. Saari J. and Vattulainen J. ISPC- 10 Bochum Germany 199 1 conference Paper 3 /O I 06 OK abstract 1.2-6. Received February 22 1993 Accepted May 4 1993

ISSN:0267-9477    

DOI:10.1039/JA9930800803

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 809 Simultaneous Multi-element Determination Using Helium or Argon Plasma for Graphite Furnace Capacitively Coupled Plasma Atomic Emission Spectrometry* Glen F. R. Gilchrist Peter M. Celliers Huacheng Yang Changbin Yu and Dong C. Liang Aurora Instruments Ltd. 191 West 6th Avenue Vancouver British Columbia Canada V5Y 1K3 Furnace atomization plasma excitation spectrometry (FAPES) has been heralded as a promising new technique for simultaneous multi-element determinations at sub-ppb concentrations. Graphite furnace capacitively coupled plasma atomic emission spectrometry (GFCCP-AES) is a specific configuration of FAPES technology. Reported here is the first application of GFCCP-AES to simultaneous multi-element ultratrace determinations.Comparison of analytical sensitivity and signal characteristics are made for Ar and He plasmas in GFCCP-AES. The emission pulses for Cd Zn TI Pb Ag Sb Au and Sn were higher and broader in the Ar plasma than in the He plasma. The noise levels were approximately equal for both plasmas hence the signal-to-noise ratio was superior for the Ar plasma. Mechanisms of vaporization atomization and excitation are discussed and compared with those proposed by other workers. Keywords Plasma furnace emission; furnace atomization plasma excitation spectrometry; graphite furnace capacitively coupled plasma; inductively coupled plasma atomic emission spectrometry; graphite furnace atomic absorption spectrometry Low pressure gas discharges in hollow cathodel and hollow anode2*' sources have been used for atomic emission spectrometry (AES). It has been ~hownl-~ that electrother- mal atomization coupled with plasma excitation spectro- scopy can provide detection limits comparable to graphite furnace atomic absorption spectrometry (GFAAS).The requirement for low pressure discharge hence evacuation of the furnace work head after loading limits the applica- tion of furnace atomization non-thermal excitation spectro- scopy (FANES) in practical analytical chemistry. Liang and co-workers4~s and Sturgeon and co-worker~.~.~ have reported the development and merits of furnace atomiza- tion plasma excitation spectrometry (FAPES) which is capable of operating at atmospheric pressure. Graphite furnace capacitively coupled plasma atomic emission spec- trometry (GFCCP-AES) is a configuration of FAPES technology for practical analytical chemistry.All of the workers cited here have recognized that the plasma furnace is a promising source for simultaneous multi-element analysis with detection limits in the picogram range for most metals and non-metals. The greater sensitivity of GFCCP-AES over inductively coupled plasma atomic emission spectrometry (ICP-AES) is due to the high atomization efficiency of GFCCP-AES the inefficiency of nebulization for sample introduction in ICP-AES and the longer residence time of excited species within the analysis volume 250-1000 ms for GFCCP versus < 1 ms for ICP- AES. An advantage of GFCCP-AES over GFAAS is that the high excitation efficiency of the r.f.plasma allows simulta- neous determination of elements at the detection limits of GFAAS.l-lO Another advantage is that non-metals some of which are difficult to analyse using GFAAS can be analysed using GFCCP-AES. Sturgeon et al.' and Banks et a[.* have reported absolute detection limits of 94 and 29 pg respectively for P. Huang and blade^,^ using an atmospheric pressure parallel plate CCP as a gas chroma- tography detector have determined that the limits of detection for F C1 Br I C N and 0 range from 10 to 78 pg SKI. The emission spectrum created within the furnace removes the need to provide a hollow cathode lamp or electrodeless discharge lamp as the source for absorption spectrometry. * Presented at the 1993 European Winter Conference on Plasma Spectrochemistry Granada Spain January 10-1 5 1993.Examination of the fundamental characteristics of the GFCCP discharge along with pragmatic studies of vapori- zation atomization and excitation mechanisms matrix effects chemical modification and practical applications are essential to realize the full potential of this technique. One of the challenges facing plasma furnace spectroscopists is to explain the observed discrepancies in the operation of He and Ar plasmas. Difficulty in igniting and maintaining Ar plasmas have been reported,1° yet Ar has a lower first ionization energy than He (15.75 ev and 24.59 eV respectively) and therefore it should be easier to ignite an r.f. plasma in Ar than in He. Hettipathirana and Blades" have calculated that the electron ionization rate (the number of ionizing collisions per cm per Torr of an electron) is 120 times higher in He than in Ar which might be the reason why He plasmas are easier to ignite.Other differences in the operation of Ar and He plasmas have been noted. The emission of C I CN and C2 observed in Ar but not in He is thought to arise from sputtering of the centre electode by Ar. l o Sturgeon et allo have reported that their Ar plasma was not sufficiently energetic to excite CO+ (19.39 and 19.66 eV) as was observed in their He plasma but Ar gave better sensitivity for Ag Cu Mn Pb Ni and Fe. Hettipathirana and Blades12 have used emission and absorption to examine the temporal characteristics of Ag Pb and Mn and to elucidate some of the vaporization and atomization mechanisms in the plasma furnace.They have shown that the centre electrode if it is not heated by the plasma can act as a second surface for condensation and re-vaporization. They have also shown that the plasma shifts the appearance temperature and the peak tempera- ture of Pb to lower values. Determination of elements in real samples may require improved sensitivity and reduced matrix interferences which can be obtained by using matrix modification platform atomization and isothermal atomization. Stur- geon et al.13 have obtained 3-10-fold improvement in detection limit and 6-1 7-fold improvement in sensitivity over wall atomization for Cd Pb and Ag by using a platform and Pd modification. They have also reported 100% recovery of Pb with up to 24 pg of NaCl added to the plasma furnace.For wall atomization and no modifier Gilchrist and Liang14 have reported 93 and 55% recovery of T1 with 0.08 and 1.2 pg of NaCl respectively added to the furnace. Hettipathirana and BladesI5 have compared810 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 matrix interferences from NaCl and NaN03 on the emission signal of Pb and Ag. Sturgeon et aLL6 have used plasma furnace emission to determine Cd and Pb in sediment and biota. The objectives of this work are to show that GFCCP-AES can be used for the simultaneous ultratrace determination of Cd Zn T1 Pb Ag Sb Au and Sn to demonstrate that stable plasmas can be ignited and maintained in Ar and to compare the signals obtained from Ar and He plasmas during multi-element analysis.Experimental Instrumentation An A1 7000 (Aurora Instruments Vancouver Canada) simultaneous multi-element plasma furnace emission spec- trometer was used for simultaneous determinations. An A1 2000 (Aurora Instruments) spectrometer was used for atomic absorption and single element atomic emission measurements. The A1 7000 and A1 2000 are unique new instruments so discussion of some of their noteworthy features is warranted here. Both spectrometers use a state- of-the art transversely heated pyrolytic graphite coated integrated contact graphite furnace that provides temporal and spatial heating profiles that are nearly isothermal. The furnace is 19 mm long with a 5.8 mm i.d. The furnace acts as a counter electrode to a 1 mm diameter pyrolytic graphite coated graphite rod positioned along the furnace central axis.An Ar or He atmospheric pressure plasma is generated by applying high voltage r.f. power to the central electrode while maintaining the furnace at ground poten- tial. The entrance optics of the A1 7000 image a region of the plasma that is 2 mm high and 0.025 mm wide. Normally the furnace is aligned so that the bright region of the plasma adjacent to the centre electrode is focused onto the entrance slit. Using this imaging system no emission signal from the heated furnace centre electrode or plasma species is observed. Blank firings and background signals plotted with the vertical scales used in Figs. 1-4 yield straight horizontal lines. The A1 7000 has a 0.75 m Rowland Circle configuration with 0.02 nm spectral band- pass.Background correction is carried out by simultane- ously monitoring the background emission 0.08 nm higher and lower than the analyte line. The data acquisition system operating at 300 Hz is fast enough to provide 100-400 signal and background measurements for each element depending on the emission pulse width of that element. The emission lines used in this work in nm were Cd 228.80 Zn 2 13.86 T1276.79 Pb 283.3 1 Ag 328.07 Sb 206.83 Au 242.80 and Sn 286.33. An r.f. oscillator and power supply have been developed to provide reliable ignition and stable running of the Ar plasma. Optimization of the oscillator for the He plasma performed by adjusting the impedance matching between the oscillator and the plasma furnace causes enhancement of the emission signals in He however sensitivity and detection limits are still inferior to those in the Ar plasma reported here.The Ar plasma could not be made to ignite and run reliably in the He-optimized oscillator. A 27.12 MHz free running oscillator was optimized to ignite and maintain an Ar plasma and used throughout this work. Temperature measurements were made using a built-in IR pyrometer focused onto the outside of the furnace directly below the dosing hole. The built-in pyrometer was calibrated against an Ircon Model 1100 IR optical pyro- meter (Ircon Niles IL USA) focused through the dosing hole onto the inside of the furnace directly below the dosing hole. Procedures Drying pyrolysis and atomization temperatures were tested to obtain conditions that gave a reasonable response for all eight elements.All elements had peak heights not more than 1 5% lower than for their individually optimized conditions except Sn which had a peak height about 50% below its optimum value. All samples were dried by heating at 90 "C for 5 s then raising the temperature at a rate of 1 K s-l for 25 s. The charring temperature was 400 "C and the furnace heating rate was 3500 K s-l. The atomization temperature was 2100 "C unless stated otherwise. Furnace cut-off experiments to investigate analyte vaporization processes were conducted as follows He or Ar plasma gas was chosen and the furnace was loaded with 20 pl of 25 ppb of Cd Zn Ti Pb Sb Au and Sn and 0.05 ppb of Ag multi-element test solution. The drying and pyrolysis steps were run in the normal manner but during the atomization cycle the furnace power was cut off after 400 ms by which time the furnace wall was heated to 1800 "C.When the furnace reached room temperature the centre electrode was removed and the furnace was cleaned at 2400 "C for 2 s. The centre electrode was then returned to its place and a normal atomization cycle was run with the dry furnace. Reagents and Materials Stock solutions containing 10 000 pg ml-I were prepared separately by dissolving the appropriate mass of pure metal (Cd Zn Pb Sb Au and Sn 99.999%) or pure metal oxide (T1 99.99%) in 50% aqua regia (HC1-HNO3 3+ 1). These solutions were transferred into calibrated flasks and made up to volume so that the final solution contained 5% aqua regia. The Ag stock solution was prepared in a similar way except that nitric acid was used instead of aqua regia.Multi-element test solutions containing Cd Zn T1 Pb Sb Ag Au and Sn were prepared daily by serial dilution of single-element stock solutions. All test solutions were made up to contain 1% v/v nitric acid and stored in plastic containers. Test solutions of 20 pl in volume were manually injected into the furnace. Results and Discussion Analytical Performance In GFCCP-AES like GFAAS discrete samples are placed into the atomizer then dried and thermally treated prior to atomization. Thermal pre-treatment is used to remove much of the solvent and matrix and in some cases to decompose the sample. During the atomization step the analyte is vaporized atomized and excited in the r.f. plasma then the analyte vapour is removed from the analytical volume by diffusion.Convolution of the supply function and the removal function produces an emission signal pulse. Figs. 1-4 show the transient emission signals of 50 ppb of Cd T1 Pb and Au respectively collected simultaneously with Zn Ag Sb and Sn (not shown here). Cadmium T1 Pb and Au were chosen for presentation because they cover a broad range of volatilities and are representative of the emission pulse shapes observed for the eight elements studied. The emission pulses of the eight elements studied were between 3.8 and 8.7-fold higher in the Ar plasma than in the He plasma. The relative performance of the Ar and He plasmas changes with a change in impedance matching between the oscillator and the plasma furnace but a 2-4-fold higher sensitivity was always observed with the Ar plasma.Table 1 gives the mean peak height values of the emission signal obtained for replicate injections of the eight-element test solution under the compromise conditions detailed earlier.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993. VOL. 8 A Time/ms Fig. 1 Emission pulses for 50 ppb of Cd test solution collected during eight simultaneous determinations in A Ar and B He plasmas 600 c. .- C 2 500 2 e c. .- 3 400 > v) c .- 9 300 .- C .- $ 200 E .- W 100 0 r I 500 1000 1500 2 Time/ms Fig. 2 Emission pulses for 50 ppb of T1 test solution collected during eight simultaneous determinations in A Ar and B He plasmas Ti me/ms Fig. 3 Emission pulses for 50 ppb of Pb test solution collected during eight simultaneous determinations in A Ar and B He p 1 as m a s 2500 I A fn J .g 2000 - ip 2 c >.cn e .- 0 1000 2000 3000 Time/ms 81 1 1 Fig. 4 Emission pulses for 50 ppb of Au test solution collected during eight simultaneous determinations in A Ar and B He plasmas Table 1 Peak height values'J for 20 pl of multi-element test solution containing 50 ppb of Cd Zn T1 Pb Sb Au and Sn and 0.1 ppb of Ag. Values are the mean k the standard deviation of at least five replicate determinations and are given for peak two of emission pulses having two peaks Element Ar plasma He plasma Cd 1100k40 289k 1 1 Zn 813k 11 1402 19 T1 630 k 32 125k 12 84k 12 Pb 507 k 44 Ag 660k 55 102k7 Sb 324 k 32 53+ 13 Au 2260k 110 530 k 45 Sn 156+ 12 18k4 Figs. 1-4 and Table 1 show that for the conditions reported here the Ar plasma provides emission signals for Cd Zn T1 Pb Sb Ag Au and Sn that are both higher and broader than those in the He plasma.The complete mechanism that causes this difference in emission signal is not yet fully understood. The enhanced signal in the Ar plasma is probably due in part to slower diffusion of the analyte atoms through Ar. However attenuation of analyte diffusion by the massive Ar atoms and ions is not sufficient to account for the whole extent of the signal enhancement. Differences in the nature of the Ar and He plasmas especially differences in electron mobilities and the number of collisions per electron probably cause some of the observed differences in sensitivities for the analyte atoms in the different plasmas.Hettipathirana and Blades'' have determined that the electron ionization rate is 120 times greater in He than in Ar which probably accounts for the easier ignition and maintenance of He plasmas. Ionizing collisions are how- ever much more energetic than the collisions that excite emission from the analyte atoms studied here. Excitation energies for these elements varied from 3.27 eV for the Ag 328.07 nm line to 5.97 eV for the Sb 206.83 nm line. A possible reason for the improved sensitivity in Ar compared with He might be found in differences in the elastic electron-neutral collision cross-sections for lower energy electrons especially in the energy range of 3-6 eV which is the most important energy range for excitation. The bulk of the plasma electrons have energies in the thermal range812 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL.8 around 0-3.0 eV. It is interesting to note that the collision cross-sections for these low energy electrons are aHe=6 x cm2 for He and Ar re~pectively.1~ This Ar cross-section is remarkably low about an order of magnitude less than in He. In the energy range of 0-5 eV the He cross-section decreases monotoni- cally while the Ar cross-section increases to a maximum value of around 4 eV which is nearly an order of magnitude larger than that for He. At 4 eV electron energy aH,=2.3x 10-{6 cm2 and aA,=6Ox 10-16~m2.17Thus in Ar the collision rate of electrons in the 4-6 eV range is nearly two orders of magnitude higher than in the thermal range of 0.0-3.0 eV.It is possible that this strong energy dependence of the elastic collision cross-section in Ar can lead to an enhancement of the electron energy distribution function in this energy range. Electrons in the thermal energy range (0.0-3.0 eV) will tend to absorb energy from the r.f. field through collisional absorption and thereby diffuse out to higher energies. However in the higher energy ranges (4-6 eV) the much stronger collision cross-section will tend to trap electrons in this energy range until they lose energy through inelastic processes (e.g. atomic excitation). Further study of this hypothesis requires detailed kinetic studies of the electron velocity distribution function including the effects of the r.f. field and the energy dependence of the collision cross-sections.cm2 and aA,=0.6 x Features of the Emission Pulses Fig. 1 shows that for the Ar plasma the Cd emission pulse has a shoulder on the trailing edge and for the He plasma the emission pulse is smaller and has only one peak. In both Ar and He the appearance time of the Cd pulse is virtually zero due to the 400 "C pyrolysis temperature and the high heating rate of the furnace. Fig. 2 shows that for the Ar plasma the T1 emission pulse has two peaks the first peak appearing at the onset of atomization while the second peak appears at about 500 ms. In the He plasma the T1 emission pulse has only one peak but it might be two peaks that are not resolved in time. Fig. 3 shows that in Ar Pb like T1 has two peaks the first peak appearing at the onset of atomization while the second peak appears at about 600 ms.In He the Pb emission pulse has a shoulder on the rising edge indicating that two peaks are present but not resolved in time. Fig. 4 shows that in Ar Au like T1 and Pb has two peaks the first occurring at the onset of the atomization step and the second appearing 1200 ms later. In He two peaks are also seen for Au the first appearing almost immediately and the second appearing at 700 ms. In the Ar plasma Zn TI Pb Ag Sb and Au have two resolved peaks and Cd has two unresolved peaks as shown by the shoulder on the trailing edge of the pulse. In the He plasma two peaks are observed only for Pb Sb and Au. The single peaks observed in the He plasma for Zn T1 Ag and Sn appear at least 100 ms after the start of the atomization cycle (Zn) and the appearance time increases in the same manner as does the appearance time of peak 2 for these elements in the Ar plasma.The above observations suggest that analyte elements have two vaporization events one from the furnace wall and one from the centre electrode. For Zn T1 Ag and Sn in the He plasma no emission pulse is observed during the initial vaporization from the tube wall because these elements are not sufficiently atomized or excited or the peaks are not resolved as with Cd in the Ar plasma. Fig. 5 shows the appearance times for the emission pulses of the eight elements studied in the Ar plasma (circles) and in the He plasma (squares). For the reasons given above the single peak observed for Cd Zn T1 Ag and Sn in the He plasma has been treated as peak 2.When double peaks are observed the first peak (open symbols) appears almost instantaneously (< 150 ms) while the 1500 0 0 0 . m * 500 m p;::. 0 Cd Zn TI Pb Ag Sb Au Sn Fig. 5 Appearance time of the emission pulses collected during eight simultaneous determinations in the Ar plasma (0 peak 1 ; 0 peak 2) and the He plasma (0 peak 1; H peak 2); for emission pulses with one peak that peak is given as peak 2 second peaks (filled symbols) appear at later times (up to 1400 ms). The appearance time of the second peaks for both Ar and He plasmas increases with a decrease in the volatility of the elements. The second peaks in the He plasma (filled squares) have appearance times 150-700 ms earlier than those in Ar plasma (filled circles).Hettipathirana and BladesI2 have used atomic absorption to determine that gas-phase analyte species initially deposi- ted on the furnace wall condense on the centre electrode and are re-vaporized and atomized when the electrode is heated by radiation from the furnace wall. This phenome- non was observed here using an A1 2000 atomic spectro- meter which has the same furnace and r.f. power supply as the A1 7000 but uses a Czerny-Turner monochromator and can be configured for absorption or emission. Hettipathi- rana and BladesI2 have reported that the plasma heated the centre electrode preventing condensation of analyte so no second peak was observed in the emission mode. For this work neither plasma heated the centre electrode and no analyte loss was observed when the plasma was on for 30 s at a furnace temperature of 400 "C.Adjusting the impedance matching between the oscillator and the plasma furnace can lead to heating of the centre electrode even when the furnace is held at room temperature. When this is done only one peak is observed. With no heating by the plasma the centre electrode may have enhanced heating by conduction in He compared with Ar since the thermal conductivity of He is 9-10 times greater than that of Ar. Thus in He earlier vaporization and atomization of the analyte condensed on the centre electrode is observed. The utility of the centre electrode as a second surface for condensation and re-vaporization is demonstrated by a two step pyrolysis programme which can be used to eliminate the early small peak from emission pulses with two peaks.The temperature for the first pyrolysis step is chosen so that all of the analyte stays in the condensed phase on the graphite tube wall. The temperature of the second step is chosen so that the analyte is vaporized and condenses on the centre electrode but is not lost from the tube. For example Pb can be pyrolysed at 550 "C in the first step then at 900 "C for 5 s in the second step without loss from the furnace. A two step pyrolysis programme was not used in this work because Ag Au and Sn could not be transferred to the centre electrode without causing Cd to be lost from the furnace prior to atomization. The observed features of the emission pulses are consis- tent with a mechanism in which the analyte is vaporized from the tube wall and then condenses on the centre electrode.When the centre electrode is heated by radiationJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 813 800 600 - UJ U .- 5 400 t E $ 200 (D >. - U .- s o A 0 . 0 . - *-& 0 500 1000 1500 2000 Time/ms Fig. 6 Results of temperature cut-off experiments for Pb in (a) He plasma and (b) Ar plasma A first firing; and B second firing m C .; 600 v) E 400 W 200 .- L I 0 500 1000 1500 2000 Time/ms Fig. 7 Results of temperature cut-off experiments for Au in (a) the He plasma and (b) the Air plasma A first firing; and B second firing from the tube wall and by conduction from the gas phase18 the analyte is re-vaporized atomized and excited by the plasma. This distillation of solvent matrix and analyte may prove to be beneficial in removing matrix interferences.Gilchrist and LiangI4 obtained 93% recovery of T1 in a 4 ppm NaCl matrix by pyrolysing at 600 "C and 55% recovery of T1 in a 60 ppm NaCl matrix by pyrolysing at 500 "C; in GFAAS without platform or modifier all the T1 would be lost from the furnace under the conditions described above.I9 Direct proof of analyte condensation on the centre electrode is presented below. Cut-off Experiments As detailed under Experimental partial heating of the furnace removal of the centre electrode purging the furnace and then replacing the electrode and re-firing allows one to determine if any analyte has condensed on the centre electrode. Figs. 6 and 7 show the results for Pb and Au respectively obtained during multi-element cut-off experi- ments run using Ar and He plasmas.Fig. 6(a) shows that for the He plasma Pb is completely vaporized and removed from the furnace during the first firing (solid line) because no emission pulse is observed during the second firing (broken line). In Fig. 6(a) a small ragged peak can be seen at the beginning of the atomization step. This peak is probably due to Pb vaporizing from the wall as molecular species (PbO PbC1220.21) partially dissociating in the gas phase then condensing on the centre electrode. Fig. 6(b) shows that for the Ar plasma some of the Pb vaporized during the first firing condenses on the centre electrode and produces an emission pulse during the second firing (broken line). Fig. 7(a) shows that for the He plasma some Au is atomized and excited during the first firing condenses on the centre electrode and produces an emission pulse appearing at 1000 ms during the second firing.Fig. 7(b) shows that the same phenomenon occurs in the Ar plasma except that the emission pulses are larger due to the greater sensitivity of Au in Ar. Signals for the remaining six elements were also collected during the cut-off experiments. For the Ar plasma Cd and T1 were completely vaporized during the first step. Emission signals for Zn Pb Sb Ag and Au were observed during both steps indicating that they were vaporized during the first firing condensed on the centre electrode and were then vaporized atomized and excited during the second firing. No emission signal was observed for Sn during the first firing but an Sn emission pulse was observed during the second firing.These observa- tions are consistent with Sn being vaporized as molecular species (SnO SnC1,20,21) during the first firing but hardly dissociating so that excited atoms are not observed. The SnO or SnCl condensed on the centre electrode might be vaporized atomized and excited during the second firing. For the He plasma Cd Zn T1 Pb and Sb were completely vaporized in the first step as shown by their emission signals in the first step and no emission signal in the second step. This is consistent with greater heating of the centre electrode by conduction in He than in Ar thereby reducing or eliminating condensation of analyte on the centre electrode. Emission signals for Ag and Au were observed during both steps indicating that Ag and Au were atomized excited and condensed on the centre electrode during the first firing.The Ag and Au were then re- vaporized from the centre electrode atomized and excited during the second firing. The absence of an Sn emission signal during either step indicates that Sn was vaporized as a molcular species (SnO SnC1220,21) and removed from the furnace without being atomized and excited sufficiently to be observed. Atomization Curves Fig. 5 shows that vaporization and atomization from the centre electrode occur earlier in a He plasma than in an Ar plasma. Fig. 8 shows the effect of atomization temperature on the emission pulses of Pb and Ag in Ar (solid symbols) and in He (open symbols). The optimum atomization temperatures of Pb and Ag in the Ar plasma are 400 and 500 "C higher respectively than those in the He plasma.Fig. 8 also shows that Pb and Ag emission pulses can be observed at I 100 "C in the He plasma while no signal can be observed at this temperature in the Ar plasma. Table 2 shows that the optimum atomization temperatures for Cd T1 Pb Ag Sb and Au are 300-500 K lower in the He plasma than in the Ar plasma. The optimum atomization temperatures for longitudinally heated GFAAS in the absence of chemical modifiers are given for comparison. One possible explanation for the earlier appearance of the analyte emission pulse for the He plasma is that He having thermal conductivity about nine times greater than Ar heats the centre electrode by conduction to a greater extent than does Ar.Heating of the electrode by the plasma was not observed prior to atomization for either plasma gas in814 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 2000 [ 1 ioo - 1200 1600 2000 Atomization temperaturePC Fig. 8 Atomization temperature versus emission intensity for Pb (A He; B Ar) and Ag (C He; D Ar); emission intensity for the largest peak (peak 2) is given when two peaks occur Table 2 Optimum atomization temperatures for GFCCP-AES in Ar and He and for GFAAS Element Ar Plasma He Plasma GFAAS* Cd 1600 1100 1500 Zn 1800 1800 2000 Ti 1900 1600 2100 Pb 1800 1400 2000 Ag 2000 1500 2400 Sb 2150 1700 2200 Au 21 50 1700 2200 Sn 2100 2 100 2 500 *Longitudinally heated graphite furnace. this work so any plasma-induced heating would have to occur during the atomization cycle when direct measure- ment of the centre electrode temperature is very difficult.The above notwithstanding the A1 7000 oscillator can be configured to heat the centre electrode if desired. Sturgeon et a[.’ have reported a decrease of 400-500 K in the appearance temperature of P in their FAPES source relative to the appearance temperature in GFAAS. Hettipa- thirana and Blades1* have reported a decrease in the appearance temperature of Pb in GFCCP-AES relative to GFAAS and no change in the appearance temperature of Ag and Mn. Both workers have suggested that the observed shifts to lower appearance temperatures are due to plasma induced dissociation of volatile analyte containing mole- cules. This mechanism does not explain why elements that do not form volatile molecular species have different appearance temperatures in Ar compared with He.Silver and Au certainly do not form volatile molecular species but have appearance temperatures in He lower than those in Ar. Investigation of heating and sputtering of the centre electrode by He and Ar and bombardment of gas-phase molecules by the plasma gases is required to elucidate the different vaporization atomization and excitation mecha- nisms in these two plasma gases. AES at concentrations that have previously been the domain of GFAAS and ICP mass spectrometry. An r.f. oscillator can be built that reliably produces a stable capacitively coupled 27 MHz Ar plasma. Four to nine times greater sensitivity is observed for the Ar plasma relative to the He plasma for the conditions reported here.Volatile metal species that vaporize prior to atomization are likely to condense on the centre electrode. Analyte atoms and molecules that have condensed on the centre electrode are re-vaporized and atomized when the electrode is heated by radiation and conduction from the tube wall. Second surface condensation and re-vaporization of solvent matrix and analyte might be advantageous for reducing matrix interferences. Sputtering or heating by the He plasma can contribute to analyte vaporization from the centre elec- trode. Financial support from the National Research Council of Canada (IRAP-M Grant No. 4 I 138W) the Science Council of British Columbia [Technology B.C. Grant No. 17(T-3)] and the Western Economic Diversification Program (B-90- WD-0592) is gratefully acknowledged.1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 References Falk H. and Ludke Ch. Prog. Anal. Spectrosc. 1988,11,4 17. Ballou N. E. Styris D. L. and Harnly J. M. J. Anal. At. Spectrom. 1988 3 1141. Frech W. Baxter D. C. and Hutsch B. Anal. Chem. 1986 58 1973. Liang D. C. and Blades M. W. Spectrochim. Acta Part B. 1989 44 1059. Smith D. L. Liang D. C. Steel D. and Blades M. W. Spectrochim. Acta Part B. 1990 45 439. Sturgeon R. E. Willie S. N. Luong V. Berman S. S. and Dunn J. G. J. Anal. At. Spectrom. 1989 4 669. Sturgeon R. E. Willie S. N. Luong V. and Berman S. S. Anal. Chem. 1990 62 2370. Banks P. R. Liang D. C. and Blades M. W. Spectroscopy 1992 7 36. Huang D. and Blades M. W. Appl. Spectrosc. 1991,45 1468. Sturgeon R. E. Willie S. N. Luong V. T. and Dunn J. G. Appl. Spectrosc. 199 1 45 14 13. Hettipathirana T. D. and Blades M. W. Spectrochim. Acta Part B. 1992 47 493. Hettipathirana T. D. and Blades M. W. J. Anal. At. Spectrom. 1992 7 1039. Sturgeon R. E. Willie S. N. Luong V. and Berman S. S. J. Anal. At. Spectrom. 1991 6 19. Gilchrist G. F. R. and Liang D. C. Am. Lab. 1993 March 25 34U. Hettipathirana T. D. and Blades M. W. J. Anal. At. Spectrom. in the press. Sturgeon R. E. Willie S. N. Luong V. and Berman S. S. J. Anal. At. Spectrom. 1990 5 635. Brown S. C. Basic Data of Plasma Physics MIT-Wiley New York 1959 p. 16. Wu S. Chakrabarti C. L. and Rogers J. T. Prog. Anal. Spectrosc. 1987 10 1 11. Fuller C. W. Anal. Chim. Acta 1976 81 199. Frech W. Lundberg E. and Cedergren A. Prog. Anal. At. Spectrosc. 1985 8 257. Gilchrist G. F. R. Chakrabarti C. L. Hughes D. H. and Ashley J. F. Anal. Chem. 1992 64 1144. Conclusions Simultaneous multi-element determination of Cd Zn T1 Pb Ag Sb Au and Sn can be carried out using GFCCP- Paper 3/01 583A Received March 18 I993 Accepted May 18 1993

