- Email: [email protected]
Synchrotron radiation total reflection for rainwater analysis
Spectrochimica Acta Part B 55 Ž2000. 1173᎐1179
Synchrotron radiation total reflection for rainwater analysis 夽 Silvana Moreira SimabucoU , Edson Matsumoto Campinas State Uni¨ ersity, Ci¨ il Engineering Faculty, Water Resources Department, P.O. Box 6021, 13083-970 Campinas-SP, Brazil Received 1 November 1999; accepted 25 January 2000
Abstract Total reflection X-ray fluorescence analysis excited with synchrotron radiation ŽSR-TXRF. has been used for rainwater trace element analysis. The samples were collected from four different sites at Campinas City, SP, Brazil. Standard solutions with gallium as internal standard were prepared for the calibration system. Rainwater samples of 10 l were added to Perspex reflector disks, dried under vacuum and analyzed for 100 s measuring time. The detection limits obtained for K-shell lines varied from 29 ng mly1 for sulfur to 1.3 ng mly1 for zinc and copper, while for L-shell the values were 4.5 ng mly1 for mercury and 7.0 ng mly1 for lead. 䊚 2000 Elsevier Science B.V. All rights reserved. Keywords: Synchrotron radiation; Rainwater; Total reflection; Trace element analysis
1. Introduction An improvement of lower detection limits ŽLD. in X-ray fluorescence analysis can be achieved in different manners. Straightforward ways to
夽
This paper was presented at Colloquium Spectroscopicum Internationale XXXI, held in Ankara, Turkey, September 1999, and is published in the Special Issue of Spectrochimica Acta Part B, dedicated to that conference. U Corresponding author. E-mail address: [email protected] ŽS. Moreira Simabuco..
improve the LD are obviously to increase the counting time and sensitivity and to reduce the background intensity. The reduction of the background intensity is achieved by applying physical phenomena such as total reflection or linear polarization of the exciting radiation or a combination of both. The most intensive X-ray source available nowadays is the synchrotron, providing outstanding properties of brilliance, linear polarization and natural collimation. TXRF is especially suitable for ultra-trace analysis of pure waters such as rain and drinking water and its advantages when compared with
0584-8547r00r$ - see front matter 䊚 2000 Elsevier Science B.V. All rights reserved. PII: S 0 5 8 4 - 8 5 4 7 Ž 0 0 . 0 0 1 5 8 - 0
1174
S. Moreira Simabuco, E. Matsumoto r Spectrochimica Acta Part B: Atomic Spectroscopy 55 (2000) 1173᎐1179
other methods such as atomic absorption spectrometry ŽAAS. and inductively coupled plasma techniques ŽICP-AES and ICP-MS. are the multielement determination and low cost w1x. The aim of this work is to apply the total reflection with synchrotron radiation ŽSR-XRF. to determine the elemental concentrations in rainwater sampling in four different sites in Campinas City, SP, Brazil.
In other words, from Eq. Ž1. we can write: Ii s Ci s i IGa sGa CGa Ii s C s i C IGa Ga sGa i
Ž2.
and having: Ri s
2. Materials and methods
Ii C IGa Ga
Ž3.
si sGa
Ž4.
and The total reflection X-ray fluorescence technique was introduced in 1971 by Yoneda and Horiuchi w2x and developed by Aiginger and Wobrauschek w3,4x. This method is based on the incidence of an X-ray beam at small angle Ždenoted critical angle. on the flat surface of a support or carrier Žfor example, quartz or Perspex. on which the sample to be analyzed is deposited. In this condition the scattering effect is minimized and thus a better peak-to-background ratio is obtained reducing in this way the detection limits. Another advantage of this technique is the small volumes required for liquid sample analysis Žmicroliters. or small masses Žmicrograms. for solid samples after chemical digestion. The quantitative analysis can be made through Eq. Ž1., because the sample can be considered as a thin film so that absorption and enhancement effects can be neglected. Ii s s i C i
Ž1.
where Ii represents the intensity Žcps. for K or L X-ray line for the element i; Ci the concentration Žin ppm or g mly1 . and Si the sensitivity for this element Žcps gy1 ml.. The thin film formed on the Perspex support does not have a regular geometry and therefore the X-ray intensities depend on the thin film position. This geometric effect w1,5x can be corrected by computing the relative intensity Ž R i . wEq. Ž3.x for each element in relation to an internal standard added in every sample and standard.
