Quantitative analysis of glass using inductively coupled plasma atomic emission and mass spectrometry, laser micro-analysis inductively coupled plasma atomic emission spectrometry and laser ablation inductively coupled plasma mass spectrometry

摘要

JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1992 VOL. 7 25 1 Quantitative Analysis of Glass Using Inductively Coupled Plasma Atomic Emission and Mass Spectrometry Laser Micro-analysis Inductively Coupled Plasma Atomic Emission Spectrometry and Laser Ablation Inductively Coupled Plasma Mass Spectrometry* Lieselotte Moenke-Blankenburg Thomas Schumann and Detlef Gunther Institute of Analytical Chemistry Martin-Luther University Weinbergweg 16 0-4050 Halle Germany Heinz-Martin Kuss Department of Instrumental Analysis University of Duisburg Lotharstr. 1 W-4 100 Duisburg Germany Michael Paul Perkin-Elmer GmbH P.O. Box 101 164 W-7770 Uberlingen Germany The traditional wet-chemical analysis of glasses is time consuming because of the laborious sample preparation sequential determinations repetitive measurements and manual work in general.Alternative methods are inductively coupled plasma atomic emission spectrometry (ICP-AES) and inductively coupled plasma mass spectrometry (ICP-MS) without and with laser ablation. Inductively coupled plasma AES and ICP- MS are suitable for dissolvable glasses and both are rapid and precise methods. Laser micro-analysis ICP-AES and laser ablation ICP-MS offer direct solid sampling rapid quantitative multi-element analysis and ease of application from set-up to final report. Analytical results for three fluorophosphate glasses obtained by five methods are compared. Keywords Inductively coupled plasma atomic emission spectrometry; inductively coupled plasma mass spectrometry; laser ablation; laser micro-analysis; glass analysis As the number of formulations and applications of glass materials continues to grow the need for rapid accurate and precise determinations of the concentrations of the major and minor constituents of the finished products becomes more pressing.Whereas the classical wet-chemical analysis requires experienced staff and a great deal of time and chemicals for the manual work involved the scope of instrumental analysis of liquids suffers from the well known problems of digestion and dissolution of many solid materials and consequently time-consuming sample prepa- ration. Although these problems are avoided in the direct instrumental analysis of solids the task of quantification is difficult because there are insufficient certified reference materials for the wide range of glass matrices.In this work various methods were compared for the analysis of glasses. Experimental Wet-chemical Analysis The determination of Sr La Al Ca Mg F and P in fluorophosphate glasses is performed after digestion separ- ation and precipitation by gravimetry titrimetry and photometry as shown in Table 1. Inductively Coupled Plasma Atomic Emission Spectrometry (ICP-AES) and Inductively Coupled Plasma Mass Spectro- metry (ICP-MS) For sample preparation in ICP-AES 0.5 g of each of the pulverized glasses was dissolved with 20 ml of concentrated nitric acid distilled water was added to 100 ml the solution was heated for 2 h on a sand-bath then diluted again with distilled water up to 1 dm3 and filtered. This solution was then diluted 1 + 9 with 1.3 nitric acid.For the preparation of samples for ICP-MS 0.5 g of pulverized glasses was mixed with 20 ml of nitric acid *Presented at the XXVII Colloquium Spectroscopicurn Interna- tionale (CSI) Bergen Norway June 9- 14 199 l . (1 + 1) and digested in a microwave oven (10 min 300 W power five times). The resulting solution was diluted with water to 1 dm3. For calibration in ICP-AES commercially available standard solutions were used for all elements to be determined. These standards were mixed and the equiva- lent masses of fluoride and phosphate were added to simulate the same matrix as that present in the fluorophos- phate glasses. The concentration