ISSN:0267-9477    

DOI:10.1039/JA9930800809

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 815 Vesicle-mediated Hig h-performance Liquid Chromatography Coupled to Hydride Generation Inductively Coupled Plasma Atomic Emission Spectrometry for Speciation of Toxicologically Important Arsenic Species* Yi Ming Liu Maria Luisa Fernandez Sanchez Elisa Blanco Gonzalez and Alfred0 Sanz-Medel? Department of Physical and Analytical Chemistry University of Oviedo C/Julian Cla veria 8 33006-Oviedo Spain The use of vesicles as mobile phases can provide a synergic combination for high-performance liquid chromatographic (HPLC) separation when coupled to hydride generation inductively coupled plasma atomic emission spectrometric (HG-ICP-AES) detection as illustrated here for arsenic speciation. The more toxic arsenic species including arseneous arsenic monomethylarsonic and dimethylarsinic acids can be separated within 10 min by using didodecyldimethylammonium bromide (DDAB) vesicles in phosphate buffer containing 0.5% methanol and a c18 reversed-phase column which had previously been modified by DDAB soluton.In addition the sensitivity of detection on-line by HG-ICP-AES of the four separated arsenic species was enhanced by the presence of DDAB vesicles. The normalized detection limits for arsenic with the proposed speciation method were down to the sub-ng level and the observed precisions were always better than *!job for the four forms of arsenic at the 10 ng level of the element in water. Arsenic recoveries (full procedure) were found to range from 93 to 108% for tap water and human urine samples by this vesicle-mediated HPLC-HG-ICP-AES technique for arsenic speciation.Keywords Vesicles; high-performance liquid chromatography; arsine generation; inductively coupled plasma atomic emission spectrometry arsenic speciation Element speciation at real-life concentration levels presents a formidable challenge to modern analytical chemistry. Although electroanalytical and ultraviolet-visible mole- cular spectroscopic methods can be used the most popular powerful and reliable analytical tools for element speciation today would appear to be the so-called 'hyphenated' techniques particularly the combination of a powerful separation technique [mainly gas chromatography (GC) or high-performance liquid chromatography (HPLC)] with an adequate element specific detector (e.g.atomic detectors).' Plasma atomic spectroscopy methods including induc- tively coupled plasma and microwave-induced plasma atomic emission spectrometry (ICP-AES and MIP-AES respectively) and recently ICP mass spectrometry (MS) are presently believed to be the most effective for element speciation owing to their high sensitivity simultaneous multi-element monitoring ability and element specific In spite of present availability of fast automatic impe- dance-matching networks that enable good coupling of the r.f. field to the ICPs into which organic solvents are injected one of problems with coupling high-performance liquid chromatography (HPLC) to ICP-AES (or MS) detec- tion is associated with the low tolerance of plasmas to the commonly used organic solvents.6 Their introduction to the ICP detector results in higher plasma background de- creased stability and increase in noise and even eventual extinction of the plasma owing to excessive organic solvent loading.Therefore the search for alternative HPLC mobile phases is obviously worthwhile because it could guarantee satisfactory performance of HPLC-ICP-AES (or MS) hybridizations for speciation purposes. Suyani et al. described the application of a micellar mobile phase of sodium dodecyl sulfate (SDS) for HPLC separation i.e. micellar chromatography of alkyltin compounds followed by ICP-MS dete~tion.~ However the high SDS concen- tration in the micellar HPLC mobile phase which was *Presented at the 1993 European Winter Conference on Plasma tTo whom correspondence should be addressed.Spectrochemistry January 10- 15 1993. necessary for achieving acceptable alkyltin separation could also cause problems such as salt deposition at the nebulizer tip and/or mass spectrometer sampling cone. On the other hand research on arsenic speciation is receiving a great deal of attention because of the high toxicity of certain arsenic compounds and the widely different toxicological effects of several arsenic species. Whereas inorganic arsine arseneous (As*l1) and arsenic (AsV) acids are highly toxic monomethylarsonic (MMAs) and dimethylarsinic (DMAs) acids are only moderately toxic. Moreover arsenobetaine and arsenocholine are considered to be non-toxic to living organisms.8 Numerous approaches have been applied to the determination of arsenic species in various matrices mainly using hyphe- nated techniques based on HPLC separation coupled to atomic spectrometric d e t e ~ t i o n .* ~ ~ ~ ~ ~ ~ ~ ~ - * ~ Separations by HPLC including ion chromatography ion-exchange and ion pair reversed-phase chromatography are most frequently used for arsenic speciation. Sheppard et a1.13 resolved As"' AsV DMAs and MMAs on a single ion chromatography column in 15 min but a slightly alkaline mobile phase (carbonate buffer pH 7.5) had to be used which is not advisable for silica based HPLC columns. Hakala et a l l 4 separated the same mixture of arsenic species in 4 min on a c18 reversed-phase column with tetrabutylammonium hydroxide ( 10 mmol dme3) in the mobile phase as the ion-pairing reagent.However the chromatographic resolution observed between AsIII-DMAs and DMAs-MMAs appeared to be inadequate. More- over Morin et a l l 8 claimed that octadecyl-bonded silica columns were not able to resolve arsenic species such as As1I1 or arsenobetaine by ion-pair reversed-phase chromatography whatever the ion-pairing reagent concen- tration used owing to their limited pH stability range. So far anion-exchange chromatography has been most frequently employed for separation of arsenic species because of their widely varying first pK values. However problems associated with the limited lifetime of these ion-exchange columns and what is more the need to resort to organic solvents and high buffer concentrations for shortening the elution times have been noted by various w o r k e r ~ .~ ~ J ~ J ~816 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 Previous work in this laboratory has shown that surfac- tant organized assemblies (e.g. micelles and vesicles) can enhance chemical generation processes of volatile species [e.g. hydride generation (HG)]. In other words ‘ordered media’ ability to organize reactants at a molecular level and thus favourably to modify the equilibria or kinetic constants,20 can be exploited to enhance the performance of analytical HG-ICP-AES. It has been shown that the detection limits of arsenic determination by vesicle-en- hanced HG-ICP-AES were improved by a factor of two with a significant improvement in the precisions of the ICP emission signals and in the tolerance to matrix interfer- ences.,’ This favourable effect of vesicles could be com- bined with their potential in HPLC separations.22 Although the utilization of surfactant-based organized assemblies in HPLC separations has greatly increased in recent years only micellar mobile phase systems have been investigated extensively.The use of vesicles as mobile phases seems so far to be largely neglected owing to their relatively high viscosities. This paper reports on the first vesicle-mediated coupling of HPLC separation to HG-ICP-AES detection and its detailed evaluation for speciation of toxicologically impor- tant arsenic species which happen to form volatile hydrides on-line with the plasma detector after direct reaction with sodium tetrahydroborate solutions. Experimental Apparatus A Knauer Model 6400 HPLC pump with an attached sample injection valve equipped with a 100 mm3 loop was used for eluent delivery and sample introduction.The analytical column was an LKB cartridge (240 x 4 mm id.) packed with 10 pm Lichrosorb RP 1 8 which had previously been modified by didodecyldimethylammonium bromide (DDAB) solution as described below. A four-channel peristaltic pump (Minipuls-2 Gilson) a 100 cm poly(tetra- fluoroethylene) (PTFE) mixing coil and a laboratory-made gas-liquid separator23 constituted the continuous hydride generator. A mass flow controller (Sho-Rate Brooks Veemendaal The Netherlands) was used for introducing an argon stream into the gas-liquid separator. An S.M. Model 500 W high-intensity ultrasonic processor equipped with a microtip (Sonics & Materials) was used for DDAB vesicle preparation.An Amicon Model 52 (W.R. Grace Danvers USA) was used for ultrafiltration experiments. A Perkin-Elmer ICP 5000 spectrometer interfaced with a microcomputer (Perkin-Elmer Model 3 500) was used for ICP emission measurements. A computer program for transient ICP emission data acquisition and processing was written in BASIC. Chemicals Stock solutions (1000 mg dm-3 of As) of arsenite DMAs and MMAs were prepared by dissolving appropriate amounts of As203 (Merck) in 25 cm3 of 0.5 mol dm-3 NaOH solution and then diluting the solution to 1 dm3 with 0.6 mol dm-3 HC1; however (CH3),AsO2Na-3H,O (Sigma) and CH3AsO(ONa)2-6H20 (Carlo Erba) were dissolved directly in water. An AsV stock solution (1000 mg dm-3) was obtained from Merck.The stock solutions were stored in polyethylene bottles at 4 “C. Working standards were freshly prepared daily by dilution in ultrapure Milli-Q water. The DDAB vesicular solution ( 1 x mol dm-3 DDAB) was prepared by adding 0.23 g of DDAB (Fluka) to 50 cm3 of water and sonicating the solution until all added Vesicular Sample (100 mm7 mobile phase - 1 cm’ min-’ HPLC pump To ICP KI+ Gas-liquid separator vesicle 1 cm3 min-’ Mixing coil Waste NaBH Peristaltic pump Fig. 1 Schematic diagram of the vesicle-mediated HPLC-HG- ICP-AES hybridization for arsenic speciation DDAB was dissolved (about 10 min with a power output of 60 W). This solution was used for preparing by simple dilutions all the other vesicular solutions mentioned in this work.Sodium tetrahydroborate solution (2% m/v) was pre- pared by dissolving NaBH powder (Probus Barcelona) in 0.5% m/v NaOH solution. Filtration of the solution through a Whatman grade 4 filter-paper before use was carried out. All other chemicals were of analytical-reagent grade and distilled and de-ionized (Milli-Q system Millipore) water was used throughout the work. Procedures Column modijkation The C ,-bonded silica reversed-phase column was modified by passing through 500 cm3 of a DDAB solution (1 x mol dm-3) in 50% methanol-50% water at a flow rate of 1 cm3 min-l. Water was then passed through the column. The modified column was kept in water when not in use. Ultrajiltration for partition experiments Single-component arsenic solutions containing 50 pg dm-3 of As were prepared for each form of arsenic species in the mobile phase.Ultrafiltration was performed by using cellulose ultrafiltration membranes (Diaflo Amicon Divi- sion). Arsenic contents of the filtrate and of the original solution were determined by the HG-ICP-AES method. Arsenic speciation Typically 100 mm3 of the working standards of the arsenic species mixture were injected directly into the HPLC system to study the experimental conditions and to evaluate the analytical performance characteristics of the hyphe- nated procedure. The mobile phase was prepared as follows 200 cm3 of NaH,PO (10 mmol dm-3) buffer solution (pH 5.75) containing 0.5% (v/v) methanol was de- gassed with helium and then 200 mm3 of mol dm-j DDAB vesicle solution were added.The HPLC eluate was first mixed with the HCl solution through the mixing coil and then mixed with sodium tetrahydroborate solution for arsine generation (see Fig. 1). A continuous stream of argon carries the generated arsine directly into the ICP injector tube that is the nebulizer was removed and the exit of the gas-liquid separator was directly connected to the plasma. The HPLC-HG-ICP-AES hybridized assembly is shownJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 - 817 - " 0 Table 1 Experimental conditions for arsenic speciation by the vesicle-enhanced HPLC-HG-ICP-AES hybridization L .- \ .- i CI .- CI c.. ;* G 4 K Chromatography- Column i x-x-x-x-x-x D Column temperature Sample loop size Mobile phase HG- (a) HCI+ Kl+ vesicle (b) NaBH Argon carrier flow Analytical line R.f.forward power Refected power Viewing height I CP-A ES- C,,-bonded silica 10 pm particle size 30 "C 100 mm3 10 mmol dm-3 NaH2P04 buffer (pH 5.75)+ 240 x 4 mm i.d. 1 x methanol mol dme3 vesicle of DDAB+0.5% Flow rate 1 cm3 min-' Flow rate 1 cm3 min-I I 10% m/v 0.1% m/v 1 x mol dm-3 DDAB 2% NaBH (stabilized by 0.5% m/v NaOH) Flow rate 1 cm3 min-I 0.70 dm3 mind' 193.69 nm 1 kW < 5 w 15 mm schematically in Fig. 1. Peak heights from the chromato- gram were used in all arsenic quantifications. Experimental conditions finally selected for operation after preliminary investigations are summarized in Table 1. Tap water and urine samples Tap water and urine samples were analysed for arsenic. These samples contained no detectable natural arsenic.They were therefore spiked with the four arsenic com- pounds at different concentration levels. Aliquots ( 100 mm3) of the spiked samples filtered through a nylon 0.45 pm filter were immediately injected into the HPLC system. Results and Discussion Chromatographic Separation The column was previously modified with DDAB as detailed above. The DDAB coating formed proved to be very stable and resistant to water and to the mobile phase passing through. Although organic solvents e.g. methanol could remove the coating the modified column was very durable in the recommended vesicular operation. In fact no significant column behaviour indicating degradation was observed after several months of daily usage. This great advantage of the modified column proposed here versus anion-exchange c01umns~~J~ could be reasonably attributed to continuous stationary phase renewal by the continuous dynamic exchanges of single DDAB molecules between the mobile phase and the modified stationary phase.The pH of the vesicular mobile phase had important effects on the arsenic separation as can be seen in Fig. 2. In the pH range of 5.5-7.0 the two inorganic arsenic species As111 and AsV could always be resolved and separated from the methylated species. Arseneous acid as HAsO (pKa=9.3) which is completely hydrophilic and not ionized in this pH range was eluted with the void volume. On the other hand arsenic acid as H3As0 (pKa=2.3 6.9 11.4) was always negatively charged in the pH range studied; that is the strongest electrostatic interaction was with the positively charged stationary phase and thus was the last to elute.For the two methy- lated arsenic species i.e. (CH,),AsO,H (pKa= 6.3) and CH3As03H2 (pKa=2.6 8.2) their elution sequence could be 5.0 5.5 6.0 6.5 7.0 pH of the mobile phase Fig. 2 Retention times of the arsenic species A As"; B MMAs; C DMAs; and D As"' as a function of the pH of the mobile phase ( I0 mmol dm-3 NaH,P04); experimental conditions as described in Table 1 changed by altering the pH of the mobile phase. At a pH lower than 6 dimethylarsinic acid is mainly in its neutral form and hence it is eluted before monomethylarsonic acid. However at pH higher than 6 it starts to dissociate and so it has a higher apparent negative charge. This increased electrostatic interaction in addition to its natural possible hydrophobic interaction with the stationary phase results in longer retention in the column (now close to that of monomethylarsonic acid).As is schematically shown in Fig. 3 in this vesicle-mediated HPLC separation process arsenic separation is probably achieved by competition of the arsenic species and HzP04- ionic components for interaction sites on the stationary phase and the mobile vesicle pseudophase. The partition coefficients between DDAB vesicles and water for the four arsenic species at the pH 5.75 of the vesicular mobile phase were estimated using an ultrafiltration method. It should be borne in mind that because the stationary phase and the mobile vesicle pseudophase probably possess a similar chemical nature (see Fig.3) the partition of the solute between water and DDAB vesicular pseudophase resembles that between water and the DDAB-modified stationary phase. Results of ultrafiltration experiments for the four arsenic species have been plotted in Fig. 4. These ultrafiltration818 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 Stationary phase Fig. 3 Simplified representation of interactions of the solute in the bulk aqueous phase (water) with the stationary phase and the mobile vesicles Aslll DMAs MMAs AsV Fig. 4 Partition estimations as obtained by ultrafiltration of the arsenic species between DDAB vesicles and water results showed that the largest vesicular partition coefficient (ie. minimum arsenic concentration in the filtrate) was obtained for the solute with higher apparent charge AsV.Neutral As1]* species exhibited the smallest partition coeffi- cient. Thus these results also suggested the involvement of ion-exchange or an ion chromatography mechanism in this HPLC separation process. However it does not appear to be a single interaction mechanism. The DMAs and MMAs showed almost the same vesicle-water partition values although they bear completely different apparent charges at pH 5.75. Moreover they can be well separated with this vesicle-mediated HPLC procedure at lower pH (Fig. 2). Obviously these solutes should not only be affected by electrostatic effects but also by hydrophobic interactions. The dependence of the solute retention time on the concentration of DDAB vesicles in the mobile phase was different to that observed in micellar liquid chromato- graphy (MLC).It is commonly assumed in MLC that an increase in the surfactant concentration in the mobile phase results in a decrease in the retention time of the solutes. The magnitude of this decrease will depend upon their partition coefficients between micelles and the stationary phase.22 It should be realized that the surfactant modified C18 station- ary phase and the vesicles allow for electrostatic interac- tions of ionic solutes simultaneously with possible hydro- phobic interactions. The observed retention behaviours in this study suggest that for the four arsenic species studied the partition between the bulk solution in the mobile phase and the modified stationary phase seems to be the dominant mechanism for separation.Possible partition between the Table 2 Effect of DDAB vesicle concentration on the retention times Concentration of DDAB vesicle/ mol dm-3 Retention time/min As111 DMAs MMAs AsV O* 2.82 4.55 5.96 9.80 1 x 10-6 2.83 4.54 5.98 9.82 5x 10-6 2.83 4.55 5.96 9.82 1 x 10-5 2.83 4.55 5.94 9.82 I x 1 0 - 4 2.83 4.56 5.95 9.81 *The column behaviour was not reproducible owing to the continuous DDAB desorption and release from the stationary phase in the column. U 5 10 15 [N aH *PO I/mmol dm” Fig. 5 Effect of buffer concentration on retention times for A AsV; B MMAs; C DMAs; and D As111; experimental conditions as described in Table 1 bulk aqueous solution and the vesicles in the mobile phase does not seem to play a significant role. As shown by the results in Table 2 it follows that the four arsenic species could still be separated even if no vesicle was added to the mobile phase.It should be pointed out however that the column behaviour was by no means reproducible in the absence of vesicles in the mobile phase owing to the continuous DDAB desorption and release to the passing buffered aqueous solution. Further studies to probe the mechanism involved in this vesicle-mediated liquid chro- matography are under way in this laboratory. The effect of NaH,PO buffer concentration in the mobile phase was most significant in the elution of AsV. As expected the retention time of this solute decreased with the increase in the buffer concentration as can be seen in Fig. 5. When the NaH,PO concentration in the mobile phase was below 5 mmol dmq3 the retention of AsV in the column was too long and what is even worse a great peak broadening was observed.A typical chromatogram obtained in the conditions detailed under Procedures is shown in Fig. 6. The four arsenic species separation was accomplished in less than 10 min. The resolution factors (R) between each of these arsenic species [R = 1.172 (t2 - t l ) / ( w1 + w2) where t is the retention time and w is the peak width at half the peak height] were always far greater than 1 which is usually considered acceptable. l 3 HG-ICP-AES Detection In order to obtain efficient HG and transport to the ICP along with minimum peak broadening at the exit of the HPLC column a gas-liquid separator with the smallestJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL.8 819 I 4.5 min I Time - Fig. 6 Typical chromatogram showing arsenic speciation in water A Asi1'; B DMAs; C MMAs; and D AsV. Experimental conditions as described in Table 1. Arsenic content As111 and MMAs 10 ng; DMAs and AsV 20 ng t a) From HPLC 50 CI .- 40 E e 2 20 .= 30 Y ul .- 2 lo Y 0 MMAs ASU' DMAs Fig. 8 Vesicular enhancement of arsenic detection by HG-ICP- AES following HPLC separation. Arsenic content as for Fig. 6 Table 3 Figures of merit of the vesicle-mediated HPLC-HG- ICP-AES method As species RT*/min DLt/ng RSDS (n = 10) (%) As1'' 2.9 0.5 3.6 DMAs 4.5 1 .O 5.0 MMAs 5.8 0.6 3.0 AsV 9.4 1.2 2.0 *Retention time. TDetection limit calculated as 3 times the baseline noise. /Relative standard deviation 10 ng of arsenic for each species was injected. ( b ) Argon FromHPLC 1 -3 To ICP Fig.7 Gas-liquid separators tested (for comparison) (a) U-type; and ( b ) Browner-type possible void volume should be aimed at. In fact most of the HG-based HPLC-ICP-AES couplings described so far utilize a U-tube type gas-liquid separator. The U-type separator depicted in Fig. 7(a) was tested but a laboratory- made glass bead filled separator [Fig. 7(b)] described p r e v i o ~ s l y ~ ~ * ~ ~ provided much better ICP detection charac- teristics than the U-type separator. The analytical signal was higher and the base-line noise was greatly depressed while no significant broadening in the chromatographic peaks was observed (in spite of its relatively bigger size). The enhancing effect of vesicles2' on the ICP on-line signal of the separated arsenic species was verified.As can be seen in Fig. 8 in the presence of DDAB vesicles ( 1 x 1 0-3 mol dm-3) the ICP emission signal increased substantially for all of the four arsenic species under study. The relatively higher concentration of DDAB vesicles needed to produce the satisfactory enhancement in the ICP detection as compared with that for the separation was introduced with the arsine producing reagents after the HPLC separation (see Fig. 1). Unnecessary build-up of DDAB in the HPLC column (resulting in increased pressure drops across the column) from the mobile phase passing through can be easily avoided in this way. Characteristics of the Method The reproducibility of the arsenic separation by this novel chromatographic system was initially evaluated. The col- umn efficiency and the resolution values were studied over a period of 6 months.The column modification procedure was also performed three times (each time the old DDAB coating on the column was removed by passage of a large volume of methanol). No substantial fluctuation in reten- tion times and resolution values was observed from the data collected over these months. The combination of versatil- ity stability efficiency and low cost should make this vesicle-mediated HPLC procedure an attractive choice not only for arsenic separation but also for separations of other anions and possibly neutral species. The precision of the determinations of the arsenic species by this hyphenated method was investigated using ten injections of a mixed arsenic solution containing all of the four arsenic species each present at 100 pg dm-3 each.The relative standard deviations (RSDo/o) of the peak height results calculated for each form of arsenic were always better than 5% (Table 3). Typical calibration graphs for each of the arsenic compounds investigated are shown in Fig. 9. These were obtained by injecting the mixed standard arsenic solutions of different concentrations and plotting peak height for each arsenic species against the amount injected onto the column. As can be seen the four arsenic species showed different sensitivities. This result could be partly due to the different kinetics of the on-line continuous hydride generation of each arsenic compound and also due to the chromatographic peak broadening observed for DMAs and AsV.However the sensitivities would be more similar if peak areas instead of peak heights were used for calibration. The detection limits obtained for each species are shown in Table 3 and were estimated as three times the signal-to-noise ratio and referred to total arsenic injected in water and are seen to be 0.5-1.2 ng of arsenic. Real Samples Arsenic speciation in tap water and human urine by using this hybrid technique was also evaluated. As real samples of820 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 ~~ ~ Table 4 Recoveries of arsenic species added into tap water and human urine samples Experiment 1 Experiment 2 Samples Arsenic Tap water As1i1 DMAs MMAs AsV Human urine As111 DMAs MMAs AsV Spikedlpg dm-3 I50 250 150 250 I50 250 150 250 Recovered* (O/o) 95 98 102 103 96 I05 102 100 Spikedlpg dm-3 50 100 100 100 50 100 50 I00 Recovered* (O/O) 97 I02 100 104 101 I08 99 93 *Mean of two analyses. 60 1 1 0 10 20 30 40 50 Arsenic content/ng Fig.9 Calibration graphs for the four toxic arsenic species under study A As111; B MMAs; C DMAs; and D AsV the local tap water and control human urine available did not contain detectable All1 DMAs MMAs and AsV they were spiked with the four targeted arsenic compounds so as to produce various known arsenic concentrations in the real samples. The spiked samples were filtered through a 0.45 pm nylon filter and analysed immediately. The recovery values (mean of two injections) for all of the four arsenic species are shown in Table 4 and indicate that the proposed method can be used for arsenic speciation in such types of real samples.Conclusion The HPLC separation with HG-ICP-AES detection can be a synergic combination through the use of vesicles as mobile phases for the speciation and determination of As11' DMAs MMAs and AsV. These four arsenic compounds can be separated by using a DDAB vesicular mobile phase on a previously modified C,,-bonded silica column within 10 min. The vesicle-mediated HPLC procedure and column have proved to be fairly robust. Chromatographic behav- iour was most durable and reliable in fact no degradation behaviour was noticed after months of daily usage. On the other hand the sensitivity of the HG-ICP-AES detection of the separated arsenic compounds was enhanced by resort- ing to a post column for further addition of DDAB vesicles with the HG reagents.Moreover with a laboratory-made glass bead filled Browner-typez4 gas-liquid separator a smooth gas-liquid separation and thus reduced plasma noise and increased signals were achieved without signifi- cant chromatographic peak broadening. In brief the vesicle-mediated HPLC-ICP-AES technique proposed here provides a comparatively low-cost efficient and robust separation along with improved ICP perform- ance. Therefore this approach should be able to be extended to the speciation of other elements forming volatile species whose generation can be enhanced by surfactant-based organized media such as Pb Cd and Hg.zo Financial support from Comision Interministerial de Ciencia y Technologia (CICYT) (Project Ref.PB9 1-0669) is gratefully acknowledged. Y. M. L. thanks the Spanish Ministry of Education and Science for a postdoctoral fellowship. References 1 2 3 4 5 6 7 8 9 10 I 1 12 13 14 15 16 17 18 19 20 21 22 23 24 Trace Element Speciation Analytical Methods and Problems ed. Batley G. E. CRC Press Boca Rabon FL 1989. Environmental Analysis Using Chromatography Interfaced With Atomic Spectroscopy eds. Harrison R. M. and Rapso- manikis S. Ellis Horwood Chicester 1989. Chau Y. K. Analyst 1992 117 571. Ebdon L. Hill S. and Ward R. W. Analyst 1987 112 I . Van Loon J. C. and Barefoot R. R. Analyst 1992 112 563. Boom A. W. and Browner R. F. Anal. Chem. 1982 54 1402. Suyani H. Heitkemper D. Creed J. and Caruso J. Appl.Spectrosc. 1989 43 962. Hodgson E. Mailman R. B. and Chambers J. E. Dictionary of Toxicology Macmillan London 1988. Ebdon L. Hill S. Walton A. P. and Ward R. W. Analyst 1988 113 1159. Beauchemin D. Siu K. W. M. Mclaren J. W. and Berman S. S. J. Anal. At. Spectrom. 1989 4 285. Heitkemper D. Creed J. Caruso J. A. and Fricke F. L. J. Anal. At. Spectrom. 1989 4 279. Spall W. D. Lynn J. G. Andersen J. L. Valdez J. G. and Gurley L. R. Anal. Chirn. Acta 1986 58 1340. Sheppard B. S. Caruso J. A. Heitkemper D. T. and Wolnik K. A. Analyst 1992 117 971. Hakala E. and Pyy L. J. Anal. At. Spectrom. 1992 7 191. Rauret G. Rubio R. and Padro A. Fresenius' J. Anal. Chem. I99 1,340 157. Violante N. Petrucci F. La Torre F. and Caroli S. Spectroscopy 1992 7 36. Colon L. A. and Barry E. F. J. High Resolut. Chromatogr. 1991 14 608. Morin Ph. Amran M. B. Lakkis M. D. and Leroy M. J. F. Chromatographia 1992 33 58 1. Chana B. S. and Smith N. J. Anal. Chim. Acta 1992 197 177. Sanz-Medel A. Fernandez de la Campa M. R. Valdes-Hevia M. C. Aizpun Fernandez B. and Liu Y. M. Talanta in the press. Aizpun Fernandez B. Valdes-Hevia M. C. Fernandez de la Campa M. R. Sanz-Medel A. Talanta 1992 39 I 5 17. Ordered Media in Chemical Separations eds. Hinze W. L. and Armstrong D. W. American Chemical Society Washing- ton DC 1987. Menendez Garcia A. Sanchez Uria J. E. and Sanz-Medel A. J. Anal. At. Spectrom. 1989 4 581. Pyen G. S. Long S. and Browner R. F. Appl. Spectrosc. 1986 40 246. Paper 3 / O 1 004 J Received February 19 1993 Accepted April 30 1993