Si s
the result is: R i s Si ⭈ Ci
Ž5.
where Ii and Ci are the intensity and concentration of the i th-element in the sample; IGa and CGa the intensity and concentration of the internal standard Ga in the sample; si and sGa are the sensitivity for element i and for the internal standard Ga; R i the relative intensity for the i th-element and Si its relative sensitivity Žnon-dimensional.. In a relative intensity Ž R i . vs. concentration Ž Ci . plot, the slope of the calibration line represents the non-dimensional sensitivity Ž Si . element i. Table 1 Site and date sampling for rainwater samples Site
Sample
Date
Water treatment plant 1 and 2
Ch1-1 Ch1-2 Ch1-3 Ch1-4 Ch1-5 Ch1-6 Ch2-1 Ch2-2 Ch3-1 Ch3-2 Ch4-1
3r29r98 4r29r98 5r17r98 6r19r98 7r20r98 8r03r98 3r29r98 8r03r98 3r29r98 8r03r98 3r29r98
Water treatment plant 3 and 4 Water treatment plant Capivari River Water supply Atibaia River
S. Moreira Simabuco, E. Matsumoto r Spectrochimica Acta Part B: Atomic Spectroscopy 55 (2000) 1173᎐1179
1175
Fig. 2. Non-dimensional sensitivity curve for synchrotron radiation X-ray total reflection.
Fig. 1. Linear regressions for calculating the sensitivity for the elements Ni, Co, Mn, V and Pb.
2.1. Instrumentation For the X-ray detection a hyperpure Ge detector was employed, with 140 eV resolution at 5.9 keV ŽMn K ␣ line., while for the excitation a white beam of synchrotron radiation w6x with 2 mm width and 1 mm height under total reflection conditions was used.
ple. Then an aliquot of 10 l was pipetted onto a Perspex disk and dried in vacuum thus giving rise to a shaping thin layer of approximately 5 mm diameter Žtwo replicates.. Polished quartz is normally used as sample carrier and it can be cleaned and reused while Perspex is disposable and cheap. The samples were excited for 100 s and X-ray spectra obtained were evaluated by the software QXAS w7x in order to obtain the X-ray intensities. 2.3. Sampling Data The site and date sampling for rainwater samples are shown in Table 1.
2.2. Sample and standard preparations To set up the calibration curve, five standard solutions containing the elements V, Cr, Mn, Fe, Co, Ni, Cu, Zn, As, Se, Tl and Pb were prepared at different and well-known concentrations Žfrom 0.3 to 3.8 g ly1 . with Ga addition as the internal standard to eliminate errors caused by excitationrdetection geometry. For the sample preparation, 1 ml of each sample was taken and 10 l of Ga Ž1025 g mly1 . was added resulting in a Ga concentration of 10.148 g mly1 as internal standard in the sam-
Fig. 3. Detection limit for synchrotron radiation total reflection ŽSR-TXRF. for 1000 s.
1176
S. Moreira Simabuco, E. Matsumoto r Spectrochimica Acta Part B: Atomic Spectroscopy 55 (2000) 1173᎐1179
Table 2 Measured and certified values in the standard reference material ŽDrinking Water Pollutants . Element
Cr As Se Ba Pb
Concentration Žg mly1 . Certified value
Measured value
10.00" 0.49 10.00" 0.49 5.00" 0.05 100.00" 0.53 10.00" 0.49
9.74" 0.09 10.05" 0.14 5.00" 0.25 99.98" 0.53 9.86" 0.61
3. Results and discussion With the X-ray intensities obtained from standard samples, relative intensities Ž R i . were calculated by using Eq. Ž3. and, e.g. the relationship
for the elements Ni, Co, Mn, V and Pb can be seen in Fig. 1. On the basis of the measurements performed on standard solutions, the calibration curve for synchrotron radiation total reflection as shown in Fig. 2 was obtained. The detection limits were calculated using w8,9x below: LMDi s 3 ⭈
(
Ii Ž BG . C ⭈ Ga t IGa ⭈ Si
Ž6.