ranges were Al= 8-25.9 La=0.3-10.1 Ca=0.9-30 Sr= 1.6-50 Mg=0.3-10 F=3.3-104.7 and P03=0.7-23.1 mg dm-3.In ICP-MS the following calibration ranges were used Al=0.1-0.8 La=O.O5-0.4 Ca=0.1-1 Sr=O.l-2 and Mg=0.05-0.4 mg dm-3. For internal standardization a spectroscopic standard solution containing 0.1 mg dm-3 each of Sc and Rh was added to the sample solution before it was made up to the final volume. The selected spectral lines and m/z ratios are given in Table 2 and the ICP instrumentation and the experimental conditions in Table 3. Laser Micro-analysis Inductively Coupled Plasma Atomic Emission Spectrometry (LM-ICP-AES) and Laser Ablation Inductively Coupled Plasma Mass Spectrometry Laser sampling is becoming recognized for its capability of analysing solids directly and greatly reducing or eliminating the need for sample preparation. With transparent glasses it is recommended that rough surfaces are used rather than polished surfaces.The laser instrumentation and the experi- mental conditions are given in Table 4. The LMA 10 laser microscope3 was coupled with the Spectroflame ICP spectrometer with the parameters given in Table 3. The Model 320 laser sampler was coupled with the Elan 5000 spectrometer with the parameters power 1.2 kW plasma flow rate 15.5 dm3 min-l carrier flow rate 0.84 dm3 min-I and auxiliary flow rate 0.6 dm3 min-'. For quantification in LM-ICP-AES liquid-solid calibra- (LA-ICP-MS)252 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1992 VOL. 7 Table 1 Classical determination of the major components of fluorophosphate glass Digest ion Element separation Sr Na2C03 Ca Filtrate from Sr precipitation from Ca precipitation hydroxides of La and Al dissolved in HCI Mg Filtrate La Na2C03 NHj-NHdCI A1 HNO F Distillation P Na2C03 + K2C03 Precipitation Analytical method With HNO 8Ooh; as aqueous Sr dried ( I 30 "C) With HZC2O4; as CaC20 heated (900 "C) With (NH4)2HP0,; as MgNH4P04.6H20 heated (900 "C) as aqueous h 2 ( c 2 0 4 ) 3 heated (900 "C) With HZC2O4; With NH,-NH,CI; as hydroxides of La and Al heated ( 1 100 "C) With KBr-PbNO,; as PbBrF Reaction with ammonium molybdate Gravimetry Gravimetry CaO Gravimetry Mg2P2O7 Gravimetry L a 2 0 3 Argent imet ry Photometry AgN03 + NHdSCN Molybdenum Blue Table 2 Wavelengths used in ICP-AES and m/z ratios used in ICP-MS Wavelength/ Element nm mlz Al 396.1 27 La 379.4 139 Ca 3 17.9 44 Sr 346.5 86 Mg 279.8 24 F N.d.* N.d.(LA-ICP-MS 19)t P N.d. N.d. (LA-ICP-MS 3 I)? N.d. =not detectable. t LA=laser ablation. ~~~~ ~ Table 3 Instrumentation and experimental conditions in ICP- AES and ICP-MS Parameter ICP-AES2 ICP-MS Type Spectroflame Elan 500 Power 1.25 kW 1.25 kW Gas Ar Ar Plasma gas flow rate 10.7 dm3 min'' 12.0 dm3 min-l Camer gas flow rate 1.5 dm3 min-l Auxiliary gas flow rate 0.5 dm3 min-' 1.4 dm3 min-l 1 .O dm3 min-' Nebulizer Cr oss-flow Cross-flow Optics Paschen-Runge Quadrupole gratings 2400 mass filter grooves mm-l Detector Photomultiplier Channel electron multiplier tion'-' was used. Fig. 1 shows the experimental apparatus. The double gas flow meets the requirements for analysis of both the solid analyte and the liquid standard under identical spectroscopic conditions and for simultaneous multi-element analysis for at least two elements i.e.the element to be determined and a reference element. In the first step aqueous standard solutions were nebu- lized in the normal manner but carried by only one part of the divided argon stream. The other part of the gas stream was allowed to flow through the laser ablation chamber but without laser action. Intensities were measured and calibra- tion graphs constructed for the element to be determined Table 4 Laser instrumentation and experimental conditions Parameter LM-ICP-AES LA-ICP-MS TY Pe Laser resonator Wavelength Operation mode Repetition rate Energy output Ablation pattern Crater diameter LMA 10 laser micro- analyser Ruby 694 nm Q-swi tched (6 steps Qs used 50 ns) Single shot 100 mJ Single spot 100 pm Model 320 laser sampler Nd Y AG 1064 nm Q-swi tc