ISSN:0267-9477    

DOI:10.1039/JA9930800815

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 82 1 Sensitive Method for Determination of Lead by Potassium Dichromate-Lactic Acid Hydride Generation Inductively Coupled Plasma Atomic Emission Spectrometry* M. C. Valdes-Hevia y Temprano M. R. Fernandez de la Campa and Alfredo Sanz-Medelt Department of Physical and Analytical Chemistry Faculty of Chemistry University of Oviedo Julian Claveria s/n 33006 Oviedo Spain Continuous flow plumbane generation for sample introduction into an inductively coupled plasma (ICP) and further determination of lead by atomic emission spectrometry (AES) has been investigated in order to increase the detection limits of conventional nebulization ICP-AES. Continuous hydride generation from different media is discussed and the generation of plumbane using potassium dichromate with lactic acid has been selected.Sensitivity selectivity and accuracy of the corresponding determination of lead by ICP-AES are investigated. The proposed method (detection limit 2 ng ml-I precision 1.3% at the 50 ng ml-' level) has been applied to the determination of low levels of lead in soft drinks sediments and lichens. The results obtained show good agreement with certified or expected values. Keywords Lead; inductively coupled plasma atomic emission spectrometry; continuous hydride generation; potassium dichromate-lactic acid There is currently much concern regarding the adverse effects of lead on health' and the considerable amount of organic and inorganic lead pollutants emitted into the atmosphere.Thus in recent years many countries have given special attention to reducing the lead content in gasolines replac- ing tetraalkyllead species additives in petrols by other chemicals and enforcing adequate control of lead concen- tration in the en~ironment.~~~ Therefore the present inter- est in developing more sensitive selective and reliable analytical methods to determine traces and ultratraces of this element in samples of biological and environmental interest is understandable. Analytical atomic spectrometry techniques particularly inductively coupled plasma atomic emission spectrometry (ICP-AES) have been widely used for this purpose. However the low levels to be analysed are demanding the use of more sensitive methods than those provided by conventional nebulization ICP-AES. The introduction of lead into the atomizer as a vapour by plumbane generation has been widely applied to enhance the sensitivity of lead determinations by atomic absorption spectrometry (AAS).4-23 It is recognized today that the efficient generation of lead hydride the least known of the hydrides of the Group 14 elements because of its instabil- it^,^ requires the simultaneous presence of an oxidant and a complexing agent of metastable lead(Iv) before final reduction to PbH4 with NaBH,.In fact many oxidants and complexing agents have been reported to increase the sensitivity of the lead hydride based techniques i.e. potassium dichromate with tartaric acid malic acid5v6 or lactic acid;7 hydrogen peroxide with citric acid,8 nitric acid6y7v9 or hydrochloric a ~ i d ; ~ .~ ~ ammonium persulfate with nitric a ~ i d ; ~ - ~ J ~ - l ~ potassium hexacyanoferrate(~~r),l~ am- monium cerium(II1) nitrate with oxalic acid,Is or nitroso R- salt.23 The generation of plumbane has also been achieved from organic medial6-** and from 'ordered media' (micelles and vesicles).26 Madrid et aL7 compared different systems and methodologies for plumbane generation by AAS tech- niques. They concluded that to their knowledge the best oxidant system was potassium dichromate-lactic acid.7 * Presented at the 1993 European Winter Conference on Plasma t To whom correspondence should be addressed. Spectrochemistry Granada Spain January 10-1 5 1993. Therefore the purpose of the present work was to examine the ability of the potassium dichromate-lactic acid system to improve the analytical performance characteristics of continuous hydride generation (HG) ICP-AES techniques for the determination of lead.As a result a sensitive and selective analytical method for the determination of lead by continuous HG-ICP-AES has been developed using the dichromate-lactic acid reaction medium. The validity of this method has also been demonstrated by the successful determination of lead at trace levels in some environmental and food samples. Experimental Instrumentation An ICP Philips Model PU7000 was used for ICP-AES detection. Other details of the experimental flow system and the ICP-AES set-up used are given in Table 1 and Fig. 1. Reagents A 1000 pg ml-l lead@) stock standard solution was stabilized in 0.5 moll-' HN03 (Merck).Working solutions were freshly prepared daily by diluting appropriate aliquots of the stock solution with ultrapure water. Sodium tetra- hydroborate(II1) solutions were prepared by dissolving NaBH (Carlo Erba) in ultrapure water (Milli-Q) and stabilized in a 0.1% m/v sodium hydroxide solution. Solutions were prepared weekly and filtered before use. Potassium dichromate (Merck) and lactic acid (Merck 90%) solutions were prepared by appropriate dissolution of the required amounts of the reagents in ultrapure water. All mineral acids and metal salts used were of analytical- reagent grade and ultrapure water (Milli-Q) was used throughout. General HG-ICP-AES Procedure Continuous plumbane generation In the flow system described here (Fig. I) the sample a 0.3% m/v potassium dichromate and 3% v/v lactic acid solutions were passed through a four channel union cross by a peristaltic pump at a rate of 0.75 ml min-l each.The resulting mixture (see Fig. 1) was merged with a 5% m/v822 Tetrahydroborate- Acid Oxidant Sample (lead) - JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 0.75 ml min-' 0.75 ml min-' Table 1 Instrumentation ICP Spectrometer Gas-liquid separation interface Peristaltic pump (four channels) Microwave oven for sediment and lichen samples digestion Philips Model PU7000 equipped with 40 MHz source unit Grid type nebulizer provided with the instrument Gilson Minipuls 2 Milestone. Digestion module MLS- 1200; exhaust module EM-5; automatic capping module AC- 100; closed vessels SV- 1401 I0 Table 2 Optimum conditions for lead hydride generation by ICP-AES Plasma experimental conditions Wavelengthhm Radio frequency forward powerlkW Nebulizer pressure/lb in-* Coolant gas flow rate/] min-' Auxiliary gas flow ratell min-l Final sample flow ratelm1 min-' Drain flow rate/ml min-' Integration ti me/s Chemical parameters Potassium dichromate Lactic acid Sodium tetrahydroborate 220.353 0.7 30 13 0 3 0.3% m/v (flow rate 0.75 ml min-') 3% v/v (flow rate 0.75 ml rn1n-l) 5% m/v in NaOH 0.1% m/v (flow rate 0.75 ml min-I) sodium tetrahydroborate solution (flow rate 0.75 ml min-l) via an ordinary T-piece producing a final flow rate of 3 ml min-l.This solution feeds the grid nebulizer of the ICP detuned in order to separate the volatile species. Thus PbH goes to the plasma and is separated from the liquid phase going to waste.Lead was measured at the 220.353 nm emission line under the conditions given in Table 2. Background correction at 220.371 and 220.339 nm was carried out. Dissolution of Sediments and Lichens A 0.200 g sample was weighed directly into a poly(tetra- fluoroethylene) (PTFE) vessel and 1.5 ml of HN03 0.5 ml of HClO and 0.5 ml of HF were added. After tightly capping the vessels using the Milestone capping station the sample carousel was placed in the microwave oven using the heating programme shown in Table 3. After cooling the vessels were uncapped and 3.5 ml of a solution of 6% m/v H3B03 were added for complexing the excess HF; then the PTFE vessels were placed again in the microwave oven and heated for 5 min at 60 W power.The sample solutions were then filtered and the contents transferred into 100 ml calibrated flasks and diluted with ultrapure water (Milli-Q). This solution was the sample (lead) used for PbH generation ICP-AES. Separate solid ICP (grid nebulizer) portions (1.000 g each) from each sample were dryed at 105-1 10 "C and used to determine the water content of the sample. Juice samples were directly analysed as received without any sample pre-treatment following the continuous flow plumbane generation described above. Results and Discussion Optimization of Instrumental and Chemical Parameters Using continuous gas-liquid separation and ICP detection the effect of ICP and chemical generation variables such as nebulizer gas pressure forward r.f.power coolant gas flow sample flow rates etc. were studied by following the general HG-ICP-AES procedure and a univariant-type experimen- tal search. The optimum ICP instrumental values obtained are summarized in Table 2 and were selected to study optimum chemical parameters (concentration of reactants flows etc.) for continuous PbH generation and introduc- tion into the plasma. Maximum signal-to-background ratio was always the optimization criterion. The results observed have been plotted in Fig. 2 for potassium dichromate lactic acid and NaBH and show that the optimum concentrations are 0.3% m/v 3% v/v and 5% m/v respectively. The effect of sample flow rate has been studied from 1 to 10 ml min-I and as expected the signal increased almost linearly with the sample flow rate as shown in Fig.3 in which the observed results have been plotted. A flow rate of 3 ml min-' was selected as a compromise in order to avoid flooding of the spray chamber in the commercial set- Table 3 Microwave oven heating programme Step Time/min PowerIW I 5 60 2 5 120 3 10 180 4 5 0 5 10 120 Fig. 1 Schematic diagram of the continuous HG flow system usedJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 823 Dichromate (% m/v) 0 0.,2 0;4 0.,6 0. j 2 4 6 8 10- Lactic acid (% v/v) or tetrahydroborate (YO m/v) Fig. 2 Optimization of A tetrahydroborate; B lactic acid; and C dichromate concentrations % .- E 1001 I I I I w l 3 5 7 9 Sample flow rate/mI min-' Fig. 3 Effect of sample flow rate on lead signal A maximum intensity; and B background intensity.[Pb]=100 ng ml-l and drain flow rate= I 1 ml min-l. Other conditions as in Table 2 up but the results in Fig. 3 show a way of increasing the detection power if necessary; wider diameter drain tubes and a higher sample flow can be used without flooding. Analytical Performance Characteristics Using the experimental conditions given in Table 2 the analytical parameters defining the performance of the proposed method were evaluated. The calibration graph was linear up to 0.5 pg ml-I of Pb and the detection limit for the element ( 3 ~ ~ ) was 2 ng ml-I. The within-run precision evaluated by analysing ten replicates containing 50 ng ml-I of Pb turned out t o be 2 I .3%. It should be stressed however that working at a sample flow rate of 9 ml min-' (see Fig.3) and draining with adequate tubing able to drain out at 1 1 ml min-l the detection limit observed was 0.7 ng ml-1 with a within-run precision evaluated (over 10 ng ml-1 of Pb) of around f 3%. Effect of the Nature of the Acid Used for PbH4 Generation The mineral acids tested were HCl HC104 HN03 H2S04 and HF (with boric acid) because they are currently used in many sample digestions. All these mineral acids except sulfuric acid which produced maximum intensity at a concentration of 0.05 mol 1 - I and at a level of 0.2 mol 1 - l caused the emission signal of Pb to disappear showed similar behaviour to that illustrated for nitric acid in Fig. 4 curve A. As citric acid is normally present in the soft drinks to be analysed its addition to the HG medium was also investigated in more detail.Results showed that inadequate HG of lead from dichromate-lactic acid was obtained; the lead recovery decreased almost exponentially with increas- ing citric acid concentration under the conditions of the proposed procedure as shown in Fig. 4 curve B. Lead(rv) metastable complex formation with lactic acid in the proposed procedure which seems to act as an intermediate favouring plumbane generati~n,~ could be hindered by citric acid acting as a competitive ligand. For comparison purposes Fig. 4 curve C shows the observed effect of citric acid using a different PbH4 generation system namely ammonium persulfate-nitric acid-cetyltrimethylammonium bromide (CTAB).26 It can be seen that the effect of citric on Pb signals is much less in this latter PbH4 generation medium.Kinetic Studies In order to further investigate the nature of the observed enhancement using potassium dichromate-lactic acid,7 batch AAS signals were studied by comparing the peak shape of the transient AAS signal obtained when generating PbH4 from the oxidant systems2? potassium dichromate- lactic acid (0.3% m/v K2CrZ0 2% v/v lactic acid); ammonium persulfate-nitric acid-CTAB; and ammonium persulfate-nitric acid [3% m/v (NH4)2SZOs 2% v/v HN03 and 1 x mol 1 - I CTAB]. The results are shown in Fig. 5 and indicate that the presence of micelles and vesicles seems to accelerate the HG kinetics in the ammonium persulfate-nitric acid system but the potassium dichro- matic-lactic acid system provided higher and sharper lead peaks.Interference Studies Five possible sources of lead interferences were investigated for selective PbH4 generation (i) hydride forming elements; (ii) transition metals; (iii) alkali metals; (iv) alkaline earth metals; and ( v ) anions. All the elements tested and the level of tolerance observed in the determination of 0.1 ppm of lead by the proposed method are summarized in Table 4. [Nitric acidl/mol I-' 0 0.15 0.30 0.45 0.60 0.75 I I I I I n t 7 2 1 1 I I I I I 0 0.05 0.10 0.15 0.20 0.25 [Citric acidl/mol I-' Fig. 4 Effect of A nitric acid; and B citric acid on the ICP-AES signal of lead with HG from potassium dichromate-lactic acid medium; and C effect of citric acid on the ICP-AES signal of lead with HG from ammonium persulfate-nitric acid-CTAB medium824 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL.8 Table 4 Interference studies on 100 ng m1-I of Pb Amount of interferenv Pb interferent Recovery* Interferent pg ml-I mass ratio (O/O) 50 10 50 10 50 10 10 10 50 50 50 50 50 50 200 1000 1000 1000 1000 1000 1000 1000 5000 5000 1000 1000 1000 0.5 1:500 1:lOO 1300 1:lOO 1:500 1:lOO 1:lOO 1:lOO 1:500 1:500 1:5 1500 1:500 1:500 1:500 1:2000 1 10000 1 10000 1 10000 1:10000 1 10000 1 I0000 1 10000 1 :50000 1 :50000 1 10000 1 10000 1 10000 82 77 103 78 82 93 84 78 123 102 94 99 99 100 93 104 81 103 103 100 100 100 I00 102 100 94 102 90 * The errors (precision) observed in each case (expressed as relative standard deviation) were within f 3%. i- I Time -t Fig. 5 Effect of the different oxidant systems on the lead batch AAS peak profile A potassium dichromate-lactic acid absor- bance 0.447 5-fold reduced signal; B ammonium persulfate-nitric acid-CTAB absorbance 0.082; and C ammonium persulfate- nitric acid absorbance 0.039 The observed interferences due to hydride forming elements can be explained by competitive reactions in which these elements successfully compete with lead for the NaBH to form their corresponding hydrides at the expense of plumbane production.Copper always interfered in this determination at the levels studied. This interference could be caused by a reduced rate of evolution.27 High levels of alkali alkaline earth metals or common anions were found not to affect lead hydride generation (see Table 4). Analysis of Real Samples The method established for lead determination by continu- ous HG-ICP-AES from a mixture of potassium dichrom- ate-lactic acid was applied to the determination of low levels of lead in certified sediments and in lichens. Certified reference materials (CRMs) sediments and lichens from the Community Bureau of Reference (BCR) were analysed according to the general recommended procedure by reference to an aqueous calibration line and also by standard additions techniques (to check for matrix interferences).Background correction (at 220.37 I and 220.339 nm) was employed. The results obtained in each case corrected for dry mass in pg g-* can be seen in Table 5 for sediments and the certified lichen. Commercial fruit juices were analysed directly without any sample pre-treatment by HG-ICP-AES and the results obtained by the proposed method were compared with the Table 5 Analysis of real samples Obtained value/ Certified value*/ Sample lug g-' Pg g-l BCR CRM 227 Estuarine Sediment Calibration line 128.3 f 2.8 146f3 Standard additions 144.7 f 3.4 146k3 BCR CRM 320 River Sediment Calibration line 41.3+ 1.3 42.3k 1.6 Standard additions 42.9 2 2.1 42.3 k 1.6 BCR RM Lichens TP-24 Calibration line 5.84 k 0.0 I 5.7 1 f 1.30 HG-ICP-AESf/ ETGAAS/ ng ml-1 ng m1-l Apple 66.4 f 7.3 65.3 f 1.9 Orange 13.0 +- 5.2 11.3 +.1.0 Pineapple 32.6 f 2.7 33.0 k 3.7 * Recommended value given for Lichens T2-240. f Determination using standard additions.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 825 results for lead obtained by electrothermal atomic absorp- tion spectrometry (ETAAS) for the same samples. The lead content in the three commercial juices analysed was evaluated by using the technique of standard additions because of the interference of citric acid (see Fig.4 curve B) an important component in these commercial juices usually present at a concentration level of 0.1 mol 1-* (2% m/v). The validation of these latter results was carried out by ETAAS. The results obtained by both methods can be seen comparatively in Table 5 and show very good agreement between the expected and the observed values of lead content. It should be noted that the precision obtained by HG-ICP-AES was worse than that with ETAAS prob- ably because of matrix interferences which indicates the necessity of using the standard additions technique.Conclusions The results demonstrate that a mixture of potassium dichromate and lactic acid provides improved analytical performance characteristics (higher sensitivity good preci- sion and low interference level) for the determination of lead by ICP with plumbane generation. This is probably due to enhanced kinetics and efficiency of PbH4 generation (see Fig. 5). The detection limit of the ICP-AES determina- tion of lead in water with HG using the recommended procedure is 2 ppb which compares favourably with 13 ppb by HG-ICP-AES using ammonium persulfate-nitric acid and with 9 ppb adding CTAB surfactant to the latter medium.26 Conventional nebulization of lead by ICP-AES observed in this set-up was 20 ppb. The method seems to be more selective than the more common method used for plumbane generation ( i e .ammonium persulfate-nitric acid ~ y s t e m ~ - ~ ~ ' ~ - ~ ~ ) and has proved to be adequate for the determination of low levels of lead in environmental samples of varied matrices ( i e . sediments plant material and fruit juices). We gratefully acknowledge Fundacion para el Foment0 en Asturias de la Investigacion Cientifica Aplicada y la Technologia (FICYT) and Comision Interministerial de Ciencia y Technologia (CICYT) for financial support and a grant to M. C. V. H. y T. The loan of the ICP instrument from Unicam Analytical Systems (Cambridge) is also gratefully acknowledged. References 1 Beockx R. Anal. Chem. 1986 58 274A. 2 Taylor A. Clin. Endocrinol. Metab. 1985 14 658. 3 Off. J. Eur.Comm. 1980 23 11. 4 Fleming H. D. and Ide R. G. Anal. Chim. Acta 1976,83 67. 5 Brindle I. D. and Le X. Anal. Chem. 1989 61 1175. 6 Jin K. and Taga M. Anal. Chim. Acta 1982 143 229. 7 Madrid Y. Meseguer J. Bonilla M. and Camara C. Anal. Chim. Acta 1990 237 181. 8 Vijan P. E. and Wood G. R. Analyst 1976 101 966. 9 Aroza I. Bonilla M. Madrid Y. and Camara C. J. Anal. At. Spectrom. 1989 4 163. 10 Jin K. Taga H. Yoshida H. and Himike S. Bunseki Kagaku 1978 27 759. 1 1 Madrid Y. Bonilla M. and Camara C. J. Anal. At. Spectrom. 1988 3 1097. 12 Madrid Y. Bonilla M. and Camara C. J. Anal. At. Spectrom. 1989 4 167. 13 Sanz J. Basterra P. Galban J. and Castillo J. R. Mikro- chim. Acta 1989 1 271. 14 Tao R. and Zhou H. G. Fenxi Huaxue 1985 13 253. 15 Li J. Liu Y. and Lin T. Anal. Chim. Acta 1990 231 151. 16 Aznarez J. Palacios F. Vidal J. C. and Galban J. Analyst 1984 109 7 13. 17 Aznarez J. Vidal J. C. and Carnicer R. J. Anal. At. Spectrom. 1987 2 55. 18 Nerin C. Olavide S. and Cacho J. Anal. Chem. 1987 59 1918. 19 Bonilla M. Rodriguez L. and Camara C. J. Anal. At. Spectrom. 1987 2 157. 20 Thompson K. C. and Thomerson D. R. Analyst 1974 99 595. 21 Madrid Y. Bonilla M. and Camara C. Analyst 1990 115 563. 22 Yan X.-p. and Ni Z.-m. J. Anal. At. Spectrom. 1991 6,483. 23 Zhang S.-z. Han H.-b. and Ni Z.-m. Anal. Chim. Acta 1989 221 85. 24 Paneth F. and Rabinowitsch E. Ber 1925 58 1138. 25 Ikeda M. Nishibej J. Mamada S. and Tujino R. Anal. Chim. Acta 198 1 125 109. 26 Valdes-Hevia y Temprano M. C. Aizpun Fernandez B. Fernandez de la Campa M. R. and Sanz-Medel A. Anal. Chim. Acta 1993 in the press. 27 Nakahara T. in Sample Introduction in Atomic Spectroscopy ed. Sneddon J. Elsevier Amsterdam 1990 ch. 10. Paper 3/00284E Received January I I993 Accepted March 16 1993

ISSN:0267-9477    

DOI:10.1039/JA9930800821