where Ii Ž BG. is the background intensity for the element i; IGa the internal standard intensity ŽGa.; CGa the internal standard concentration ŽGa., Si the non-dimensional sensitivity for the element i and t the measuring time. Detection limits Žin g ly1 or ng mly1 . for synchrotron
Fig. 4. Characteristic X-ray spectrum of rainwater ŽCh1-4. measured for 100 s with SR-TXRF.
S. Moreira Simabuco, E. Matsumoto r Spectrochimica Acta Part B: Atomic Spectroscopy 55 (2000) 1173᎐1179
1177
Fig. 5. Concentration Žng mly1 . of S, Ca, Fe, Zn, Cl and Ti for rainwater samples.
radiation total reflection, calculated using Eq. Ž6. above, were extrapolated for 1000 s measuring time ŽFig. 3.. As can be visualized in this figure, the detection limits obtained for the K-shell changed from 29 ng mly1 for sulfur to 1.3 ng mly1 for zinc and copper. For the L-shell, the detection limits were 4.5 ng mly1 for mercury and 7.0 ng mly1 for lead. In general, the detection limits values obtained
are in good agreement with those reported by other workers in different synchrotron radiation facilities w1,5,10,11x. A standard reference material ŽDrinking Water Pollutants, Aldrich, 41,393-3. containing Cr, As, Se, Ba and Pb was analyzed in order to test the procedure. Table 2 shows the results and it can be see that the measured values agree well with the certified values. As can be seen in Table 1, the
1178
S. Moreira Simabuco, E. Matsumoto r Spectrochimica Acta Part B: Atomic Spectroscopy 55 (2000) 1173᎐1179
Fig. 6. Concentration Žng mly1 . of K, Ni, Mn, Cu, Cr and V for rainwater samples.
accuracy of this method is approximately 3% and the precision approximately 2%. Fig. 4 presents the spectrum for rainwater sample ŽCh1-4. measured by SR-TXRF for 100 s. The concentrations of rainwater samples were calculated using the equation presented in Fig. 2 and the results are presented in Figs. 5 and 6.
4. Conclusions The elements Ca, S, Fe, Zn and Hg presented the highest concentrations in every sample. Com-
paring the Ch1-1, Ch2-1, Ch3-1 and Ch4-1 samples collected in the same day Ž29 March 1998. and different sites, a significant variation was not observed for the same element but for the other three samples, Ch1-6, Ch2-2 and Ch3-2, sampled on 3 August 1998, a small variation can be observed in the concentration for the same element. When the Ch2-1 and Ch2-2 samples are compared, which were sampled at the same site but on different dates Ž29 March 1998 and 3 August 1998., it can be noted that the values for the Ch2-2 sample are higher than for the Ch2-1 sam-
S. Moreira Simabuco, E. Matsumoto r Spectrochimica Acta Part B: Atomic Spectroscopy 55 (2000) 1173᎐1179
ple. The same fact happened for the Ch3-1 and Ch3-2 samples, collected in the same date cited above. This can be explained by the fact that August was a rainier month than March, thus, the particulate material present in the atmosphere is higher. When it rains this material is dragged together with rainwater, resulting in higher concentrations in the sample.
Acknowledgements Research was partially performed at the National Synchrotron Light Laboratory ŽLNLS. and financial support came from FAPESP. References w1x R. Klockemkamper, A.Von Bohlen, Elemental analysis ¨ of environmental samples by total reflection fluorescence: a review, X-ray Spectrom. 25 Ž1996. 156᎐162. w2x Y. Yoneda, T. Horiuchi, Optical flats for use in X-ray spectrochemical micro-analysis, Rev. Sci. Instrum. 42 Ž1971. 1069. w3x H. Aiginger, P. Wobrauschek, A method for quantitative X-ray fluorescence analysis in the nanogram region, Nucl. Instrum. Methods 114 Ž1974. 157᎐158.