hed 8 ns 10 Hz 250 mJ Single spot I mm and for a reference element.In the second step intensities for the laser-ablated solid (analyte and reference element) were measured using both flow streams one flowing through the ablation chamber transporting the aerosol and the other transporting a blank solution. The two argon streams were mixed in front of the ICP torch so that the water or liquid introduction into the plasma was the same thus maintaining similar plasma conditions. The desired concentration of analyte was calculated by using the equation csa = (Cla/clr) csr where c is concentration sa stands for solid analyte cIa and cIr represent the liquid concentration of the analytical element a and the reference element r and sr respectively represents the solid reference element.The results of the quantitative analyses of three fluoro- phosphate glasses are compared in Table 5 . The performance of the analytical chemical methods is characterized by the time needed for sample preparation and analysis by the accuracy of the results and the precision of the procedure. Whereas the classical wet-chemical analysis is very time consuming (at least 1 week is required for a double determination of seven elements in one sample) instru- mental methods such as ICP-AES ICP-MS LM-ICP-AESJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1992 VOL. 7 253 Table 5 Quantitative analysis of fluorophosphate glasses No. 143 (used as external standard for LA-ICP-MS of fluorophosphate glasses No.277 and No. 289) No. 277 and No. 289. All results given at the 95 confidence level with n- 1 degrees of freedom Classical determination XkAX Glass Element (mass percent.) No. 143 Al 9.5 f 0.08 La 3.3 f 0.02 Ca 11.8f0.12 Sr 20.6 f 0.23 Mg 2.7f0.02 F 41.8 f 0.46 P 3.9 f 0.03 n 5 Internal standard No. 217 A1 9.2 f 0.08 La 2.8 * 0.01 Ca 11.2-cO.lI Sr 20.3 20.23 Mg 3.1 fO.02 F 41.8 20.46 P 4.5 k 0.04 n 5 Internal External standard standard No. 289 Al 9.5 f 0.08 La 2.8 2 0.01 Ca I1.6f0.12 Sr 20.3 2 0.23 Mg 2.8 ? 0.02 F 4 1.4 k 0.45 P 4.3 k 0.04 n 5 Internal External standard standard * ND = not detectable RSD (I 0.6 0.4 0.7 0.8 0.4 0.8 0.6 0.6 0.4 0.7 0.8 0.4 0.8 0.6 0.6 0.4 0.7 0.8 0.4 0.8 0.5 ICP-AES LM-ICP-AES X k A x RSD X f A t (mass percent.) () (mass percent.) 9.2 20.10 3.1 f 0.03 1 I .7 f 0.20 2 I .3 f 0.35 2.6 f 0.02 - N D* 10 9.5 -+ 0.06 2.620.01 11.5k0.08 20.1 k 0.23 2.5 -+ 0.05 - ND* 10 9.6 k 0.08 2.7 f 0.02 I I .3 f 0.05 2.5 f 0.03 19.920.1 1 - ND* 10 1.5 I .3 2.4 2.3 1.3 N D* - 0.9 0.8 I .o I .6 3.0 N D* - 12.2 k0.3 I 3.2 2 0.04 I1.9k0.14 21.420.35 2.6 f 0.02 N D* - 10 La; Ca for La 12.5 2 0.35 2.6 2 0.06 I1.3f0.16 20.6 k 0.18 2.6 f 0.05 - N D* 10 La; Ca for La 1.1 13.420.58 1.1 3.0+0.05 0.6 10.720.17 0.8 20.3 ? 0.16 1.5 2.620.05 - - N D* N D* La; Ca for La 10 RSD (I 3.6 1.7 1.7 2.3 2.I ND* - 3.9 3.2 2.0 1.2 2.7 ND* - 6.1 2.2 2.2 1.1 2.5 ND* - ICP-MS LA-ICP-MS x+Ax RSD XkAA RSD (mass percent.) () (mass percent.) () 9.4 f0.03 0.45 3.2 f0.02 0.65 11.5fO.03 0.31 2 1.2 2 0.05 0.35 2.5 20.01 0.32 - - ND* N D* 10 Sc for Mg Al Ca; Rh for Sr La 9.9 1 2 0.03 1 0.43 2.8 I 2 0.020 1 .OO 11.51 k0.035 0.43 2.54k 0.006 0.35 2 1.3 I f 0.070 0.46 L - ND* N D* 10 Sc for Mg Al Ca; Rh for Sr La 9.89 2 0.028 0.39 2.8620.017 0.22 1 1.73 4 0.024 0.28 21.36+0.049 0.32 2.57 2 0.008 0.4 I ND* ND* L - 10 Sc for Mg Al Ca; Rh for Sr La 9.69 2 0.08 2.39 f 0.06 10.73 f 0.10 20.49 -+ 0.I0 2.54 k 0.04 34.172 1.69 3.98 k 0.04 5 and "Sr Glass No. 143 9.65 f 0.12 2.3 I amp; 0.09 10.51 f0.I 20.55 -e 0.07 2.5420.06 35.83 rt 1.94 4.18 rt 0.02 5 "Sr and Glass No. 143 0.8 2.6 0.95 0.5 I .6 4.96 0.9 I .24 3.97 0.95 0.36 2.48 5.43 0.63 and LA-ICP-MS offer the possibility of rapid analyses (a few hours for sample preparation and calibration and seconds or minutes for each analysis).Accuracy characterizes the agreement between measured values (or their mean) and the true value. In this work the traditionally determined compositions were assumed to be true values (the analyst at the glass factory' guaranteed the accuracy of the results from experience over many years and certified the composition of each glass as given in the first column of Table 5; no suitable reference materials of this type of glass are commercially available). Comparison of the results of the four instrumental methods with the results of the wet-chemical method shows good agreement with one exception. The determination of aluminium (b) (a) 5 r L Ar I 6a 1 I - 5 I tiAr Ar G 3 t-. I Ar I 6b Fig. 1 Quantification in LM-ICP-AES by liquid-solid calibration. 