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 827 Boron Determination in Steels by Inductively Coupled Plasma Atomic Emission Spectrometry. Comparative Study of Spark Ablation and Pneumatic Nebulization Sampling Systems* Aurora G. Coedo Teresa Dorado Ester Escudero and Isabel G. Cob0 Centro Nacional lnvestigaciones Metalurgicas Consejo Superior de Investigaciones Cientificas Gregorio del Amo 8 28040 Madrid Spain An experimental study for the determination of boron in steels by inductively coupled plasma atomic emission spectrometry is presented. A comparison is made of spark ablation and pneumatic nebulization (after microwave digestion) sampling systems. A one-step microwave digestion procedure for total boron content using diluted aqua regia (HCI+HNO 3+1) and high pressure vessels was developed.The influence of microwave power and time on the dissolution of boron compounds is discussed. The strongest available conditions (0 Q 600 V 600 Hz) were required for spark ablation sampling. The stability of spark sampling during the spark ablation-ICP process was tested by plotting iron and boron emission versus sparking time. The iron content of collected and dissolved spark-produced particles was analysed and showed that the amounts of aerosol from different boron steels samples during 90 s sparking processes were fairly similar. The analytical performance of both systems was evaluated. Using pneumatic nebulization after microwave digestion of the sample a detection limit for boron of 2.6 ,ug g-1 and overall relative standard deviation (RSD) values of 1-3.5% were found.For spark ablation the detection limit for boron was 0.65pg g-l the overall RSD ranging from 0.5 to 1.5%. A comparison of the data for British Chemical Standards (BCS) Certified Reference Materials (CRMs) Carbon Steel Residual Series (Group B) and Spectroscopic Standard (SS) 456/1 to 460/1 indicated that the accuracy of both methods was satisfactory. Keywords Spark ablation; microwave digestion; inductively coupled plasma atomic emission spectrometry; boron determination; steel The importance of the effect caused by very low concentra- tions of boron on the physical properties of steels (hot workability hardenability creep resistance etc.) requires precise and accurate determination of this element. Com- pared with most other alloying elements the amount of boron added to steel is extremely small and commonly the boron content in boron treated carbon steels ranges from 0.0005% or less to about 0.005% (too little boron may be ineffective and too much can cause difficulties in rolling or forging).Boron was probably the first element in steelmak- ing to require analytical techniques capable of determining this element at the pg g-I level. Ambrose et aL2 have reviewed methods for the determina- tion of boron in steel. These workers also examined and commented on the method adopted by the European Committee for Iron and Steel Standardization/Technical Committee 20 (ECISS/TCZO) and presented analytical results for the boron content of some certified reference steel samples using different methods.The procedure suggested by Thiering,3 and the study mentioned above form the basis of the International Organization for Standardization (SO) standard method for boron determinati~n.~ This interna- tional standard is applicable to boron contents in steel of between 0.0005% and 0.012% (m/m) and specifies a spectrophotometric procedure using curcumin which is sensitive to experimental conditions during the dissolution steps in order to obtain quantitative dissolution of all the boron compounds and low blank values. The technique of microwave digestion in pressure vessels has progressed significantly over recent year^,^-^ and offers a very effective and fast dissolution system in addition to low blank values. These characteristics make the system very attractive for use in inductively coupled plasma (ICP) met hods.Over the past several years there has been growing interest in developing direct solid sample introduction *Presented at the 1993 European Winter Conference on Plasma Spectrochemistry Granada Spain January 10- 15 1993. systems for use with ICP spectrometry. Broekaert et aL9 reported a study of some techniques for direct solid sampling in plasma spectrometry. Although several s t u d i e ~ I ~ - ~ ~ have shown that the spark ablation-ICP combination is a convenient means of direct solid steel sample analysis it is difficult to ascertain the physical and chemical characteristics of the spark-produced aerosol and consequently the analyti- cal performance of the technique. Watters et af.14 presented a detailed study of the physical properties and the chemical composition of a spark-produced aerosol and corresponding erosion craters on brass certified reference materials (CRMs). In this paper a comparison is made between pneumatic nebulization (from microwave dissolved samples) and spark ablation sampling systems used with ICP spectro- metry for analysing boron in steels.In an effort to establish a microwave dissolution proce- dure a study was conducted to select the most important variables conditioning the microwave digestion process acid mixture digestion vessels proportion of sample to reagent required power time and number of samples. For spark ablation the first assays aimed to prove the similarity of the amounts of aerosol produced (from different steel samples) and the stability of its chemical composition.The aerosol particles produced for the spark- ing process were collected and dissolved in aqua regia (HCl+HN03 3+ 1) and the iron content was analysed by ICP spectrometry to determine the total mass eroded. The stability of the aerosol composition was verified by plotting the boron and the iron emission intensities versus spark erosion time curves. British Chemical Standards (BCS) CRMs Carbon Steel Residual Series Group B and Spectroscopic Standard (SS) 45611 to 460/1 were used for testing both methods. Experimental Instrumentation Inductively coupled plasma atomic emission spectrometry (ICP-AES) measurements were performed with a Jobin828 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL.8 Table 1 Selected working parameters Microwave oven Power1W 360 Time/min 30 Digestion vessels HPV 80 Voltage1V 600 Repetition rate (f)/Hz 600 Sampling spark Capacitance1pF 1 InductancelpH 20 Resistance (R)lQ 0 Gas flow rate/l min-l Permanent carrier gas 2.1; analysis gas 0.80 ICP-AES Power1W 950 Gas flow rate/l min-' 14 Observation height1mm 13 Yvon JY 24 plasma spectrometer purging with nitrogen. Spark ablation sampling was carried out with a Jobin Yvon JY-SAS sparking unit. Microwave digestion solutions were obtained using a Milestone MLS- 1200 microwave oven and HPV 80 high pressure vessels. Table 1 lists the selected operating conditions. Samples British Chemical Standard CRMs Carbon Steel Residual Series BCS-CRM and SS 456/ 1 to 460/ 1 are boron-carbon steels with certificate boron contents between 0.001 5 and 0.01 18%.The BCS-CRM samples come in chip form for wet analysis while SS samples come in disc form for direct solid analysis. The boron content values are similar in both types of sample and they were re-certified in 1988 in accordance with the study conducted by the European Committee for Iron and Steel Standardization (ECISS/TC 20). The Fe-0 sample with a boron content below 1 ppm was employed as a 'blank'. Results and Discussion Microwave digestion Boron is present in steel both in acid-soluble and non- soluble forms. The most significant non-soluble boron compound is BN. The dissolution step not only increases analysis time but can also cause errors when ICP-AES is used by increasing the spectral background and decreasing overall analytical sensitivity .In this paper a microwave digestion system was used to completely dissolve the sample. Tests were performed on different acid mixtures (containing different proportions of hydrochloric nitric hydrofluoric phosphoric and per- chloric acids) and different proportions of samples and reagents to find the simplest procedure for stable and quantitative dissolution of all types of boron compounds. After this selection different microwave digestion pro- grammes were applied varying the operational parameters of the oven (microwave power and time). Two types of pressure vessels were used low pressure SV-140 [Pmax= 18 bar ( 1 bar= 1 x 1 O5 Pa) V= 140 ml] and high pressure HPV- 80 150 bar V= 80 ml). When a closed vessel is used the pressure rises with an increase in microwave power and depending on how long the microwave acts.Consequently the temperature of the digestion mixture increases also enhancing efficiency. The BCS 46011 sample was used to test the influence of the parameters studied. Coedo and Dorado15 concluded that by using an ICP method this CRM with a certificate value for total boron of 28 ppm contains a high proportion = SV 140 (low pressure) 0 HPV 80 (high pressure) 240 360 500 600 PowerW Fig. 1 Effect of microwave power on boron dissolution of BCS- CRM 46011 for two types of pressure vessel for a time of 30 min = SV 140 low pressure 0 HPV 80 high pressure 3o I ,n 20 n n n I I I I I I U 5 10 20 30 45 Time/s Fig. 2 Effect of time on boron dissolution of BCS-CRM 46011 for two types of pressure vessel.Power 600 W for S V 140 and 360 W for HPV 80 of non-soluble boron particles ( ~ ~ 5 3 % of the total boron content) and so can be used to evaluate the efficiency of the dissolution procedures. The results obtained by varying the microwave power for the two types of vessels and for a fixed period of 30 min are shown in Fig. 1. The influence of time for a fixed power value is shown in Fig. 2. This power value was set at 600 W for SV-140 (the maximum value that could be applied with the selected reagents without risking vessel failure) and at 360 W for HPV-80 (this level of power was enough to recover all the boron content when this type of vessel was used). The BCS 460/1 sample was completely dissolved only when high pressure vessels were used.The volume of the high pressure vessels (HPV 80) is 80 ml the volume of the low pressure vessels (SV 140) being 140 ml. This difference means that under the same operating conditions (same portion of test sample reagents and oven parameters) the pressure reached in the HPV vessels is greater and conse- quently the temperature and the dissolution process effi- ciency is higher. It could be suggested that a high pressure is required to dissolve all the boron compounds in sample BCS 46011 owing to the presence of a high amount of boron nitride and of a high carbon content (0.45%) producing fine carbon particles that help to retain other products such as boron compounds. As a result of this study the following operating dissolu- tion procedure was adopted.A 0.250 g sample was dissolved in an HPV-80 vessel (maximum pressure 150 bar) with 5 ml of HC1+ 2 ml of HN03 + 10 ml of H 2 0 + 2 drops of HF by applying a one-step microwave programme of 360 W lasting 30 min. After cooling the solutions obtained were made up to 50 ml with water in graduated polyethy- lene flasks. The efficiency of the method was verified by simultaneous treatment of 2-6 samples.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 829 Spark Ablation Since the use of the spark to generate sample aerosol divides the sampling and excitation processes into two separate events systematic errors in each step should be examined in order to evaluate the results provided by the combined spark ablation-ICP technique. Most studies have centred on the erosion of the sample surface yet the analytical information is carried to the plasma by the particles that are formed.A high sample ablation and rate of analyte introduction into the plasma provide higher levels of sensitivity and higher detection capacity. The amount of analyte reaching the plasma increases both with the voltage and the repetition rate. However the particle size is increased by raising the voltage applied not by raising the repetition rate. A high repetition rate is preferable to a high voltage because of the particle size produced. An increase in the particle size results both in decreased stability and poor signal-to-background ratio of ICP signals (the finer the particles the better the stability in the plasma).Conse- quently source parameters resistance (R) voltage (V) and repetition rate (f) which have a clear influence on the analyte produced must be changed in line with the nature and characteristics of the samples to be analysed. The condition and hence the burn is stronger with decreasing R but increasing V or f Owing to the hardenability of boron steels and considering that fand R have a direct influence on the amount of material eroded but not on the particle size these parameters were set to obtain strong burns (R = 0 SZ and f=600 Hz). Tests were conducted using different voltage spark values from 350 to 600 V (maximum available voltage). The most repeatable and accurate results were obtained by applying the strongest conditions (600 V). As a result of this study the operating conditions listed in Table I were selected for spark sampling.As the samples heated up during the sparking process they were cooled in a carbon dioxide stream to allow consecutive measurements to be made under the same conditions with a view to achieving better sampling repeatability. Under the established conditions a spark ablation-ICP process lasted about 85 s (30 s for pre-integration and cleaning 15 s for transfer and 40 s for a ‘three-point’ mode measurement with background correction). To minimize sparking times peak intensities were measured employing a ‘three-point’ mode. In this mode the window includes three points with a distance between them of 0.0003 nm and the intensity value is a weighted average of these three points. The aerosols produced from 90 s spark ablation attacks of boron-carbon steels were trapped and dissolved in diluted aqua regia. These particles were removed from the end of the 0.75 m long tube (5 mm id.) which transports the analyte from the sparking chamber to the ICP torch with the aid of an argon gas flow of 0.8 1 min-l by introducing the end of the tube into the acid solution (10 ml of aqua regia+ 40 ml of water).After completing the dissolution by heating the iron content of these solutions was determined by ICP-AES in order to calculate the amount of analyte produced during the period of time from the beginning of the spark process to the acquisition of a spark ICP measurement ( ~ 8 5 s). The results show good repeatability of the amount of eroded material produced during sparking processes from the different boron-carbon steel samples tested.This amount was 0.700-+0.025 mg of iron for sparking periods of 90 s. The profiles of the boron and the iron emission intensities versus spark time show the signal stability of the aerosol produced under the selected spark conditions at least for the duration of the spark ablation-ICP measurement (Fig. 3). ICP Measurements Before measurement a preliminary study was conducted in order to choose the boron analytical line. Under the operating parameters selected for the JY 24 ICP instrumen- tation the lowest detection limit and the best values of relative standard deviation (RSD) were obtained for the line at 208.959 nm. The scans around the selected boron line are shown in Fig. 4. These scans correspond to the ‘blanks’ and to the calibration samples employed with both sampling systems.After having considered these scans values for background correction were measured at 208.925 nm (0.035 nm left of the boron peak). By using pneumatic nebulization blank values measured with background correction were approxi- mately half the total emission intensity. There are two components in the blank total emission a specific peak signal and an elevation of the background. Table 2 gives the analytical performance expressed in terms of background equivalent concentration (BEC) detection limit (DL) and precision. To calculate the DL (defined as the concentration of a solution which gives an absorbance equal to three times the SD of the blank) the following formula was used The BEC values (boron concentration giving a net analyte signal equal to the background signal) were 138 pg g-’ for pneumatic nebulization and 46 ,ug g-l for spark ablation.In the first system the BEC value was so high because of the 208.959 nm 0.01 nn i H 208.959 nm 208.959 nm 30 50 70 90 Time/s Fig. 3 Effect of sparking time on emission intensity of A Fe; B B using SS 45911; and C B using SS 45611 Fig. 4 Emission profiles. (a) Spark ablation 1 SS 45911; 2 SS 4561 1 ; and 3 Fe-0. (b) Pneumatic nebulization (0.5 g per 100 ml) I BCS-CRM 45911; 2 BCS-CRM 45611; and 3 Fe-0. (c) Solutions (100 ml) 1 0.5 g of Fe plus 0.7 ppm of B; 2 0.5 g of Fe plus 0.1 ppm of B; 3 0.5 g of Fe; and 4 blank830 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL.8 Table 2 Certified and found values for SS samples x=mean of six determinations; T=interval of uncertainty of the mean for a level of probability of 95%. T= t95,5 where t95,S = 2.57 1 Total Blpg g-I Samples SS 45611 SS 45711 SS 45811 SS 45911 SS 46011 CENIM * L=low calibration standard. i H = high calibration standard. Microwave digestion Spark ablation ~~ ~ Certified X T a 15 14.7 0.31 0.26 25 24.4 0.50 0.43 61 59.6 0.81 0.70 118 120.0 1.32 1.14 28 26.8 0.49 0.42 - 49.5 0.47 0.4 1 X T a Calibration L* 25.3 0.90 0.78 62.0 1.12 0.97 Calibration H t 27.3 0.85 0.74 48.7 0.90 0.78 background elevation produced by the iron matrix and of the elevation of the specific signal as a result of the dissolution process (as can be seen in Fig. 4). At levels five times higher than the corresponding DLs RSDs (n=6) of 1.3 and 3.5% were obtained using pneumatic nebulization and spark ablation respectively.Calibration After verifying the linearity of the boron emission intensi- ties versus boron concentrations within the interval of boron contents considered only two samples (‘low’ and ‘high’) were used to obtain the calibration graphs. Using pneumatic nebulization two calibration samples were prepared with 0.01 and 0.07 mg of boron added from a standard boron solution in the presence of 0.5 g of pure iron (Fe-0 sample) per 100 ml. These two calibration solutions are equivalent to steel samples containing 0.0020 and 0.0 14% of boron respectively. Two samples were used for calibration with spark ablation sample SS 456/ 1 (0.00 1 5% of boron) as ‘standard low’ and sample SS 459/1 (0.01 18% of boron) as ‘standard high’.The ICP calibration graphs obtained from pneumatic nebulization and spark ablation sampling systems are shown in Fig. 5. In these graphs the BCS-CRM and SS samples not used for calibration are interpolated by plotting their emission values against the certified concentrations. The correlation coefficients obtained with the linear regres- sion of all the data points were 0.9992 and of 0.9990 respectively. The results obtained from the above calibration graphs for the BCS-CRM and SS samples series and for a Centro Nacional de Investigaciones Metalurgicas (CENIM) sample are shown in Table 3. The mean values obtained with both methods for the boron content in the CENIM sample were compared by employing the Student’s t-test.This sample 1 !!? 1200 .- C 5. 1000 L 2 800 fJ 600 4- >. 400 200 c. .- - .- cn c I 1 I I I 1 0 0.002 0.004 0.006 0.008 0.010 0.012 0.014 - B (%I Fig. 5 Calibration graphs obtained using A pneumatic nebuliza- tion; x spark ablation; and U BSC-CRM (pneumatic nebuliza- tion) and SS (spark ablation) of samples was analysed first in disc form by spark ablation sampling and then by pneumatic nebulization of a dissolution of chips obtained from the disc sample. As the same number of determinations were carried out with the two methods (n = 6) mutual agreement was tested by the ratio XA -XB t= 0 2 + a2B JI1-l where x= mean; o= SD; v= degrees of freedom; ad ( 1 -a) is the probability level. For a probability level (1 -a)=0.95 and 10 degrees of freedom (v=2n-2= lo) in Student’s tables t = 2.228.The value calculated from figures in Table 3 t=2.026 is lower than the tabulated one. This means that the difference between the two means is statistically insignificant and can be explained by random errors alone. Conclusion The high pressure microwave digestion method is a valid system for complete dissolution of boron compounds in steels. The sample dissolution time is reduced from hours to 30 min. A further advantage was seen when six samples were dissolved simultaneously. As no evaporation occurred during the digestion process only a few millimetres of aqua regia were required as digestion reagent leading to low blank values. This allows lower DL to be attained than with conventional dissolution procedures. As a result of reduced sample handling and the impossibility of airborne contami- nation the risk of contamination is substantially reduced.By employing strong spark conditions (0 SZ 600 V 600 Hz) spark ablation can be used as a solid sampling system for the determination of boron in steels. The amounts of spark aerosol produced from boron steel surfaces prepared under the same conditions during a spark ablation ICP process are fairly similar and their elemental composition (B and Fe) remains stable and matches the bulk compo- sition reasonably well. Table 3 Analytical performance Sampling system BEC*/pg g-l DLtlpg g-l RSDS (0.5 g per 100 ml) 138 2.60 1.3 Microwave digestion Spark ablation 46 0.65 3.5 * Background equivalent concentration giving an emission signal equal to twice the total emission of the blank.t Detection limit producing an emission signal equal to three times the SD of the blank measured with background correction. $ Relative standard deviation (n=6) at five times the DL.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 83 I The ICP-AES technique with the two tested sampling systems could replace traditional chemical methods for boron analysis in steel samples. The drawback to the microwave digestion method is its higher background values and consequently its lower signal-to-background ratios and higher DL (2.5 times that achieved with spark ablation) however the precision values are better. The main obstacle to spark ablation sampling is the need to use solid standard samples for calibration. The accuracy of both methods is demonstrated by the results obtained for the CRMs analysed.The assays concerning development of the microwave sample procedure form part of a research project financially supported by the European Community for Steel and Carbon (ECSC) No. 7210/GD/936 (E2. 3/89). References Steel and Its Heat Treatment ed. Thelning K. E. Butter- worths London 2nd edn. 1984 p. 405. Ambrose A. D. Harine M. Staats G. and Weichert E. Steel Res. 1989 60 363. Thierig D. Fresenius'Z. Anal. Chem. 1982 310 154. 4 5 6 7 8 9 10 11 12 13 14 15 IS0 10153 199 1. International Organization for Standardiza- tion P.O. Box 56 CH- 12 1 1 Geneva 20 Switzerland. Introduction to Microwave Sample Preparation eds. Kingston H. M. and Jassie L. B. ACS Professional Reference Book Washington D.C. 1988. Kammin W. R. and Brandt M. J. Spectrosc. Znt. 1989,1 50. Gilman L. B. and Engelhart G. Spectrosc. Znt. 1990 2 16. Progress of Analytical Chemistry in the Iron and Steel Industry Commission of the European Communities. EUR 141 13 Luxembourg 1992 pp. 293-295. Broekaert J. A. C. Leis F. Raeymaekers B. and Zaray Gy. Spectrochim. Acta 1988 39 339. Mandoki A. Boron in Low Alloy Steels; Analytical Report Instruments S.A. Jobin Yvon Longjumeau 199 1. Lemarchand A. Labarraque G. Masson P. and Broekaert J. A. C. J. Anal. At. Spectrom. 1987 2 481. Prell L. J. and Koirtyohann S. R. Appl. Spectrosc. 1988,42 1221. Coedo A. G. Dorado L. T. Seco J. L. and Cobo I. G. J. Anal. At. Spectrom. 1992 7 11. Watters R. L. Jr. DeVoe J. R. Shen F. H. Small J. A. and Marirenko R. B. Anal. Chern. 1989,61 1826. Coedo A. G. and Dorado L. Rev. Metal. Madrid 1985 21 87. Paper 2/05430B Received November 9 1992 Accepted February 8 1993