1179
w4x P. Wobrauschek, H. Aiginger, Total reflection fluorescence spectrometric determination of elements in ng amounts, Anal. Chem. 47 Ž1975. 852᎐855. w5x W. Ladisich, R. Rieder, P. Wobrauschek, Quantitative total reflection X-ray fluorescence analysis with monoenergetic excitation, Wiley, Chichester, 1994. w6x C.A. Perez, M. Radtke, H.J. Sanchez, H. Tolentino, R. ´ ´ Neuenshwander, W. Barg, M. Rubio, M.I.S. Bueno, I.M. Raimundo, J.R. Rohwedder, Synchrotron radiation X-ray fluorescence at the LNLS: beamline instrumentation and experiments, X-ray Spectrom. 28 Ž1999. 320᎐326. w7x Quantitative X-ray analysis system ŽQXAS. software package, IAEA, Vienna. w8x L.A. Curie, Limits for quantitative detection and quantitative determination, Anal. Chem. 40 Ž3. Ž1968. 586᎐593. w9x W. Ladisich, R. Rieder, P. Wobrauschek, H. Aiginger, Total reflection X-ray fluorescence analysis with monoenergetic excitation using rotating anode X-ray tubes, Nucl. Instrum. Methods Phys. Res. 330A Ž1993. 501᎐506. w10x L.M. Muia, F. Razafindramisa, R. Van Grieken, Total reflection X-ray fluorescence analysis using an extended focus tube for the determination of dissolved elements in rain water, Spectrochim. Acta Part B 46 Ž1991. 1421᎐1427. w11x P. Hoffmann, V.K. Karandashev, T. Sinner, H.M. Ortner, Chemical analysis of rain and snow samples from ChernogolovkarRussia by IC, TXTF and ICP-MS, Fresenius J. Anal. Chem. 357 Ž1997. 1142᎐1148.
Synchrotron radiation total reflection for rainwater analysis 夽 Silvana Moreira SimabucoU , Edson Matsumoto Campinas State Uni¨ ersity, Ci¨ il Engineering Faculty, Water Resources Department, P.O. Box 6021, 13083-970 Campinas-SP, Brazil Received 1 November 1999; accepted 25 January 2000
Abstract Total reflection X-ray fluorescence analysis excited with synchrotron radiation ŽSR-TXRF. has been used for rainwater trace element analysis. The samples were collected from four different sites at Campinas City, SP, Brazil. Standard solutions with gallium as internal standard were prepared for the calibration system. Rainwater samples of 10 l were added to Perspex reflector disks, dried under vacuum and analyzed for 100 s measuring time. The detection limits obtained for K-shell lines varied from 29 ng mly1 for sulfur to 1.3 ng mly1 for zinc and copper, while for L-shell the values were 4.5 ng mly1 for mercury and 7.0 ng mly1 for lead. 䊚 2000 Elsevier Science B.V. All rights reserved. Keywords: Synchrotron radiation; Rainwater; Total reflection; Trace element analysis
1. Introduction An improvement of lower detection limits ŽLD. in X-ray fluorescence analysis can be achieved in different manners. Straightforward ways to
夽
This paper was presented at Colloquium Spectroscopicum Internationale XXXI, held in Ankara, Turkey, September 1999, and is published in the Special Issue of Spectrochimica Acta Part B, dedicated to that conference. U Corresponding author. E-mail address: [email protected] ŽS. Moreira Simabuco..