1 Laser microscope; 2 sample chamber with analytical sample; 3 poly(viny1 chloride) tube; 4 Y-piece as adapter for LM-ICP; 5 ICP; 6a bottles for calibration samples; 6b bottle with blank solution and 7 analytical sample254 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1992 VOL.7 causes some problems when using LM-ICP-AES. This could be because of the tendency of AlF3 to sublime at 1291 "C whereas the fluorides of La Mg Sr and Ca vaporize at 2330,2240 2460 and 2500 "C respectively and therefore a fractionated ablation was assumed. Precision measures the agreement between the results OY repeated measurements. Random (statistical) errors occur and are represented by the standard deviation (SD). The results in Table 5 are characterized by the confidence interval which is the product of the standard deviation and Student's t-factor (for p= 0.95 as the statistical confidence level and the number of degrees of freedom f = n - I) divided by the square root of the number of parallel determinations.The precision in comparison with the wet- chemical process is best with ICP-MS and acceptable with the other methods. The advantages of laser micro-analysis and laser ablation techniques are the minimized sample preparation and handling and avoidance of contamination and losses. The detectability of F+ is noteworthy because F has an ionization energy (1 7.42 eV) greater than that of Ar (1 5.76 eV). Whereas most elements are ionized to more than 90 in the high-temperature environment of the ICP F is ionized to about 2 x 1 0-4 assuming local thermodynamic equilibrium T,,,(Ar)=6680 K and f i e = 1.47 x lOI4 ~ m ' ~ the degree of ionization of F is 0.000 19 19 (ref.8)J. By direct laser ablation with a high-repetition laser sample masses in the milligram range could be ablated and the high concen- tration of F in the fluorophosphate glasses may therefore be detectable. Despite this a high-power Q-switched laser is able to ionize the elements in the sample partly and the free F ions would be transported within the ablated material by the carrier gas stream into the plasma. Interference from OH3+ can be excluded because the argon aerosol is dry only a small peak of a few counts per second being recorded when analysing the camer gas without the sample aerosol. Conclusions The methods presented here for the determination of major elements in fluorophosphate glasses demonstrate again that instrumental techniques based on the ICP with AES or MS without and with laser ablation are promising and could replace traditional wet-chemical methods.Comparison of the results shows that both the accuracy and precision are good. References Scheidhauer G. Jenoptik Jena Jena Germany unpublished data. Schumann T. Diploma Thesis Martin-Luther University Halle 1990. Moenke-Blankenburg L. Laser Microanalysis Wiley New York 1989. Thompson M. Chenery S. and Brett L. J. Anal. At. Spectrom. 1989 4 1 1. Gunther D. and Gackle M. Doctoral Thesis MartinLuther University Halle 1990. Moenke-Blankenburg L. Gackle M. Gunther D. and Kam- mel J. in Plasma Source Mass Spectrometry eds. Jarvis K. F. Grey A. L. Williams J. G. and Jarvis I. Royal Society of Chemistry Cambridge Special Publication No. 85 1990 pp. Moenke-Blankenburg L. and Giinther D. G e m . Geol. 1992 95 85. Horlick G. Tan S. H. Vaughan M. A. and Shao Y. in Inductively Coupled Plasmas in Analytical Atomic Spec trome- try ed. Montaser A. and Golightly D. W. VCH Weinheirn 1987 p. 364. 1-17. Paper I /02909F Received June 17 1991 Accepted October 9 1991