ISSN:0267-9477    

DOI:10.1039/JA9930800827

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 833 Application of Ultrasonic Nebulization for the Determination of Rare Earth Elements in Phosphates and Related Sedimentary Rocks Using Inductively Coupled Plasma Atomic Emission Spectrometry with Comments on Dissolution Procedures* 1. B. Brenner and E. Dorfman Geological Survey of Israel 30 Malkhe Israel Street Jerusalem Israel 95507 Low concentrations of all the rare earth elements (REE) can be determined in geological materials using ultrasonic nebulization and inductively coupled plasma atomic emission spectrometry (USN-ICP-AES). Using a matrix matched calibration procedure to compensate for an acid effect on REE spectral line intensities caused by high concentrations of nitric acid in-house and international geological reference materials were analysed.Provided that a total decomposition method was used the data obtained compared well with those from the literature. Yttrium La Ce Eu and Yb were also determined directly because of their superior limits of detection relatively high abundance in geological samples and freedom from spectral line interference as a result of the high resolution of the spectrometers. Various sample decomposition procedures were evaluated for the determination of REE in sedimentary rocks. The solutions obtained by decomposition with various acids (HCI HNO HC104 and HF) and sodium peroxide sintering and leaching were preconcentrated using a DOWEX 50 W x 8 ion-exchange column. Significant differences in Y La and Ce concentrations occurred when the sediment samples were submitted to different methods of sample decomposition. Indeed REE distributions between acid soluble and total samples differed significantly having different Y:Ce Y:La and La:Ce ratios.Keywords Inductively coupled plasma atomic emission spectrometry; ultrasonic nebulization; rare earth elements; phosphates; sedimentary rocks The growing interest in the distribution of the rare earth elements (REE) Ce in particular in sedimentary rocks has increased owing to their application in the unravelling of sedimentary processes such as the role of redox conditions in ancient and modern basins of deposi- tion,'- contribution of terrestrial provenance^,^-^ and rock-water interactiow6 Redox conditions can lead to fractionation of Ce relative to the other REE resulting in significant deviations from the smooth REE chondrite distributions. The distribution of REE in the individual fractions of sediments (phosphorites siliceous ferru- ginous and calcareous phosphates bituminous shales carbonaceous clays and their acid insoluble residues etc.) can also vary significantly. The partition of REE in littoral sediments is controlled by the individual fractions of the sediments and might be fractionated during estuarine mixing.Thus the distribution of REE can be used as a diagnostic tool for determining the role of terrestrial detritus in deposition basins and for characterizing redox conditions in ancient and modern sedimentary basins of d e p o ~ i t i o n . ~ ~ ~ Owing to the fact that the REE concentrations in some of these fractions in particular those that contain high amounts of siliceous and organic material may be low it is necessary to preconcentrate the REE fraction and exclude the interfering matrix.Indeed the limits of detection (LODs) that can be obtained using conventional nebuliza- tion with inductively coupled plasma atomic emission spectrometry (ICP-AES) are inadequate for the determina- tion of the low concentrations of some of the REE. This disadvantage can be overcome by using enhanced methods of sample introduction such as ultrasonic nebulization (USN) which results in a 10-fold improvement in the LODs of the REE.' In the present investigation the superior LODs for the REE using USN-ICP-AES are applied to determine low REE concentrations in geological samples and decomposition residues.*Presented at the 1993 European Winter Conference on Plasma Spectrochemistry Granada Spain January 10- 15 1993. In previous investigation^^.^ it was reported that chemi- cally resistant residues were observed when geological materials were decomposed in a poly(tetrafluoroethy1ene) (PTFE) open dish using a mixture of HF-HC104. These residues consist mainly of Zr- and Ti-bearing minerals and were subsequently decomposed using alkali fusions. Shol- kovitz demonstrated that the HF dissolution excluded a significant portion of the heavy REE (HREE Ho-Lu) whereas this fraction was decomposed using LiBOz fusion. In contrast fusion and HF dissolution techniques yielded similar concentrations for the light (LREE La-Nd) and middle REE (Sm-Dy).He concluded that shelf and slope marine sediments are more similar in REE composition to shales than previously rep~rted.~ Sholkovitz3 stressed the importance of these fractions in the interpretation of fluviatile influxes oceanic abundances and the behaviour of REE in diagenesis. Therefore the determination of REE in small amounts of residues is an additional analytical challenge when using ICP-AES. In the present study several methods of decom- position for the determination of the REE were evaluated using the following geological reference materials Cana- dian Certified Reference Material Project (CCRMP) SY-2 and SY-3 Silicate Rock; Community Bureau of Reference (BCR) 32 Morrocan Phosphate Rock; National Institute of Standards and Technology (NIST) Standard Reference Material (SRM) 120B Phosphate Rock (Florida) and 694 Phosphate Rock (Western); South African Bureau of Stan- dards (SABS) SARM 1 Granite SARM 2 Lujavrite and SARM 3 Syenite; and in-house reference materials.The following decomposition routines were studied HC1; HN0,-HC10,-HF-HCl; and sodium peroxide sintering and cold leaching. The resultant solutions were analysed by simultaneous and sequential ICP-AES after the major elements and fusion cations were excluded and the REE concentrated using a cation-exchange procedure described by Brenner et al.' and Watkins and N01an.~ Solutions were also analysed directly for trace and major elements which included La Y Yb Eu and Ce using procedures described by Brenner et a1.lo834 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 Experimental Decomposition Procedures Several dissolution routines were applied to several in- house and international geological reference materials phosphorites ferruginous and siliceous phosphates bitumi- nous and calcareous shales and silicate samples.Hydrochloric acid A sample weighing 2 g was stirred continuously with 75 ml of 2 mol 1-l HCl for 1 h at room temperature then filtered through a Whatman 40 filter-paper into a 50 or 100 ml calibrated flask. Solutions were made up to volume with 2 mol 1-l HCl. The residues were ashed overnight at 600 "C; the ashes were then digested to dryness in a Pt dish at 300 "C with 15 ml of HF and 10 ml of HN03. The salts were then dissolved in 2 mol 1-l HCl and made up to 50 ml.Calcareous and phosphatic fractions of the samples were dissolved using this procedure. Sodium peroxide sinter A sample weighing 0.5 g was decomposed by sintering with 2 g of sodium peroxide in a graphite or Zr metal crucible in an oven at 500 "C for 30 min. After cooling the sintered mass was disintegrated by treatment with water and the mass transferred into a beaker containing 25 ml of HCl dissolved and made up to 100 ml. The procedure used has been described previously.1o With this technique a total analysis was obtained. Mixed acid (hydrochloric-nitric-perchloric-hydrofluoric) procedure A sample weighing 1 g was treated with 30 ml of HF (38-40% v/v) and 10 ml of concentrated HN03 in a platinum or Teflon dish. An additional 25 ml of concen- trated HN03 were added after the mixture was heated to dryness.Residues were then heated consecutively with 20 ml of HCl and 20 ml of HC104 until the solution was clear. Heating was continued to dryness to remove perchloric acid. After cooling the salt residue was dissolved in 10 ml of 2 mol 1-l HC1. With this technique a total analysis was obtained. However resistant REE-bearing minerals might be only partially Cold sodium peroxide extraction This procedure was employed for the decomposition of samples containing high contents of organic matter. A 10-20 g sample of bituminous shale or carbonaceous clay was mixed for 24 h with 150 ml of a 10% m/v solution of sodium peroxide in order to decompose selectively the organic-bound fraction. The soluble fraction separated by centrifugation was evaporated to about 20-30 ml then treated by boiling with 50 ml of HN03 and 25 ml of H202 until the solution was colourless.The heating was continued to dryness and the solid residue was dissolved in 50- 100 ml of 2 mol 1 - l HCl. Hydrofluoric acid Chert samples containing very high amounts of silica were treated with HF in order to remove the silica and to extract the REE bound to t h e inorganic fraction. A sample weighing 10 g was treated with HF-HC104 (2+ 1 v/v) and the solution was heated to dryness. The residue was dissolved in 25-50 ml of 2 mol 1-l HCl. REE Chromatographic Separation The procedure employed is similar to that used by Watkins and Nolan9 and Brenner et aL7. Sample solutions produced from the decomposition procedures were loaded onto a column of DOWEX 50 Wx8 ion-exchange resin pre- equilibrated with 2 mol 1-l HCl.Matrix elements (Ca Mg Na K Fe and Al) were eliminated using 100 ml of mixed acid 3 mol 1-l HN03-2 mol 1-l HC1 (3+ 1) followed by :25-50 ml of 2 moll-' HC1. The REE eluted with 100 ml of '7.2 mol 1-l HN03 were determined by simultaneous and sequential USN-ICP-AES. Several REE (Y La Eu Yb and Ce) were also determined directly using the polychromator and a multi-element routine designed to determine the major minor and trace elements.lo REE Calibration Multi-element REE calibration standards were prepared by stepwise dilution of stock solutions prepared from Specpure REE oxides (Johnson Matthey UK). A graded series of calibration standards were prepared in which the LREE concentrations exceeded the concentrations of the HREE in accordance with their abundances in the materials under investigation (1-500 pg 1-l).This calibration scheme was usually adequate for most of the geological materials investigated. In certain cases calibration ranges were ex- tended or samples diluted. The step involving removal of HN03 was avoided. However the USN is sensitive to changes in the acid content" (Fig. 1). Consequently a matrix match calibration procedure was adopted where the standards were accurately matched to the samples by adding 7.2 mol 1-' HN03. 1.05 0 1 2 3 4 5 6 7 8 [HNOJmol 1.' Fig. 1 Effect of HN03 on REE (1 mg I-') spectral line intensity using ultrasonic nebulization with desolvation A Ho; B Y; C Gd; D Yb; E La; and F Lu Instrumentation and Operating Conditions A high resolution 1 m JY 38 sequential system was employed for the determination of all the REE.A JY 48,48 channel polychromator was employed for the simultaneous determination of Y La C Yb and Eu. The lack of significant spectral line interferences is attributed to the high resolution of the sequential spectrometer (6 pm in the first order). A Cetac USN was employed. The instrumental configuration and operating conditions are listed in Tables 1 and 2.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 835 Table 1 Instrumentation Monochromator Grating Dispersion Polychromator Grating Dispersion R.f. generator Torch Jobin Yvon JY 38 3600 grooves mm-I range 180-490 nm 0.27 nm mm-l 6 pm in the first order (25 pm slit-widths) Jobin Yvon JY 48 1 m 2550 grooves mm-l spectral range 170-450 nm 0.35 nm mm-I PlasmaTherm 2.5 kW < 10 W reflected Jobin Yvon Ryton demountable Results and Discussion Limits of Detection The spectral lines employed were listed in a previous investigation.' The LODs were determined by calculating the concentration of the analyte that yielded a signal twice the standard deviation of the blank signal (n = 10).Limits of detection of USN-ICP-AES are compared with ICP-MS and other ICP-AES literature values in Table 3. The similarity of the USN-ICP-AES LODs to the ICP-MS values quoted by JarvisI3 is noteworthy; Ce Pr La and Tb are 10-fold better using ICP-MS whereas Lu is 10-fold better by USN- ICP-AES. Taking into consideration that implementation of these LODs using ICP-MS is dependent on the limita- tions of total salt and acid concentrations in ICP-MS the performance of USN-ICP-AES is commendable.Quantification limits in the solid samples were calculated based on a 1 +49 dilution (superior to those reported by Jarvis and JarvisI2 and permit the determination of all the REE in most types of geological materials. These levels are a prerequisite when REE are to be determined in siliceous and bituminous materials and silicate rocks such as SARM 2 syenite. They can be improved by using larger amounts of sample. Precision and Accuracy In general the relative standard deviations (RSDs) (stan- dard deviatiodmean x 100 n= 5-1 0) varied as a function of concentration (Fig. 2) and at 1-100 mg kg-I were about lo% except for concentrations that were near the limit of quantification (RSDs 10-25%).For concentrations greater Table 2 ICP and USN operating conditions Pneumatic nebulizer Nebulizer gas flow rate/l min-l Ultrasonic nebulizer Desolvation temperature/"C Cooling Heating Outer Intermediate Sheath (Trassy-Mermet) Pneumatic USN Gas flow rate/l min-' Aerosol carrier gas flow rate/ml min-l Washout periodls Pneumatic (Meinhard) USN Integration period/s Pol ychromator Monochromator Meinhard TR-C-20 1.2 Cetac U 500AT 45 psi - 5-0 140 14 0.2-0.4 0.2 0.1 0.95 30 50 5 0.5 ~~~~~ ~~~ ~~~ than 100 mg kg-l RSDs improved with increasing concen- tration to < 1%. In order to estimate the accuracy of the present proce- dures several international reference materials were de- composed in triplicate employing sodium peroxide sinter and HF-containing mixed acid decomposition procedures to furnish total REE contents. It should be noted that all REE were determined Tm and Tb included which are usually too low to be determined by ICP-AES using a conventional sample introduction system.Recommended value^'^-'^ listed in Table 4 are in good agreement with the published values. In general the data have a variation of < 10% when compared with recommended values. Values are also listed for SARM 2 which has very low REE contents. In this case the dilution factor was 1 +24. Unfortunately the precision of the determinations of SARM 2 varied from 10-35%. The efficiency of the ion-exchange columns for quantita- tive preconcentration of the REE was ensured by analysing the original decomposition solutions for Ce and Yb (a LREE and a HREE respectively) using the JY 48 simulta- neous spectrometer. The correlation between these values and those obtained by column preconcentration and sequential USN-ICP-AES are illustrated in Figs.3 and 4. Table 3 LODs (20 obtained by ICP-MS USN-ICP-AES and ICP-AES using conventional nebulization; all data in pg 1-1 Element Lu Tm Y Gd Ho Tb DY Sm Er Yb Eu La Nd Pr Ce W avelengt h/nm 26 1.52 31 3.126 37 1.03 342.246 345.6 350.9 17 353.17 1 359.26 369.265 369.41 9 38 1.97 398.852 406.66 41 7.939 4 18.66 ICP-MS (Ref. 13) 0.05 0.01 0.1 0.1 0.04 0.03 0. I 0.2 0.06 0.06 0.06 0.08 0.2 0.09 0.1 Meinhard (Ref. 7) 0.08 0.3 0.4 0.3 0.2 2.3 0.1 0.9 0.6 0.4 0.5 1 2.1 1.6 5.9 USN (Ref. 7) 0.008 0.06 0.30 0.09 0.05 0.3 0.04 0.3 0.15 0.03 0.06 0.35 0.5 0.8 1.2 ICP-AES (Ref.1 2) 0.2 1.5 2. l t 0.56 0.65 1.8 l . l t 0.25t 0.26 2.6 5.2t 3-87 8.7 - - Quantification limit*/pg kg-* (USN) 4 30 150 45 25 150 20 150 75 150 300 175 250 400 600 *Quantification limit is 10 x LOD. Data are reported as concentrations in the solid sample based on a 1 +49 dilution. ?Different spectral line.836 JOURNAL OF ANALYTI.CAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 Table 4 Total REE concentrations in reference materials determined (Det) by ICP-AES and USN; data in mg kg-l RSDs (O/O) are based on five separate determinations. Recommended data (Rec) are from Jarvi,s1*.l3 and G~vindaraju'~ Element Parameter Y La Ce Pr Nd Sm Eu Gd Tb DY Ho Er Tm Yb Lu Det Oh RSD Rec Det % RSD Rec Det Oh RSD Rec Det Oh RSD Rec Det O/o RSD Rec Det Yo RSD Rec Det Yo RSD Rec Det O/o RSD Rec Det % RSD Rec Det Oh RSD Rec Det O/o RSD Rec Det % RSD Rec Det Oh RSD Rec Det Oh RSD Rec Det % RSD Rec NIST 120B 167 5 172 95 5 88 115 4 115 18 8 17.4 75 6.8 75 17 15 16 3.8 10 3.6 19.5 13 18.9 2.2 15 2 16.3 13 17.2 3.8 13 3.9 1 1 8 11.7 1.35 14 1.1 10.9 7 10.9 1.85 12 1.7 BCR 32 268 6 262 117 16 115 33 15 35 13.8 11 60 8 68 14 8 12 3.5 11 3.5 19 12 19.5 2.1 15 3.3 16.5 10 4 10 14 12 - - - - 1.85 25 2 14.4 2 14.2 2.5 4 2.6 NIST 694 189 20 125 128 5 103 26 20 25 14 2 70 5 41 12 2 4.9 3 25 1.5 16 15 - - 1.6 7 1.2 15 10 9 3.5 9 11 6 1.2 3 9 4 7.2 2.1 0.97 - - - 17 SY-2 90 4 117 75 10 66 168 5 161 18 4 18.2 76 8 72.9 20 8 14.8 2.6 11 2.62 15.7 4.5 15.4 2.4 13 18.7 6 18.4 4.3 5 4.29 13.1 1 12.5 2.1 6 16.6 3 16.5 2.7 5 2.7 - - SY-3 64 5 0.8 685 1345 1 1330 2200 3 2340 170 2 225 710 15 749 134 6 126 18 9 20.5 126 9 112 14.2 8 114 12 130 21 2 28.7 67 7 76.8 9.3 8 63 9 67.2 7.9 11 8.3 - - SARM I 145 2 130 95 3 105 220 3 200 20 5 21.7 70 4 73 21 7 16 0.44 23 0.4 14 12 13.7 2.4 20 2.3 18 9 15 4.2 16 3.7 17 8 13 2.3 20 2 15 11 14 20 2.1 1.6 SARM3 SARM2 27 9 22 285 2 50 285 3 270 18 15 19 48 7 48 5.7 16 5 1.4 22 1.2 3.4 17 4 3.4 - - - 2.9 3.1 0.8 0.6 1.6 2.2 3.4 17 3 2.5 13 3 0.44 31 0.4 15 25 21 1.2 2.5 4.5 10 5 13 10 11.9 1.5 15 1.18 6.5 10 6 1 18 1 0.3 0.3 0.9 0.83 0.16 0.1 0.5 0.4 0.06 0.05 0.13 0.12 0.06 20 - - 20 20 - 20 - - 0.1 0.06 0.1 0.0 1 30 30 The correlation coefficients for Ce and Yb are 0.99636 and 0.998 1 respectively (n = 2 1 and 14 respectively) indicating that these elements can be determined directly without preconcentration.This procedure was employed to cross check data obtained by USN-ICP-AES and to ensure column performance. Reference materials NIST SRM 120B Phosphate Rock (Florida) and BCR 32 Morrocan Phosphate Rock were treated using the decomposition procedures described previously. The effect of sample decomposition on the partition of the LREE is illustrated in Figs. 5 and 6. Evidently the Y La and Ce values for the sodium peroxide sintering procedure are maximum. Thus these data indi- cate that significant differences in Y La Ce and Nd concentrations can also occur when sediment samples are submitted to different modes of sample decomposition. Preliminary results on the partition of the REE in the various fractions of the samples investigated are listed in Table 5 and Fig.7. Several phosphate and shale samples were treated with HCl and sodium peroxide and the residues analysed. While the interpretation of these trends is beyond the scope of this article it is obvious that the acid insoluble and organic fractions have different REE distribu- tions from those of the whole samples. These patterns could be controlled by the mineralogy of the fractions derived from different parent rocks alteration processes leading to REE fractionation secondary phases that preferentially accommodate REE etc. It is clear that the present technique is capable of determining very low REE concentrations. Nevertheless at the present time the precision of determination of the HREE were inadequate to resolve the question of whether they were also discriminated.Conclusion The advantage of the USN lies in obtaining superior LODs permitting small amounts of sample to be analysed. TheseJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 837 loo c c3 v) K 1 000 0 0 0 00 El 000 0 0 0 0.1 1 I I I 1 10 100 1000 [REEI/ mg kg-' Fig. 2 Relation between RSD (%) of determination and REE concentration 1 g of solid sample dissolved in 100 ml 0 Y 0 0 F Brn - Q I I 1 I I 1 10 100 1000 E [Cel (sequential USN-ICP-AES)/ mg kg" Fig. 3 Correlation of sequential USN-ICP-AES (with column preconcentration) and simultaneous direct determination of Ce I1 4 18.66 nm using ICP-AES 0 0 10 100 [Ybl (sequential USN-ICP-AES)/ mg kg-' Fig.4 Correlation of sequential USN-ICP-AES (with column preconcentration) and simultaneous direct determination of Yb I1 369.419 nm using ICP-AES include residues that were produced by treating the samples with selective extractions and those that are chemically resistant. Furthermore all the REE including Tb and Tm can be determined and samples containing low REE contents can be analysed. 200 150 0 Y 0 E 100 E w w 50 0 HCI a Mixed acid / Y La Ce Nd Fig. 5 REE concentrations in NIST 120 B Phosphate Rock (Florida) treated with various decomposition procedures ' /I HCI 300 a Mixed acid 250 69 Recommended yo) 200 Y 150 h w w E 100 50 O V f f f / Y La Ce Nd Fig. 6 REE concentrations in BCR 32 Morrocan Phosphate Rock treated with various decomposition procedures Phosphate - Bituminous shale Fig.7 Y:La Y:Ce and La:Ce ratios in various fractions of phosphates (P) bituminous shales and cherts from southern Israel; Res = residue Tot = total Org = organic fraction and NASC = North American Shale Composite838 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 Table 5 Y La and Ce concentrations in various fractions of phosphates bituminous shales and cherts from southern Israel Sample Y NIST 120 Phosphate Rock Residue* 25 Total? 167 Residue 24 Total 2 60 BCR 32 Phosphate Rock IB-7 Bituminous Shale Residue Organic$ Total Residue Total IB-8 Phosphorite 1.25 0.005 45 0.6 45 IB-9 Bituminous Phosphorite Residue 2.2 Organic Total 62.7 IB-10 Chert Residue Total 0.02 2.2 North American Shale Composite 27 La 8.6 94. 3 124 0.8 0.007 1 7 0.3 1 '7 0.4 0.0002 21 - 10.9 32 Ce 3.8 116 4 35 0.45 0.008 6 0.6 8 0.2 0.0002 9 0.2 0.8 73 Y:La Y:Ce 2.9 6.6 1.8 1.4 8.0 6.0 2.1 7.4 1.6 2.8 0.7 0.6 2.6 7.5 2.0 1 .o 2.6 5.6 5.5 11.0 3.0 7.0 - 0.1 2.4 2.8 0.8 0.4 - - La:Ce 2.3 0.8 0.8 3.5 1.8 0.9 2.8 0.5 2.1 2.0 1 .o 2.3 - 1.1 0.4 *Residue HCl insoluble residue.tTotal sodium peroxide or mixed acid procedure with HF. $Organic fraction soluble in sodium peroxide solution. A comparison of several dissolution procedures for the determination of REE in phosphates and related sedimen- tary rocks indicated that satisfactory data were obtained when a reliable total decomposition procedure such as a sodium peroxide sinter or a mixed HN03-HC104-HF-HCl procedure was employed. Data obtained from selective dissolution procedures indicate that significant concentra- tions of the REE are also located in the insoluble silicate fraction.The data produced in this study indicate that the REE concentrations in bituminous and siliceous fractions are very low. This study was sponsored by the Earth Science Administra- tion Ministry of Energy and Infrastructure Israel. Their financial assistance is sincerely appreciated. The authors are indebted to Dr. A. Bein Director of the Geological Survey of Israel for his constructive discussions. 0. Joffe per- formed cross check REE determinations using simulta- neous multi-element analysis. The assistance of I. Segal and R. Binstock in instrument operation was most valuable. T. Minster supplied a sample of bituminous shale for which we are thankful. References 1 De Baar H. J. W. German C. R. Elderfield H. and Van Gaans P. Geochim. Cosmochim. Acta 1988 52 1203. 2 Elderfield H. and Sholkovitz E. R. Earth Planet. Sci. Lett. 1987 82 280. 3 Sholkovitz E. R. Chem. Geol. 1990 88 333. 4 Sholkovitz E. R. Am. J. Sci. 1988 288 236. 5 Murray R. W. Buchholtzen-Brink M. R. Gerlach D. C. Russ G. P. 111 and Jones D. L. Geochim. Cosmochim. Acta 1991,55 1875. 6 Smedley P. Geochim. Cosmochim. Acta 1991 55 2767. 7 Brenner I. B. Binstock R. Dorfman E. and Halicz L. ZCP Inf Newsl. 1992 18 473. 8 Brenner I. B. Watson A. E. Steele T. W. Jones E. A. and Goncalves M. Spectrochim. Acta Part B 1981 36 785. 9 Watkins P. J and Nolan J. Chem. Geol. 1992 95 131. 10 Brenner I. B. and Eldad H. ICP In Newsl. 1986 12 243. 11 Brenner I. B. Segal I. Long G. and Dorfman E. presented at the 1993 European Winter Conference on Plasma Spectro- chemistry Granada Spain January 10- 1 5 1993. Paper No. 12 Jarvis K. E. and Jarvis I. Geostand. Newsl. 1988 12 1. 13 Jarvis K. E. Chem. Geol. 1990 83 89. 14 Govindaraju K. Geostand. Newsl. 1989 13 1. P 1-24. Paper 3/0 1 9 70E Received April 6 1993 Accepted May 5 1993