improve the LD are obviously to increase the counting time and sensitivity and to reduce the background intensity. The reduction of the background intensity is achieved by applying physical phenomena such as total reflection or linear polarization of the exciting radiation or a combination of both. The most intensive X-ray source available nowadays is the synchrotron, providing outstanding properties of brilliance, linear polarization and natural collimation. TXRF is especially suitable for ultra-trace analysis of pure waters such as rain and drinking water and its advantages when compared with
0584-8547r00r$ - see front matter 䊚 2000 Elsevier Science B.V. All rights reserved. PII: S 0 5 8 4 - 8 5 4 7 Ž 0 0 . 0 0 1 5 8 - 0
1174
S. Moreira Simabuco, E. Matsumoto r Spectrochimica Acta Part B: Atomic Spectroscopy 55 (2000) 1173᎐1179
other methods such as atomic absorption spectrometry ŽAAS. and inductively coupled plasma techniques ŽICP-AES and ICP-MS. are the multielement determination and low cost w1x. The aim of this work is to apply the total reflection with synchrotron radiation ŽSR-XRF. to determine the elemental concentrations in rainwater sampling in four different sites in Campinas City, SP, Brazil.
In other words, from Eq. Ž1. we can write: Ii s Ci s i IGa sGa CGa Ii s C s i C IGa Ga sGa i
Ž2.
and having: Ri s
2. Materials and methods
Ii C IGa Ga
Ž3.
si sGa
Ž4.
and The total reflection X-ray fluorescence technique was introduced in 1971 by Yoneda and Horiuchi w2x and developed by Aiginger and Wobrauschek w3,4x. This method is based on the incidence of an X-ray beam at small angle Ždenoted critical angle. on the flat surface of a support or carrier Žfor example, quartz or Perspex. on which the sample to be analyzed is deposited. In this condition the scattering effect is minimized and thus a better peak-to-background ratio is obtained reducing in this way the detection limits. Another advantage of this technique is the small volumes required for liquid sample analysis Žmicroliters. or small masses Žmicrograms. for solid samples after chemical digestion. The quantitative analysis can be made through Eq. Ž1., because the sample can be considered as a thin film so that absorption and enhancement effects can be neglected. Ii s s i C i
Ž1.
where Ii represents the intensity Žcps. for K or L X-ray line for the element i; Ci the concentration Žin ppm or g mly1 . and Si the sensitivity for this element Žcps gy1 ml.. The thin film formed on the Perspex support does not have a regular geometry and therefore the X-ray intensities depend on the thin film position. This geometric effect w1,5x can be corrected by computing the relative intensity Ž R i . wEq. Ž3.x for each element in relation to an internal standard added in every sample and standard.
Si s
the result is: R i s Si ⭈ Ci
Ž5.
where Ii and Ci are the intensity and concentration of the i th-element in the sample; IGa and CGa the intensity and concentration of the internal standard Ga in the sample; si and sGa are the sensitivity for element i and for the internal standard Ga; R i the relative intensity for the i th-element and Si its relative sensitivity Žnon-dimensional.. In a relative intensity Ž R i . vs. concentration Ž Ci . plot, the slope of the calibration line represents the non-dimensional sensitivity Ž Si . element i. Table 1 Site and date sampling for rainwater samples Site
Sample
Date
Water treatment plant 1 and 2
Ch1-1 Ch1-2 Ch1-3 Ch1-4 Ch1-5 Ch1-6 Ch2-1 Ch2-2 Ch3-1 Ch3-2 Ch4-1
3r29r98 4r29r98 5r17r98 6r19r98 7r20r98 8r03r98 3r29r98 8r03r98 3r29r98 8r03r98 3r29r98
Water treatment plant 3 and 4 Water treatment plant Capivari River Water supply Atibaia River
S. Moreira Simabuco, E. Matsumoto r Spectrochimica Acta Part B: Atomic Spectroscopy 55 (2000) 1173᎐1179
1175
Fig. 2. Non-dimensional sensitivity curve for synchrotron radiation X-ray total reflection.
Fig. 1. Linear regressions for calculating the sensitivity for the elements Ni, Co, Mn, V and Pb.