ISSN:0267-9477    

DOI:10.1039/JA9930800833

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 839 Analysis of Glasses From the V,O,-As,O,-BaO System Using Inductively Coupled Plasma Atomic Emission Spectrometry* S. Del Barrio and R. Benito lnstituto Tecnologico Geominero de Espafia Rios Rosas 23 Madrid 28003 Spain F. J. Valle lnstituto de Ceramica y Vidrio (CSIC) 28500 Arganda del Rey Madrid Spain Glasses from the V20,-As,O,-BaO phase equilibrium system are interesting for applications in the production of thermistors and electron multiplier channels. The high volatilization rates of arsenic during the melting process make it necessary to carry out chemical analysis of the glasses in order to locate the final composition in the glass-formation area in the corresponding phase equilibrium system. Inductively coupled plasma atomic emission spectrometry is proposed for determining V205 As~O and BaO.A sodium hydroxide procedure was used to separate the main components from the impurities. The only spectral interference produced is due to the V I emission at 193.682 nm which overlapped partly with 193.698 nm of As 1. The analyses were carried out by using an internal standard Sc II at 361.384 nm. The determination of analytical short-term precision and precision of the method was also carried out. The relative standard deviations were 0.53,1.19 and 0.80 for V205 As,O and BaO respectively. The results obtained were compared with those obtained using gravimetry for As and Ba and spectrophotometry for V. Keywords Inductively coupled plasma atomic emission spectrometry; glasses; determination of V,O As203 and BaO The As203 introduced into a glass chemical composition can act either as a refining agent or network former,' and is present in the compositions of the V20,-As203-BaO phase equilibrium system for the latter purpose.A research project developed at the Spanish Instituto de Cerdmica y Vidrio (CSIC) since 1978 has been to determine precisely the glass building area within the above mentioned system (Fig. 1) and to study the future application of such materials to the production of thermistors electron photomultiplier channels and low-valency vanadium oxides2 During the research the necessity of knowing the composition of glasses in order to predict their properties has arisen. The percentages of the components expressed as oxides shift significantly from the theoretical values because during the fusion stage volatilization takes place that is difficult to control particularly for As203 and less appreciably for V205.Thermogravimetrically arsenic behaves in different ways A Glass building p area Mol(%) Fig. 1 Tentative glassy state area in the V20,-Asz0,-Ba0 system; 0 glass and e devitrified glass ~ ~ ~~~~~~~~ ~ ~~ *Presented at the 1993 Winter Conference on Plasma Spectro- chemistry Granada Spain January 10- 1 5 1993. 1 I I 1 0 2 0 0 1 0 0 6 0 0 8 0 0 VC Fig. 2 Thermogravimetric curves for A As203 and B M-5 glass (69% V2O5-23% A~~O3-896 BaO). The total volatilization of the As203 sample (0.17023 g) cannot be recorded as a single line because of the loss-in-mass scale used (100 mg).At 405 "C the top of the scale is reached depending on whether it is present as raw material (As~O~ As2S3 As2S5 etc.) or if it is inside the glass network. In the thermogravimetric plots presented in Fig. 2 both behav- iours can be observed. Line A corresponds to As~O~; a sample with 0.1 7023 g of As203 is fully volatilized at 600 "C with a heating rate of 6 "C min-I. Line B corresponds to a V205-As203-BaO glass containing 24.3 mol-Oh As203; a 0.14389 g sample of this material under the same heating conditions starts to volatilize only at temperatures >750 "C. This behaviour has great importance in the chemical analysis of V20S-As203-BaO glasses because it enables its disgregation with melting reagents when the working temperature is less than 750 "C e.g.using sodium hydroxide as a fusion reagent (attack temperature between 400 and 500 0C).3 The aqueous solution of the attacked sample contains the major components (V as V03- or V0.,H2 As as As2- or As03- Ba as Ba(OH) soluble) and840 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 Table 1 Specification of the ICP and operating conditions Plasma source FrequencyIMHz Induced power/W Reflected power/W Argon plasma gas flow rate/l min-' Argon aerosol camer gas flow ratell min-l Cooling water for coils/l min-I Sample introduction Sample delivery Measurement time Observation heigh t/mm Analytical lines/nm v I1 As I Ba I1 s c I1 27.12 MHz 1000 t 5 17 0.75 1.5 Pneumatic cross-flow nebulizer Gilson Miniplus 11 peristaltic pump 2 ml min-I 10 integrations of 5 stages per step 16 292.402 193.696 4 9 3 -409 36 1.384 the impurities (Al Fe Mn Ti Cr etc.) are separated as precipitated hydroxides or basic salts in the alkaline medium.The objectives of this work are to use inductively coupled plasma atomic emission spectrometry (ICP-AES) for the determination of the macroconstituents of glasses belonging to the V20,-As203-Ba0 system after sample fusion with sodium hydroxide; and to compare the results with those obtained using other well-known techniques for these materials (As gravimetry as Mg2As20,; Ba gravi- metry as BaSO,; V spectrophotometry as VOZ3+). Experimental Instrumentation A Jarrell-Ash Model ICAP-6 1 multichannel spectrometer with an ICP source was used equipped with a polychroma- tor (1 5 10 grooves mm-l and 32 photomultiplier tubes).The specifications of the ICP source and operating conditions are given in Table 1. Reagents Standard stock solutions for V As Ba and Y were used (Inorganic Venture). Sodium hydroxide (NaOH-H,O Suprapur Merck) was used for the sample fusion. High- purity HN03 (65%) and de-ionized water ( > l 8 mQ cm) were used for all preparations of standard and sample solutions. Sample Preparation and Procedure All theoretical compositions of glasses used in the present work are reported in Table 2. The sample prepared for analysis was ground in a tungsten carbide vibrating mortar to a particle size of less than 63 pm. A 0.2000 g of the ground sample was placed in a zirconium crucible and mixed with 0.50 g of sodium hydroxide (carbonate free).The mix was covered with a further 0.50 g of sodium hydroxide and transferred to a muffle furnace. The temperature was gradually raised until any water present was driven off and then held at 400 "C for 15 min until complete fusion occurred. The crucible was removed from the furnace and allowed to cool and 20 ml of de-ionized water were added. The crucible was transferred to a steam-bath to facilitate lixiviation of the melt which was filtered through a No 58g3 Scheleicher blue ribbon paper. The precipitate (impurities) was transferred to the filter with a jet of hot water and washed thoroughly with hot water. The filtrate was completely transferred into a 200 rnl calibrated flask acidified with 5 ml of HN03 (65%) and diluted to volume with de-ionized water.Standard Preparation and Calibration Five multi-element standards of 100 50 25 10 and 5 pg ml-I of V and 20 10 5 2 and 1 pg ~ m - ~ of As and Ba were prepared from 1 mg ml-l stock solutions and the blank of the reagents. These concentration ranges cover all three element contents in the analysed materials after diluting the original solutions 1 +9 and 1 + 19. Scandium (50 pg ml-l) was added to all diluted sample and standard solutions as an internal standard. - ~~~ ~ ~ ~ Table 2 Composition of selected glasses Theorical composition Glass sample Oxide M-2 v205 As203 BaO M-3 v2os As203 BaO M-4 v 2 0 5 A m 3 M-5 v 2 0 5 M-6 v 2 0 5 A N 3 BaO AS203 BaO BaO Mol (Oh) 85.0 8.5 6.5 70.5 12.0 17.5 80.0 12.0 8.0 69.0 23.0 8.0 60.0 28.0 12.0 Mass (%) 84.50 12.98 2.5 1 71.71 13.27 15.0 80.16 13.08 6.75 68.48 24.82 6.69 59.65 30.28 10.05 Table 3 Precision for V As and Ba with regard to sample dilution Glass sample Element Dilution M-2 V 1+9 As 1 +9 Ba 1 +9 M-6 V 1 +9 As 1 +9 Ba 1+9 1+19 1+19 1+19 1+19 1+19 1+19 RSD 0.18 0.30 0.32 0.49 0.18 0.27 0.23 0.29 0.36 0.54 0.20 0.27JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 84 1 ~ ~~ Table 4 Analytical results for macroconstituents in glasses of the V205-As203-Ba0 system Mass* (O/O) Spectrophotometry Gravimetry Gravimetry Glass Oxide ICP-AES V023+(H202) As207Mg2 BaSO 85.10 k0.35 85.61 k 0.80 - - - 9.95 k 0.26 - 10.17 k 0.14 71.35k0.39 71.60k0.85 - - - 10.87+0.30 - 11.04k0.14 M-2 v2°5 BaO 4.2 I k 0.05 - - 4.59k 0.19 As205 M-3 VZOS As205 BaO 16.90t0.10 - - 17.2 1 t0.28 M-4 v2os M-5 VZ05 AS205 BaO As205 BaO M-6 v 2 0 5 AS205 BaO 80.73 k 0.35 11.10k0.17 7.98 k0.06 69.87 t 0.40 20.69 k 0.2 1 8.90 t 0.07 59.86 k 0.42 24.31 k0.19 14.98It0.11 8 1.03 k 0.91 - - - 10.93-tO.36 - - - 8.43 k0.16 - 20.60 k 0.4 1 - - - 9.30 k 0.12 59.95 t 0.65 - - - 23.85 +- 0.40 - - - 15.29 k 22 - - 70.13 t 0.79 *Mean k standard deviation for the results of the chemical analysis of ten different samples taken from the same original specimen.1395 1 Time - Fig. 3 Wavelength scans in the vicinity of As I 193.696 nm A As 12.5 pg ~ m - ~ ; and B V 100 pg ~ m - ~ The linear calibration curves obtained were sufficiently accurate as shown by the correlation coefficients for the regression lines V r=0.9997; As r=0.9996; and Ba r=0.9995. Results and Discussion Spectral Interferences The proposed fusion process with sodium hydroxide as it suppresses metallic element impurities eliminates the spectral interferences of the impurities on the analyte emission lines.One exception is the V line at 193.682 nm which partially overlaps the As line at 193.696 nm (Fig. 3). The spectral interference was corrected on the basis of the equation where c is the real concentration of As caAs the apparent concentration of As cv the concentration of V and kAs,v the interelement correction coefficient of V for As. The kAs,v was determined by measuring the emission intensities at the specified wavelength for As using the single element solutions 10 pg ml-I of As; 50 pg ml-I of V; and 100 pg ml-I of V. Thus the value for the kAs,v was found to be 0.09.Effect of Dilution on the Precision of Results The optimum dilution working conditions for the determi- nation of V As and Ba that led to the best precision are summarized in Table 3 which also includes the relative standard deviation (RSD) values obtained for 20 ratio intensity measurements (short-term precision). It is in- ferred that the highest precision (lowest RSD) is obtained by diluting the sample solution (1+9) for the three elements. Analytical Results The chemical analysis was carried out for each glass on ten samples taken from the same original material following the optimum working conditions reported in Table 3. Table 4 shows the results of ICP-AES compared with the reference techniques for the five analysed glasses including the standard deviation (a) corresponding to the precision of ten analysed samples (precision of the method).The analytical results obtained using the three techniques are fairly acceptable and permit the location of five com- positions within the glass formation area in the V20J- As203-BaO phase equilibrium system. The method precision is significantly better using ICP-AES. By comparing data in Tables 2 and 4 it becomes clear that during glass production As203 volatilizes consider- ably V205 volatilizes to a lesser extent but BaO is not volatilized. If V205 were not volatilized its content in the glass would be higher than the theoretical value because of the need for compensation of As203 losses. Conclusions During the production process of glasses belonging to the V205-As20,-Ba0 system some volatilization of As203 and V205 takes place which needs to be controlled in order to determine precisely the glass building area. To reach this objective ICP-AES offers results comparable with those of other accepted techniques for the determination of V As and Ba.842 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 References 3 Voinovitch I. A. Debras-Guedon J. and Louvrier J. L 'Analyse des Silicates Hermann Paris 1968. 1 2 Fernandez Navarro J. M. El Vidrio Consejo Superior de Investigaciones Cientificas Madrid 1991. Jurado. J. R. Thesis Universidad Complutense de Madrid Paper 3/01 7426 Received March 26 1993 1978. Accepted May 14 1993