2.1. Instrumentation For the X-ray detection a hyperpure Ge detector was employed, with 140 eV resolution at 5.9 keV ŽMn K ␣ line., while for the excitation a white beam of synchrotron radiation w6x with 2 mm width and 1 mm height under total reflection conditions was used.
ple. Then an aliquot of 10 l was pipetted onto a Perspex disk and dried in vacuum thus giving rise to a shaping thin layer of approximately 5 mm diameter Žtwo replicates.. Polished quartz is normally used as sample carrier and it can be cleaned and reused while Perspex is disposable and cheap. The samples were excited for 100 s and X-ray spectra obtained were evaluated by the software QXAS w7x in order to obtain the X-ray intensities. 2.3. Sampling Data The site and date sampling for rainwater samples are shown in Table 1.
2.2. Sample and standard preparations To set up the calibration curve, five standard solutions containing the elements V, Cr, Mn, Fe, Co, Ni, Cu, Zn, As, Se, Tl and Pb were prepared at different and well-known concentrations Žfrom 0.3 to 3.8 g ly1 . with Ga addition as the internal standard to eliminate errors caused by excitationrdetection geometry. For the sample preparation, 1 ml of each sample was taken and 10 l of Ga Ž1025 g mly1 . was added resulting in a Ga concentration of 10.148 g mly1 as internal standard in the sam-
Fig. 3. Detection limit for synchrotron radiation total reflection ŽSR-TXRF. for 1000 s.
1176
S. Moreira Simabuco, E. Matsumoto r Spectrochimica Acta Part B: Atomic Spectroscopy 55 (2000) 1173᎐1179
Table 2 Measured and certified values in the standard reference material ŽDrinking Water Pollutants . Element
Cr As Se Ba Pb
Concentration Žg mly1 . Certified value
Measured value
10.00" 0.49 10.00" 0.49 5.00" 0.05 100.00" 0.53 10.00" 0.49
9.74" 0.09 10.05" 0.14 5.00" 0.25 99.98" 0.53 9.86" 0.61
3. Results and discussion With the X-ray intensities obtained from standard samples, relative intensities Ž R i . were calculated by using Eq. Ž3. and, e.g. the relationship
for the elements Ni, Co, Mn, V and Pb can be seen in Fig. 1. On the basis of the measurements performed on standard solutions, the calibration curve for synchrotron radiation total reflection as shown in Fig. 2 was obtained. The detection limits were calculated using w8,9x below: LMDi s 3 ⭈
(
Ii Ž BG . C ⭈ Ga t IGa ⭈ Si
Ž6.
where Ii Ž BG. is the background intensity for the element i; IGa the internal standard intensity ŽGa.; CGa the internal standard concentration ŽGa., Si the non-dimensional sensitivity for the element i and t the measuring time. Detection limits Žin g ly1 or ng mly1 . for synchrotron
Fig. 4. Characteristic X-ray spectrum of rainwater ŽCh1-4. measured for 100 s with SR-TXRF.
S. Moreira Simabuco, E. Matsumoto r Spectrochimica Acta Part B: Atomic Spectroscopy 55 (2000) 1173᎐1179
1177
Fig. 5. Concentration Žng mly1 . of S, Ca, Fe, Zn, Cl and Ti for rainwater samples.
radiation total reflection, calculated using Eq. Ž6. above, were extrapolated for 1000 s measuring time ŽFig. 3.. As can be visualized in this figure, the detection limits obtained for the K-shell changed from 29 ng mly1 for sulfur to 1.3 ng mly1 for zinc and copper. For the L-shell, the detection limits were 4.5 ng mly1 for mercury and 7.0 ng mly1 for lead. In general, the detection limits values obtained
are in good agreement with those reported by other workers in different synchrotron radiation facilities w1,5,10,11x. A standard reference material ŽDrinking Water Pollutants, Aldrich, 41,393-3. containing Cr, As, Se, Ba and Pb was analyzed in order to test the procedure. Table 2 shows the results and it can be see that the measured values agree well with the certified values. As can be seen in Table 1, the
1178
S. Moreira Simabuco, E. Matsumoto r Spectrochimica Acta Part B: Atomic Spectroscopy 55 (2000) 1173᎐1179
Fig. 6. Concentration Žng mly1 . of K, Ni, Mn, Cu, Cr and V for rainwater samples.
accuracy of this method is approximately 3% and the precision approximately 2%. Fig. 4 presents the spectrum for rainwater sample ŽCh1-4. measured by SR-TXRF for 100 s. The concentrations of rainwater samples were calculated using the equation presented in Fig. 2 and the results are presented in Figs. 5 and 6.