ISSN:0267-9477    

DOI:10.1039/JA9930800839

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRYy SEPTEMBER 1993 VOL. 8 843 Determination of Nickel in Biological Samples by Inductively Coupled Plasma Atomic Emission Spectrometry After Extraction With 1,5-Bis[ phenyl-( 2-pyridyl)methylene]thiocarbonohydrazide* E. Vereda Alonso A. Garcia de Torres and J. M. Can0 Pavon Department of Analytical Chemistry Faculty of Sciences University of Malaga 29071 Malaga Spain A sensitive inductively coupled plasma atomic emission spectrometric method for the determination of trace amounts of nickel after extraction of the metal into isobutyl methyl ketone (IBMK) containing 1,5-bis[phenyl(2- pyridyl)methylene]thiocarbonohydrazide (BPTH) is described. A systematic study was made to determine the optimum conditions for extraction of the metal into IBMK.The chief advantage of the method lies in the maximum allowable aqueous-to-organic phase volume ratio of 37. The detection and quantification limits for nickel were 0.3 and 1.1 ng ml-l respectively and the calibration is linear from 0.4 to at least 150 ng ml-l. The method has been used for the determination of nickel in various biological samples [mineralized by microwave digestion in sealed poly(tetrafluoroethy1ene) containers] as well as in human urine with good results. Keywords Solvent extraction; inductively coupled plasma atomic emission spectrometry; nickel; biological samples Inductively coupled plasma atomic emission spectrometry (ICP-AES) has been widely recognized as a suitable tech- nique for the determination of trace elements the particu- lar advantages being the multi-element capability large dynamic range and effective background correction.How- ever several problems have been indicated by researchers e.g. spectral interferences owing to matrix components nebulizer blockage owing to the high solids content of the solution or analyte emission enhancement. In addition ICP-AES has a significant drawback because the detection power is sometimes inadequate to comply with require- ments for the determination of trace elements in biological samples. For this reason liquid-liquid extraction is one of the most frequently used sample pre-treatment techniques for the determination of trace metals by ICP-AES.'-3 The extraction serves the dual purposes of concentrating the metals of interest and separating them from an interfering matrix.The extent of the concentration achieved depends on the ratio of the aqueous-to-organic phase volume. On the other hand isolation from the matrix significantly de- creases any background signal caused by concomitants and the solids content of the solution. The relevance of nickel has grown markedly over the last few years as a result of toxicological and physiological research on this element from the health point of view."8 The fact that nickel can cause allergies and in some of its compounds is carcinogenic has led to the appreciation that monitoring nickel levels in biological materials is of great importance. As under normal conditions nickel occurs in body fluids and tissues at very low concentrations (usually the ng ml-I level) there is an increasing demand for suitable sensitive and selective analytical methods.In the present work an ICP-AES method is described for the determination of trace amounts of nickel in biological materials after extraction of the metal into isobutyl methyl ketone (IBMK) containing 1,5-bis[pheny1-(2-pyridyl)me- thylene]thiocarbonohydrazide (BPTH). The complex formed is soluble in IBMK so much so that it allows the use of aqueous-to-organic phase volume ratios of up to 37 ( i e . much higher than those typically afforded by other extrac- tants) and hence the determination of concentrations down to 37 times lower than those afforded by the direct non- *Presented at the 1993 European Winter Conference on Plasma Spectrochemistry Granada Spain January 10- 15 1993.extractive method. In addition the extraction step en- hances the selectivity. The method thus developed was applied to the detemination of nickel in various biological materials prepared by microwave digestion in sealed poly(tetrafluoroethy1ene) containers and two human urines were analysed without mineralization. The advantages of microwave dissolution include the faster reaction rates that result from the high temperatures and pressures attained inside the sealed containers. Another advantage of microwave dissolution is a decrease in blank values as compared with open-beaker work both because contamination from the laboratory environment is much lower and because closed vessels make it possible to use smaller amounts of reagents. Experimental Apparatus The ICP-AES measurements were made on a Perkin-Elmer 4 0 sequential emission spectrometer.The system was controlled by an IBM XT-286 computer which was used to develop the method and in acquiring and storing the data. A standard torch and a Meinhard concentric glass nebulizer (controlled by a peristaltic pump working at a flow rate of 1 ml min-I) were used during the experiments. The r.f. generator of the nebulizer is internally mounted as a 'free- running' oscillator type with a nominal central frequency of 40 MHz and a nominal operating power level of 900 W. The pH measurements were made with the aid of a Crison Digit- 501 pH meter furnished with a combined glass-calomel electrode. Separating funnels were shaken on a Gallenkamp flask agitator. For sample digestion a domestic microwave oven Panasonic Model NN-8507/8557 with a 700 W magnetron and rotating antenna was used.The oven was placed in a laboratory fume-hood. The samples were placed in Parr 478 microwave acid digestion bombs. The bombs were cleaned before use by soaking for one day in 10% nitric acid followed by repeated rinsing with water. All glassware used was soaked in 10% nitric acid for one day and rinsed with water just before use. Reagents All chemicals were of at least analytical-reagent grade and doubly distilled de-ionized water was used throughout.844 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 Table 1 Operating conditions for the ICP Wavelength Background correction R.f. generator frequency Photomultiplier voltage Outer gas flow rate Intermediate gas flow rate Nebulizer (injection) gas flow rate Plasma viewing height Integration time Read delay Peristaltic pump flow rate 352.454 nm -0.047 and 0.028 40 MHz incident power 1.1 kW 600 V 12 1 min-1 0.5 1 min-l 0.4 1 min-l 17 mm above induction coil 300 ms 20 s I ml min-1 A stock solution of Nil1 was prepared from the nitrate and standardized gravimetrically with dimethylglyoxime.Stan- dards of working strength were made by appropriate dilution daily as required. A boric acid-sodium borate buffer of pH 8.8 was prepared by mixing 25 ml of 0.2 mol 1-1 H$O3 and 15 ml of 0.05 mol 1-1 Na2B40,.10HZ0 (0.2 mol 1-1 sodium borate) in a 100 ml calibrated flask and diluting to the mark. A 0.2% m/v BPTH solution in IBMK was prepared by dissolving the appropriate amount of reagent prepared and purified as described under Synthesis of the reagent in hot IBMK and diluting to 100 ml. The solution was stable for at least one week in a refrigerator.A 0.2 mol 1-1 solution of glycine (Sigma) was also used. Synthesis of the reagent A 1 .O g mass of the thiocarbonohydrazide was dissolved in 80- 100 ml of distilled water and 4.3 g of 2-benzoylpyridine in 10 ml of ethanol were added. The mixture was then refluxed for 4 h and the yellow product was filtered washed with hot ethanol and dried in a vacuum desiccator (yield 70% m.p. 165-1 75 "C). Elemental analysis gave C 68.62% H 4.59% N 19.14% and S 7.40%; C2sH20N6S requires C 68.80% H 4.59% N 19.27% and S 7.37%. The structure of the compounds was confirmed by infrared (IR) nuclear magnetic resonance (NMR) and mass spectrometry.Procedures Recommended procedure Aliquots of samples or standard solutions containing nickel were placed in separating funnels. The pH of the aqueous phase was adjusted to 8.8 by means of the borate buffer solution. Then 5 ml of 0.2% BPTH in IBMK were added (the maximum volume ratio of the aqueous to organic phase was 37+ 1 for a single-stage extraction of 99-100%). The mixture was shaken vigorously on the mechanical agitator at 3000 rev min-I for 7 min. The phases were allowed to separate and the organic solvent layer was transferred into a poly(propy1ene) centrifuge tube (some samples could need centifugation for up to 5- 10 rnin to improve separation between the layers). The organic phase was inserted into the plasma bulk by means of a peristaltic pump and nickel was determined according to the instrumental conditions given in Table 1.Three replicate measurements of the two individual sample preparations of the reference material or the three individual preparations of the human urine were made and the nickel concentration was determined from the calibra- tion graph; alternatively the standard additions method could also be applied satisfactorily. Sample preparation Biological tissues. The certified reference materials (CRMs) analysed to determine the accuracy of the proposed procedure were National Research Council Canada (NRCC) CRMs DORM-1 Dogfish Muscle and TORT-1 Lobster Hepatopancreas; and National Institute of Stan- dards and Technology (NIST) Standard Reference Ma- terials (SRMs) 1572 Citrus Leaves.These samples were first dried in accordance with the norms of the respective analysis certificates. Two replicates of each dried sample were prepared according to the following procedure. Samples (200-400 mg) were weighted directly into the digestion vessels on a digital electronic balance and 4 ml of concentrated HNO were added. The mixture was allowed to stand for 30 rnin and then 2 ml of concentrated HCl were added. Next the bomb was closed and placed in the microwave oven together with a beaker filled with 10 ml of water. This procedure minimizes the risk of damage of the magnetron due to the fact that the small sample loads could cause magnetron failure because most of the microwave power is reflected back to the waveguide and magnetron.The programme of the microwave oven was 360 W for 4 min followed by 10 rnin at 180 W. With this programme it is possible to place three bombs at the same time in the microwave oven. The oven was allowed to cool for the same time as the programme duration (14 min). After digestion the solutions were evaporated by heating to a small volume (1-2 ml) to eliminate the HNO and neutralized with NaOH solution then 10 ml of borate buffer of pH 8.8 and 4.0 ml of 0.2 moll-' glycine (to avoid interferences) were added and the solutions were diluted to 100 ml with de-ionized water in a calibrated flask. Finally two identical volumes (50 ml) of each reference material solution were placed in separating funnels and the deter- mination of nickel was completed as described under Recommended procedure using 2 ml of 0.2% BPTH in IBMK as the organic phase.Human urine. The urine was acidified to 1% v/v in concentrated HN03 and stored frozen until required for the determination of nickel. In 100 ml separating funnels 50 ml of acidified urine were taken and neutralized by addition of 3 mol 1-1 NaOH solution. Then 5 ml of the borate buffer of pH 8.8 and 2 ml of 0.2 mol 1-1 glycine were added; 3 ml of 0.2% BPTH solution in IBMK were used as the organic phase. Three identical volumes of each urine were taken for the reaction procedure and the determination of nickel was carried out as described under Recommended procedure using the standard additions method. Results and Discussion Influence of Experimental Variables on the Extraction of Nickel Preliminary experiments showed the extraction of the NP-BPTH complex to be quantitative over the pH range 8.5-12.0 outside which the yield was much lower.All subsequent studies were carried out at pH 8.8 this pH was adjusted using a borate buffer solution. The volume of added buffer had no effect. Increasing the ionic strength produced no significant changes in the extraction. The minimum shaking time was determined by varying the shaking time from 1 to 10 min 5 rnin were sufficient however prolonged shaking had no adverse effect on the extraction; thus a shaking time of 7 rnin was selected for this study. The reagent concentration in the organic phase was varied while keeping the final volume at 5 ml. The results obtained showed that the extracted fraction remained constant for reagent concentrations equal to or greater than 1.15 x 10-3 mol I-' (0.065%).A concentration of 0.2% of845 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 Table 2 Tolerated levels 2.5 ng ml-l of Ni" of foreign species in the determination of observation height were established to achieve the best signal-to-noise (SIN) ratios. The operating conditions for the spectrometer are recorded in Table 1 . Ion or species Tolerated ratio Lil NaI K1 Mg" Ca" Sr" Barl CdIt HgI HgII PbIr Vv AsV krlIr*,'Agr A P bromide chloride flouride iodide nitrate phosphate sulfate oxalate thiosulfate thiourea > 4000 co" 1500 Zn" 1300 cu" 900 *With 2.5 ml of 0.2 mol 1 - I glycine. ?With 2.5 ml of 0.2 mol I-' glycine+ 1 drop of H202.Movf Mnh Bib Sbh A& FelI* Fer1It 2000 Glycine 1000 BPTH was used in practice in order to avoid reagent depletion by other extractable ions potentially occurring in the aqueous medium. Effect of Phase-volume Ratio The volume of the aqueous phase was varied from 10 to 200 ml while keeping that of the organic phase constant at 5 ml (0.2% BPTH); hence the phase volume ratios were varied between 2 and 40. For a ratio >37 the phase separa- tions were unsatisfactory and the procedure was thus inapplicable. On the other hand when the volume of the aqueous phase was increased the volume of the organic phase after extraction decreased because IBMK is not totally immisci- ble with water; however the quantitative extraction of nickel could be verified from the calibration graphs.When a phase ratio of 37 was used (185 ml of aqueous phase and 5 ml of organic phase) only approximately 2 ml of organic phase were obtained after extraction. Under the optimum conditions the recovery factors for the extraction of nickel were calculated by means of a series of experiments in which the atomic emission of this element in the organic phase was compared with that of an appropriate standard. In all instances nickel in the range 0.03- 1 .O pg was extracted completely from the aqueous solution by a single extraction when the volume ratio of the aqueous to organic phase was ~ 3 7 . Selection of Measurement Conditions The following wavelengths were investigated 34 1.476 352.454 239.454 234.554 221.647 and 232.003 nm. The 352.454 nm nickel line was selected as it resulted in the largest differences between the sample and the noise signals [minimum background equivalent concentration (BEC)].Other operating variables such as type of nebulizer nebulizer flow rate potomultiplier voltage and plasma Calibration Graph Precision and Detection Limit The relationship between nickel concentration and emis- sion intensity was studied in the range 0.4- 150 ng ml- of nickel in the aqueous phase. A linear calibration graph with a regression coefficient of 0.999 was obtained in this range when an aqueous-to-organic phase volume ratio of 37 was used. The limits of detection and determination of the method were established according to American Chemical Society Committee of Environmental Improvement defini- t i o n ~ .~ The detection limit thus achieved for an aqueous-to- organic phase volume ratio of 37 was 0.3 ng ml-I and the determination limit was 1.1 ng ml-l (all concentrations refer to the aqueous phase). The precision of the method evaluated as the relative standard deviation obtained after analysing ten series of ten replicates was +2.7% at the 1 . 1 ng ml-I level of nickel when the volume ratio of the aqueous-to-organic phase was 37. Study of Interferences The effect of various ions on the determination of nickel by the proposed method was examined under the optimum working conditions. For this study different amounts of the ionic species to be tested were added to a 2.5 ng ml-l solution of nickel in the aqueous phase the volume of which was 185 ml whereas that of the organic phase was 5 ml.The starting point was an m/m interferent-to-nickel ratio of 4000; if any interference occurred the ratio was gradually lowered until the interference disappeared. The tolerance limits found (Table 2) show that nickel can be determined in the presence of a variety of ions including most of those which commonly occur with nickel in natural and synthetic samples. The tolerance level for some metal ions can be increased by addition of a masking agent (see Table 2). Sample Analysis In order to test the accuracy and applicability of the proposed method to the analysis of real samples some biological reference materials were analysed. Matrix inter- ference was verified by comparison of the slopes of the calibration graphs with those using the standard additions method.Only in the case of human urine were matrix effects apparent for the ICP-AES measurements and quan- tification was performed with the standard additions method. The results are given in Table 3 and 4. Mean values are shown for three replicate measurements of the two individual sample preparations of the reference material and the three preparations for human urine. As can be seen the nickel concentrations determined by the proposed Table 3 Determination of nickel in biological samples Ni concentrationlpg g-I Sample Certified Found* NRCC CRM TORT- 1 Lobster Hepatopancreas 2.3 20.3 2.0kO.l NRCC CRM DORM-1 Dogfish muscle 1.2 -+ 0.3 1.3 k 0.3 NIST SRM 1572 Citrus Leaves 0.6 f 0.3 0.6 -+ 0. I *Mean k standard deviation for three replicate measurements of the two individual sample preparations.846 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL.8 Table 4 Determination of nickel in human urine Ni concentration/ng ml-' Sample Added Found Urine 1 - 2.2 -+ 0.9* Urine 1 6.0 8.6 Urine 1 12.0 13.0 Urine 2 6.0 9.8 Urine 2 12.0 15.7 Urine 2 - 3.6 +- 0.8* *Mean k standard deviation for three replicate measurements of the three individual urine preparations. method are in close agreement with the certified values. For the analysis of human urine the results obtained show a satisfactory recovery of nickel. Conclusion The proposed method is efficient for the accurate determi- nation of nickel in biological materials. The main advan- tages offered by the fast solvent extraction of nickel prior to determination are the suppression of interferences and an increase in the sensitivity by preconcentration of the analyte in addition to fast mineralization of the samples with the microwave digestion in sealed containers. The authors are grateful to the Comision Interminis- terial de Ciencia y Tecnologia (CICYT) for supporting this study (Project PB 90-0809) and also to the Junta de A.ndalucia. I 2 3 4 5 6 7 8 9 References Miyazaki A. Kimura A. Banzho K. and Umezaki Y. Anal. Chim. Acta 1982 114 213. Whiteley R. V. and Merril R. M. Fresenius' 2. Anal. Chem. 1982 311 7. Jones J. S. Harrington D. E. Leone B. A. and Bramstedt W. R. At. Spectrosc. 1983 4(2) 49. Agency for Toxic Substances and Disease Registry Atlanta GA USA 1989 89(11) Abstract No. 930 616. Sunderman F. W. Jr. Arch. Toxicol. Suppl. 13 1989 40. Fisher A. B. Life Chem. Rep. 1989 7 149. Maximilien R. Comm. Eur. Communities [Rep]. Eur 1989 EUR 12456 Pt. 1. International Committee on Nickel Carcinogenesis in Man Scand. J. Work Environ. Health 1990 16 1. ACS Committee of Environmental Improvement Anal. Chem. 1980,52 2242. Paper 3/002 791 Received January 18 1993 Accepted May 5 1993

ISSN:0267-9477    

DOI:10.1039/JA9930800843

出版商:RSC    

年代:1993

数据来源: RSC

摘要:

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 841 Generation of Volatile Cadmium Species With Sodium Tetrahydroborate From Organized Media Application to Cadmium Determination by Inductively Coupled Plasma Atomic Emission Spectrometry* M. C. Valdes-Hevia y Temptano M. R. Fernandez de la Campa and Alfred0 Sanz-Medelt Department of Physical and ‘Analytical Chemistry Faculty of Chemistry University of Oviedo c/Julian Claveria 8 33006-Oviedo Spain Continuous flow generation of volatile cadmium species (assumed to be the hydride) as a means of gaseous sample introduction into an inductively coupled plasma atomic emission spectrometer has been investigated in detail. Different ‘organized molecular assemblies’ (micelles and vesicles) have been tried in order to improve the generation of such cadmium species using NaBH as the reducing agent.Didodecyldimethylammonium bromide (DDAB) vesicles were found to be the most appropriate for the purpose. The advantages of this ‘assumed’ cadmium hydride generation from vesicular media are discussed and critically compared in terms of sensitivity selectivity and accuracy with the determination of low levels of cadmium by inductively coupled plasma atomic emission spectrometry (ICP-AES). The proposed method for the determination of cadmium by ICP-AES using DDAB and NaBH provided 5-fold lower detection limits (1 ng ml-l) as compared with conventional nebulization ICP-AES. A precision of 2% at the 50ng ml-’ level of the metal was achieved and selectivity studies were carried out. The proposed method was applied to the determination of low levels of cadmium in tea infusion. Validation of the method was performed by analysis of the same samples by electrothermal atomic absorption spectrometry with satisfactory results. Keywords Cadmium; inductively coupled plasma atomic emission spectrometry; continuous hydride genera- tion; organized molecular assemblies; didodecyldimethylammonium bromide It is well known that cadmium and its compounds are highly toxic even at low concentration levels.’ They are also considered potentially carcinogenic in humans2 Therefore the determination of cadmium in environmental and food samples is mandatory today and this requires analytical techniques exhibiting low detection limits for this toxic element.The detection power of conventional nebulization induc- tively coupled plasma atomic emission spectrometry (ICP- AES) is not enough to directly analyse and control real-life concentrations of cadmium in environmental and food samples.The use of the hydride generation (HG) technique coupled with atomic absorption spectrometry (AAS) has proved to be a more adequate approach to increase the sensitivity for the determination of ultratrace amounts of those elements that are able to form volatile hydrides from aqueous or organic reaction media.3-8 Formation of other volatile species such as chelates and fluoride~,~-l~ carbo- nylI4-l6 or alkyl d e r i v a t e ~ ~ ~ J ~ of the corresponding elements has also been investigated as a means of increasing sample transport to the atomizer.Although determinations of cadmium using the genera- tion of volatile chlorideI9 or volatile diethylcadmiumZ0 have been reported the generation of volatile CdH2 in aqueous solutions at room temperature for analytical purposes is a rather difficult task owing to the instability of this hydride at temperatures above liquid nitrogen.21 In spite of such instability CdH2 has been proposed as the volatile species of cadmium formed as detected by AAS when treating Cd2+ solutions with NaBH in an organic medium of dimethyl- formamide.22 It has to be stressed however that the hydride nature or otherwise of the detected cadmium volatile species has not yet been demonstrated experimen- *Presented at the 1993 European Winter Conference on Plasma TTo whom correspondence should be addressed.Spectrochemistry Granada Spain January 10- 15 1993. tally. Moreover the generation of CdH2 has not been used for determination of cadmium probably because the insta- bility of CdH2 at room temperature2* would produce unrealiable not very precise analytical results. Surfactant organized media are organized molecular assemblies which have been shown to alter chemical equilibria reaction rates and other important chemical features owing to their capacity to create a special micro- environment for reactions at the molecular These aggregates have proved to be very useful in many fields of analytical c h e m i ~ t r y . ~ ~ - ~ ~ The use of surfactants in atomic spectrometry however has been rather scarce and limited mainly to enhancing the efficiency of nebulization in flame AAS.The benefits of their utilization in plasma emission spectrometry are rather c o n t r o ~ e r s i a l . ~ ~ ~ ~ ~ It has been shown previously that ‘organized media’ can improve the sensi- tivity of HG29 because of the new microenvironment created which is able to improve the kinetics of the generation of volatile species from the desired analyte. This effect has been applied to improve the determination of As and Pb by ICP-AES via arsine and plumbane genera- t i ~ n . ~ ~ ~ ’ An overview of surfactant-based organized media and the interface with atomic spectrometry is soon to be published.32 In this paper the effect of the addition of ‘organized media’ (including different normal micelles and vesicles) to an aqueous NaBH medium to form volatile cadmium species (probably the hydride) and its application to increase the sensitivity of the determination of cadmium by ICP-AES has been thoroughly investigated.Some prelimi- nary experiments carried out by batch AAS are also included as they also showed that organized media particu- larly didodecyldimethylammonium bromide (DDAB) ves- icles clearly enhanced the generation of volatile cadmium species from aqueous solutions at room temperature. As a result a vesicles-sabilized continuous (probably) HG of cadmium is reported here for the first time and a new ICP- AES method for cadmium has been developed based on this system.848 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 ~~ Table 1 Instrumentation for the experimental flow system and ICP-AES set-up Instrument ICP Spectrometer Gas-liquid separation interface Peristaltic pump (four channels) Manufacturer Philips Model PU7000 equipped with 40 MHz source unit Grid type nebulizer and spray chamber provided with the instrument Gilson Minipuls 2 Atomic absorption spectrometer Hydride generator for batch AAS High-intensity ultrasonic processor Perkin-Elmer Model 2280 equipped Perkin-Elmer Model MSH- I0 Sonics and Materials Model SOOW with chart recorder measurements (vesicles generation) Spectrometer 0 11 11 I t - - Peristaltic I I Ar nebulizer Pump Fig.1 Continuous HG flow system The proposed HG-ICP-AES vesicular method has been successfully applied to the determination of low levels of cadmium in beverages. Experimental Instrumentation An inductively coupled plasma atomic emission spectro- meter (Philips Model PU7000) was used for detection by ICP-AES.Other details of the experimental flow system and ICP-AES set-up used are given in Table 1 and Fig. 1. Reagents A 1000 pg ml-1 cadmium(I1) stock standard solution (Merck) was used. Working solutions were freshly prepared daily by diluting appropriate aliquots of the stock solution. Sodium tetrahydroborate(II1) solutions were prepared by dissolving NaBH (Carlo Erba) in ultrapure water (Milli-Q Millipore) stabilized in 0.1090 m/v sodium hydroxide solution. Solutions were prepared weekly and filtered before use. Cetyltrimethylammonium bromide (CTAB) solution ( mol 1-I) was prepared by dissolving the surfactant powder (Fluka) in water by gentle warming.Other surfac- tants such as sodium lauryl sulphate (SLS) solution ( mol l-l) (Sigma) 2% v/v Triton-X 100 (TX- 100) solution (Merck) Zwittergent-3.16 (ZW-3.16) solution ( mol 1-l) (Carbiochem-Behring) were prepared in a similar way. Vesicles of DDAB and the dihexadecyl phosphate (DHDP) mol 1 - I ) were prepared by dissolving the surfactant powders (Kodak and Aldrich respectively) in water and sonicating at room temperature (DDAB) or at 90 "C (DHDP) at a power of 60 W for about 12 min with the tip of a high-intensity ultrasonic processor (Sonics and Materials). (This procedure is referred to here as sonication.) All mineral acids and metal salts used were of analytical- reagent grade and ultrapure water (Milli-Q) was used throughout. General Procedures Continuous volatile cadmium species generation from DDAB In the flow system shown schematically in Fig.1 the cadmium sample dissolved in 0.4 mol 1-l HC1 and mol 1-1 DDAB vesicles was continuously pumped through one of the channels of the peristaltic pump at a rate of 0.75 ml min-' and merged with a 4% m/v solution of NaBH (flow rate 0.75 ml min-l) dissolved in mol 1-1 DDAB vesicles. This solution feeds the grid nebulizer of the ICP detuned in order to allow for separation of the volatile species which finally reach the torch of the plasma while the liquid phase goes to drain. Cadmium is measured at the 214.440 nm emission line under the conditions given in Table 2. Background correction using two points at 2 14.408 and 2 14.468 nm was carried out. Ultrajltration experiments Ultrafiltration experiments on vesicular solutions were performed in 50 ml stirred cells (Amicom).The pressure was held constant at 101.325 kPa and the cells were initially filled with 10 ml of the test solution. Regenerated cellulose ultrafiltration membranes having a nominal maximum relative molecular mass of 5000 u (Amicom) were used. They were washed thoroughly with ultrapure water (Milli- Q) for approximately 30 min before use. Tea samples Samples of tea infusion were analysed following the continuous flow volatile cadmium generation and general ICP-AES procedures described above. Results and Discussion Optimization of Instrumental and Chemical Parameters Using the procedure outlined above for continuous gas- liquid separation and ICP detection the effect on cadmiumJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 849 Table 2 Optimum conditions for cadmium hydride generation by ICP-AES Optimum plasma experimental conditions Wavelengthlnm 2 14.440 R.f.forward power/kW 0.9 Nebulizer gas pressure/psi* 40 Coolant gas flow rate/l min-' 13 Intermediate gas flow rate/l min-' 0 Final sample introduction Integration time/s 3 flow rate/ml min-* 1.5 Optimum chemical parameters HCI NaBH Surfactant * 1 psi=6893 Pa. 0.4 mol I-' (flow rate 0.75 ml min-l) 4% m/v in 0.1 To m/v NaOH (flow rate 0.75 ml min-I) DDAB mol I-' [HCll/mol I-' 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 8 6 2 ~~~ ~ 0 1 2 3 4 5 [NaBH,I (% m/v) Fig. 2 Concentration effect of HCI and NaBH on the cadmium signal obtained by HG-ICP-AES from DDAB signals of plasma instrumental variables such as nebulizer gas pressure forward radiofrequency (r.f.) power and coolant gas flow rate were studied as were the volatile species chemical generation variables concentrations of reactants sample flow rates etc.by following a univariant- type experimental search. The optimum ICP instrumental values obtained are summarized in Table 2. Those instru- mental settings were selected to study optimum chemical parameters (concentration of reactants flows etc.) for continuous generation and transport of CdHz to the plasma. The maximum signal-to-background ratio (S/B) at the 2 14.440 nm line of cadmium was always the optimization criterion. The results observed are plotted in Fig. 2. As can be seen the effect of the concentration of HCl is rather critical (final pH or generating pH = 1.2).Fig. 2 also shows that the optimum concentration of NaBH was not so critical and a concentration of 4% m/v was eventually selected. The effect of the concentration of DDAB on cadmium signals is shown in Fig. 3. Addition of DDAB vesicles to sample solutions produced an important in- crease in the cadmium signal and maximum enhancement was observed after the concentration of DDAB monomers reached a value of lo-* moll-' or higher (see the plateau in Fig. 3). The influence of sample flow rate was also studied from 0.25 to 5 ml min-I. As expected the signal increased with the sample flow rate up t o a value of 1 ml min-' then a plateau was reached for stationary HG/excitation condi- tions. A final flow rate of 1.5 ml min-I was therefore used for subsequent experiments.Both optimum values selected for ICP plasma conditions and for continuous chemical generation of CdH2 are summarized in Table 2. Analytical Performance Characteristics From Different Reaction Media In the light of the special aqueous-organic microenviron- ment provided by surfactant-based organized assemblies the formation of the volatile cadmium species could be influenced by the nature of the surfactant used. Therefore the effect of the addition of micelles and vesicles on the HG-ICP-AES signals of cadmium at 214.440 nm was investigated using the continuous HG system and tech- niques previously described. Several types of organized assemblies were examined including cationic (CTAB) anionic (SLS) zwitter-ionic (ZW-3.16) and non-ionic (Triton X- 100) micelles and also anionic (DHDP) and cationic (DDAB) vesicles.Using the experimental conditions given in Table 2 the analytical parameters of the corresponding HG-ICP-AES method for the different reaction media were evaluated. The observed influence of different organized media on slope detection limits and precision of the determination of cadmium by HG-ICP-AES are summarized in Table 3. The results can be compared with those obtained by conventional nebuliza- tion. As can be seen the best analytical performance was 1 I I I I I I 0 50 100 150 200 250 300 [DDABI/lO-'rnol I" Fig. 3 Concentration effect of DDAB on the cadmium signal850 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL.8 Table 3 Comparison of different media for the generation of volatile cadmium species with detection by ICP-AES Surfactant Detection Slope Precision limit* ( X 105) (Oh)t Conventional nebulization - 5 2.66 1.3 Hydride generation Water CTAB TX- 100 ZW-3.16 SLS DHDP DDAB - - - - Cationic micelles 7 3.83 2.7 Non-ionic micelles 3 6.8 1 2.8 Zwiterionic micelles 2 8.59 2.6 Anionic micelles 2 9.18 2.3 Anionic vesicles 3 7.69 2.2 Cationic vesicles 1 15.50 2.1 *Detection limit (ng ml-’)= 3a&lope. (IUPAC definition a,= the absolute standard deviation observed for ten blanks measured ?Precision (RSD %)=relative standard deviation for samples at the 50 ng ml-I of Cd level (n= 10). independently). Table 4 Ultrafiltration studies of cadmium in vesicles System Ultrafiltrated cadmium (O/o)* Relative emission intensity (ICP) Cadmium in aqueous solution 100.0 2 2.0 0 DDAB sonicated + cadmium Cadmium + DDAB without ultrasonication 93.0 3.2.0 300 simultaneously sonicated 46.7 3 1.3 700 *Studies using a 30 ng ml-I Cd solution as reference; n=7.observed using vesicles of DDAB the calibration graph was linear up to 1 pg ml-1 and the detection limit (30,) was 1 ng ml-I of cadmium (5-fold better than when using conven- tional nebulization). The precision evaluated as the repea- tibility obtained by analysing ten independent replicates of a solution containing 50 ng ml-l of cadmium was 2.1%. Therefore cationic vesicles of DDAB were selected for further work and the determination of low levels of cadmium by HG-ICP-AES enhanced by DDAB was investi- gated in more detail.Mechanism of Reaction Ultrafiltration Studies As the results showed that the most promising analytical potential was provided by vesicles of DDAB the impor- tance of the formation of vesicles for enhanced ICP-AES cadmium signals was tested using sonication experiments (see under Reagents). (Sonication is mandatory to obtain a solution of individual vesicles.) Micellar ultrafiltration has been used in some cases for the removal of alcohoW3 and cres01s~~ from aqueous streams by holding such organics in the unfiltered micelles. In order to verify if Cd2+ was present in the DDAB vesicles ultrafiltration experiments were performed in which solutions of Cd2+ were added to already sonicated DDAB or alternatively NaBH and Cd2+ solutions were added and sonicated at the same time as the sufactant solution (in order to form the vesicles).The resulting vesicular solutions were ultrafiltrated and the Cd2+ content in the ultrafiltrate was determined by elec- trothermal AAS. The results observed are shown in Table 4. These results demonstrate that when sonication of the DDAB solution to obtain single vesicles is carried out with the Cd2+ present it seems that around 50% of the total Cd2+ added remains in the vesicles (which do not pass the ultrafiltration membrane) and maximum HG-ICP-AES signals for cadmium are then obtained. If Cd2+ is added to already formed DDAB vesicles virtually all added Cd2+ seems to remain outside the vesicles in the bulk aqueous solution because it passes quantitativly into the ultrafiltra- ble fraction as shown by the results in Table 4.Then the analytical signal is clearly reduced. The beneficial effect of I V Y 1 A 600 4-4 .- C 3 2 500 $ 400 E 4-4 .- - .- 2 g 200 300 w .- .- v) .- E w 100 21 4.44 214.47 Wavelengthlnm Fig. 4 Emission spectra of 500 ng ml-* of cadmium A generation of CdHl from DDAB simultaneously sonicated; B generation of CdHz from DDAB vesicles without simultaneous sonicaton; C conventional nebulization; and D generation of CdHz from water simultaneous sonication of the sample NaBH and DDAB solutions for obtaining maximum analytical signals is also shown graphically in Fig. 4. Identification of Volatile Species On the other hand our efforts to characterize the nature of the observed volatile species of cadmium measured in the ICP (formed when the analyte and NaBH vesicular solutions merge in the continuous HG system of Fig.1) have failed so far. Attempts to ‘trap’ the possible ‘hydride’ (by passing the HG resulting gases through a liquid N2 cryogenic trap) and its eventual release by warming in the injection port of a gas chromatography-mass spectrometry instrument (for elemental identification) has been unsuc- cessful so far; probably decomposition of the hydride and deposition of CdO on the walls of containers prevents identification of the hydride.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL. 8 85 1 0.7 0.6 Q 0.5 0 co f! 2 0.4 n E .$ 0.3 c Y co h 0.2 0.1 0 125 10 20 30 40 50 60- Tern perature/"C Fig. 5 Temperature effect on cadmium signal obtained by batch HG-AAS from DDAB In any case the 'vesicular HG' technique proposed offers five times better detection limits for cadmium than conven- tional nebulization ICP-AES.Effect of Reaction Vessel Temperature on the Sensitivity of Cadmium Determination Considering the thermal instability of the assumed CdH2 formed the effect of the reaction vessel temperature was studied between 0 and 60 "C. It was found that temperature greatly affects the kinetics and efficiency of the reaction for the generation of the volatile species. For these experi- ments the generation was carried out using a cadmium concentration of 100 ppb in an adequately thermostated batch system (Perkin-Elmer MS- 10 hydride generator). The results obtained are plotted in Fig. 5 for peak height and peak area of the observed transient signals.They show that at temperatures around 0 "C the HG reaction is very rapid and of course the efficiency of CdH2 generation and transport as measured by the strength of the HG-AAS signal of cadmium was about four times better than that observed at room temperature. Therefore operating at lower temperatures for HG can provide a further decrease in detection limit for cadmium. Interference Studies Using the selected conditions for cadmium (Table 2) the effect of the presence of foreign elements on the HG-ICP- AES signal was investigated. All the potentially interfering elements tested and the level of tolerance observed in the determinations of cadmium are summarized in Table 5. Hydride-forming elements and high levels of alkali alkaline earth metals or common anions were found not to affect cadmium HG in the DDAB medium. The only interfences found were those of Zn and Ni and they occurred only when those elements were present at relatively high excesses.Analysis of Real Samples Having established the best HG conditions for the deter- minations of cadmium by HG-ICP-AES the method recommended under General Procedures was applied to the determination of low levels of cadmium in infusions of different types of teas. The cadmium added to the samples of tea was determined directly by HG-ICP-AES after extraction with hot water for 30 min for the preparation of Table 5 Interference studies on 100 ng ml-' of Cd; DDAB generation medium Interferent As1" Zn" CU" Mn" Nil1 Fe"' Hg" Pb" Mg" Ca" Na' K' Nitrate Chloride Cd:interferent mass ratio 1:lO 1:lOO 1500 1:lO 1:lOO 1500 1:lO 1:lOO 1:500 1:lO 1:lOO 1300 1:lO 1:lOO 1:500 1:lO 1:100 1500 1:lO 1:lOO 1500 1:lO 1:lOO 1:500 1:lO 1:lOO 1:500 1 10 1:lOO 1 500 1 500 1:lOOO 1300 1 1000 1 1000 1 10000 1:lOOO 1 10000 Interferent amount/ pg ml-' 1 10 50 1 10 50 1 10 50 1 10 50 1 10 50 1 10 50 1 10 50 1 10 50 1 10 50 1 10 50 50 100 50 100 100 1000 100 1000 Recovery (%) 100 98 97 99 97 75 100 101 98 100 100 101 100 93 61 98 99 98 98 100 101 99 100 99 100 100 98 97 101 100 99 99 98 100 100 101 99 98 Table 6 Analysis of cadmium in real samples with HG-ICP-AES from DDAB.Results given as the mean of three independent measurements 2 standard deviation. The analyses were performed at the same time with recalibration when necessary Cadmium concentration/ng rnl- Sample HG-ICP-AES ETAAS Ordinary tea 1 5.8 k 1.8 5.1 20.3 22.8 2 0.9 Ordinary tea 2 21.82 1.0 Jasmine tea 6.3 k 1.3 6.1 k0.2 Grey tea 1 17.9k 1.1 17.1 k0.8 Grey tea 2 11.6k0.9 10.7 +- 0.2 Orange tea 8.4k 1.2 7.9 -+ 0.4 the infusions.Background correction (at 2 1 4.408 and 2 14.468 nm) was employed and the rest of the conditions used were as given in Table 2. The results obtained by the proposed method were compared with the results obtained by electrothermal AAS (ETAAS) for the same samples and the observed compara- tive values are summarized in Table 6. As can be seen very good agreement in the values obtained for the same tea infusion samples analysed by both techniques was observed. Therefore the validity of the852 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY SEPTEMBER 1993 VOL.8 proposed organized medium-enhanced HG-ICP-AES method for the determination of low levels of cadmium in tea infusions has been demonstrated. Conclusions It has been demonstrated that the generaton of volatile cadmium (probably CdH2) at room temperature is pos- sible from a vesicular medium of DDAB using NaBH as the reducing agent. This generation from DDAB can be used to achieve sensitive and selective determination of low levels of cadmium and this method has proved to be very adequate for the determination of the metal in environmen- tal and food samples. The sensitivity of the proposed vesicular method (slope of the calibration graph) is about six times higher than that obtained with conventional nebulization ICP-AES and a five times lower detection limit (LOD=l ng ml-l) can be achieved as compared to the LOD of conventional nebulization ICP-AES (5 ng ml-* in the instrument used here). The nature of the volatile cadmium species is thought to be CdH2 (considering the analogy with other hydride fofming elements) but it has to be positively identified by alternative techniques in further work.In any case the ‘vesicular hydride generation’ technique has been applied to establish a new method for determination of the metal in real samples. In brief the potential ability of vesicles to organize reactants at a molecular level improving the kinetics of reactions has been realized here in the problematic generation of cadmium hydride at room temperature allowing for the first reported analytical application of (probably) CdHz generation to improve the sensitivity of the determination of cadmium by ICP-AES.We acknowledge the financial support from FICYT (Funda- cibn para el Foment0 en Asturias de la Investigacion Cientifica Aplicada y la Tecnologia) and CICYT (Comision Interministerial de Ciencia y Technologia) and also the FICYT grant to M C. V.-H. y T. The loan of the ICP PU7000 instrument by Unicam Analytical Systems Cam- bridge UK is also gratefully acknowledged. References 1 Cadmium and Health A Toxicological and Epidemiological Appraisal Volume 1 Exposure Dose and Metabolism eds. Friberg L. Elinder C-G. Kjellstrom T. and Nordberg C. F. CRC Press Boca Raton FL 1985. 2 International Agency of Research on Cancer ZARC Mono- graphs on the Evaluation of the Carcinogenic Risk of Chemicals on Humans Vol.1-20 (suppl. l) IARC Lyon 1979 p. 27. 3 FernAndez F. J. and Manning D. C. At. Absorpt. Newsl. 1971 10 86. 4 Dalton E. F. and Malanski A. J. At. Absorpt. Newls. 1971 10 92. 5 Godden R. G. and Thomerson D. R. Analyst 1980 105 1137. 6 Nakahara T. Prog. Anal. At. Spectrosc. 1983 6 163. 7 Busheina I. S. and Headridge J. B. Talanta 1982 29 5 19. 8 Yan D. Yan Z. Cheng G. and Li A. Talanta 1984 31 133. 9 Hilderbrand D. C. and Picket E. E. Anal. Chem. 1975 47 424. 10 Rigin V. I. Zh. Anal. Khim. 1985 XL 630. I 1 Rigin V. I. Zh. Anal. Khim. 1985 XL 1399. 12 Castillo J. R. Mir J. M. Garcia Ruiz M. E. and Bendicho C. Fresenius’ J. Anal. Chem. 1990 338 721. I3 Rigin V. I. Zh. Anal. Khim. 1986 XLI 788.14 Lee D. S. Anal. Chem. 1982 54 1182. 15 Vijan P. N. At. Spectrosc. 1980 1 143. 116 Alary J. Vandale J. Escrient C. and Haran R. Talanta 1986 33 748. 17 Nerin C. Olavide S. and Cacho J. Anal. Chem. 1987 59 1918. 18 Ebdon L. Hill S. and Ward R. W. Analyst 1986,111 1 1 13. 19 Skogerboe R. K. Dick D. L. Pavlica D. A. and Lichte F. E. Anal. Chem. 1975 47 568. 20 D’Ulivo A. and Chen Y. J. Anal. At. Spectrom. 1989,4 3 19. :2I Barbaras G. D. Dillard C. Finholt A. E. Wartik T. Wilzbach K. E. and Schlesinger H. I. J. Am. Chem. Soc. 1951,73,4585. 22 Cacho J. Beltran I. and Nerin C. J. Anal. At. Spectrom. 1989 4 661. 23 Pelizzetti E. and Pramauro E. Anal. Chim. Acta 1985 169 1. 24 Ramis Ramos G. Khashwneh I. M. Garcia de Coque M. C. and Winefordner J. D. Talanta 1988 35 41. 25 Cline Love L. J. Habarta J. G. and Dorsey J. G. Anal. Chem. 1984 56 1 133A. 26 Sanz-Medel A. Martinez Garcia P. L. and Diaz Garcia M. E. Anal. Chem. 1987 59 774. 27 Yin Yan Z. and Zhang W. J. Anal. At. Spectrom. 1989 4 797. 28 Mora J. Canals A. and Hernandis V. J. Anal. At. Spectrom. 1991 6 139. 29 Fernandez de la Campa M. R. Garcia Ortiz C. Diaz Garcia M. E. and Sanz-Medel A. paper presented at the XI1 National Meeting on Spectroscopy and V Instrumental Analysis Meet- ing (JAI) Barcelona Spain November 1990 (abstract AA- 10). 30 Aizpun Fernandez B. ValdCs-Hevia y Temprano M. C. Fernandez de la Campa M. R. Sanz-Medel A. and Neil P. Talanta 1992 39 1 5 17. 31 Valdes-Hevia y Temprano M. C. Aizpun Fernhndez B. Fernandez de la Campa M. R. and Sanz-Medel A. Anal. Chim. Acta in the press. 32 Sanz-Medel A. FernAndez de la Campa M. R. Valdes-Hevia y Temprano M. C. Aizpun Fernandez B. and Liu Y-M. Talanta in the press. 33 Lane Gibbs L. Scamehorn J. F. and Christian S. D. J. Membr. Sci. 1987 196 103. 34 Bhat S. N. Smith G. A. Tucker E. E. Christian S. D. and Scamehorn J. F. Znd. Eng. Chem. Res. 1987 26 1217. Paper 3/01 005H Received February 19 1993 Accepted May 6 1993

ISSN:0267-9477    

DOI:10.1039/JA9930800847

出版商:RSC    

年代:1993

数据来源: RSC