4. Conclusions The elements Ca, S, Fe, Zn and Hg presented the highest concentrations in every sample. Com-
paring the Ch1-1, Ch2-1, Ch3-1 and Ch4-1 samples collected in the same day Ž29 March 1998. and different sites, a significant variation was not observed for the same element but for the other three samples, Ch1-6, Ch2-2 and Ch3-2, sampled on 3 August 1998, a small variation can be observed in the concentration for the same element. When the Ch2-1 and Ch2-2 samples are compared, which were sampled at the same site but on different dates Ž29 March 1998 and 3 August 1998., it can be noted that the values for the Ch2-2 sample are higher than for the Ch2-1 sam-
S. Moreira Simabuco, E. Matsumoto r Spectrochimica Acta Part B: Atomic Spectroscopy 55 (2000) 1173᎐1179
ple. The same fact happened for the Ch3-1 and Ch3-2 samples, collected in the same date cited above. This can be explained by the fact that August was a rainier month than March, thus, the particulate material present in the atmosphere is higher. When it rains this material is dragged together with rainwater, resulting in higher concentrations in the sample.
Acknowledgements Research was partially performed at the National Synchrotron Light Laboratory ŽLNLS. and financial support came from FAPESP. References w1x R. Klockemkamper, A.Von Bohlen, Elemental analysis ¨ of environmental samples by total reflection fluorescence: a review, X-ray Spectrom. 25 Ž1996. 156᎐162. w2x Y. Yoneda, T. Horiuchi, Optical flats for use in X-ray spectrochemical micro-analysis, Rev. Sci. Instrum. 42 Ž1971. 1069. w3x H. Aiginger, P. Wobrauschek, A method for quantitative X-ray fluorescence analysis in the nanogram region, Nucl. Instrum. Methods 114 Ž1974. 157᎐158.
1179
w4x P. Wobrauschek, H. Aiginger, Total reflection fluorescence spectrometric determination of elements in ng amounts, Anal. Chem. 47 Ž1975. 852᎐855. w5x W. Ladisich, R. Rieder, P. Wobrauschek, Quantitative total reflection X-ray fluorescence analysis with monoenergetic excitation, Wiley, Chichester, 1994. w6x C.A. Perez, M. Radtke, H.J. Sanchez, H. Tolentino, R. ´ ´ Neuenshwander, W. Barg, M. Rubio, M.I.S. Bueno, I.M. Raimundo, J.R. Rohwedder, Synchrotron radiation X-ray fluorescence at the LNLS: beamline instrumentation and experiments, X-ray Spectrom. 28 Ž1999. 320᎐326. w7x Quantitative X-ray analysis system ŽQXAS. software package, IAEA, Vienna. w8x L.A. Curie, Limits for quantitative detection and quantitative determination, Anal. Chem. 40 Ž3. Ž1968. 586᎐593. w9x W. Ladisich, R. Rieder, P. Wobrauschek, H. Aiginger, Total reflection X-ray fluorescence analysis with monoenergetic excitation using rotating anode X-ray tubes, Nucl. Instrum. Methods Phys. Res. 330A Ž1993. 501᎐506. w10x L.M. Muia, F. Razafindramisa, R. Van Grieken, Total reflection X-ray fluorescence analysis using an extended focus tube for the determination of dissolved elements in rain water, Spectrochim. Acta Part B 46 Ž1991. 1421᎐1427. w11x P. Hoffmann, V.K. Karandashev, T. Sinner, H.M. Ortner, Chemical analysis of rain and snow samples from ChernogolovkarRussia by IC, TXTF and ICP-MS, Fresenius J. Anal. Chem. 357 Ž1997. 1142᎐1148.