You are on page 1of 4

Determination of majors in geological samples ICP-12

July 1993
by ICP-AES

Paul A Liberatore
Varian Australia Pty Ltd
Mulgrave, Victoria, Australia

Introduction
The classical analysis of silicate rocks for the major Reagent preparation4
element constituents requires the determination of Anhydrous lithium carbonate (73.89 g) and boric
Al2O3, CaO, Fe2O3, K2O, MgO, MnO, Na2O, P2O5, acid (122.6 g) were well mixed in a porcelain dish
SiO2 and TiO2. These have traditionally been deter- and slowly heated for four hours in a muffle furnace
mined and expressed as the oxides1. at 400°C. This produced a compound which was
The fusion method is applicable for the decomposi- predominantly lithium metaborate LiBO2.
tion of geological samples2,3 and for their subsequent The fused mass was ground to pass a BSS 250
analysis by ICP-AES. Lithium tetraborate or lithium mesh sieve. Lithium metaborate and lithium
metaborate is used to decompose the samples. tetraborate were mixed in a 1:1 ratio to produce the
Lithium metaborate is a better flux for highly sili- borate flux.
ceous ores and lithium tetraborate better for high
iron containing ores. The fused bead is dissolved in All reagents used were Analytical Reagent Grade.
10% nitric acid and the resultant solution analyzed Sample preparation4,5
directly.
The samples (0.5000 g) were mixed with 2.5 g of the
Experimental previously prepared borate flux in a 30 mL platinum
crucible and fused in a muffle furnace at 1000°C for
Instrumentation 30—45 minutes. The crucibles were removed from
A Varian Liberty Vacuum 200 Inductively Coupled the muffle furnace and the contents were carefully
Plasma Atomic Emission Spectrometer was used for swirled. If the fusion was incomplete the crucible
all measurements. The inert ‘V’ groove nebulizer was returned to the muffle furnace for a further 10
was chosen because of the corrosive matrix. The minutes. When the fusion was complete the crucible
operating parameters are listed in table 1. The was removed from the furnace and allowed to cool.
instrument was controlled by an Epson EL 2 PC. The crucible was placed in a tall 150 mL beaker. A
Teflon coated stirring bar was placed in the crucible
Nebulizer and 125 mL of 1.6 M nitric acid solution was added.
Type V-groove The beaker was placed on a hotplate stirrer, heated
Pressure 195 kPa to 40–60°C with vigorous stirring until complete
Plasma power 1.0 kW dissolution was obtained.
Torch Demountable with alumina injector Precautions4 required during this procedure are:-
Argon flow rate
Plasma 15.0 L/min (i) The melt should be clear and not have any dark
Auxiliary 1.5 L/min spots of sample or any floating material other-
Pump rate 15 rpm wise the process should be repeated.
Integration time 3s (ii) The melt and the acid must be at ambient
Replicates 3 temperature before commencing the dissolution
Viewing height Optimized on SBR process or the silica in the sample may form
PMT voltage 650 V polymeric hydrated silicic acid.
Grating order Default
Filter position Default (iii) Excessive heating of the solution may also
Background correction Dynamic cause the same polymerization to occur.
Snout purge High (iv) If constant stirring is not maintained some of
the rock constituents may precipitate from
Table 1: Instrument Operating Conditions solution.

1

It is also less susceptible to clog. Geological Survey Rock Standards were ence.54 0.490 nm K line.21 1. 308.76 0. other elements. Mn 257.15 0. directly overlaps.30 using Dynamic background CaO 8.15 minimal. The recommended for Si determina- high resolution of the Liberty ICP.S. The concentrations used for has sufficient signal while having calibration were those values ascribed to the rock minimum interferences from sample as shown in table 2. ration procedure described and was used as the Mg 279. which overlaps the analyte peak ging than glass concentric nebulizers. High Al concentrations will cause an elevated back- ground which can be corrected Al2O3 11. The Fe line directly Fe2O3 6. 766. A demountable interferences from other ele.When dissolution was complete.079 nm This line was chosen because it working standard.18 correction.87 interferences from other ele- ments. AES minimizes spectral interference from Ca.12 high Fe to Mn ratios this will MgO 2.12 0.41 Na 588. BCR-1 and PCC-1 were presented as to NO band emission can be unknowns. For sample preparation without any sample present.610 nm This primary line has minimum Standard values used Uncertainty interferences from other ele- (wt%) (wt%) ments. lower susceptibility to degradation in comparison with the glass concentric nebulizer particularly for P 177. found in the fusion samples. Al 396. The only docu- A blank solution was prepared by performing the mented interference is Mn. purposes for this study.495 nm There is a nearby weak Cu line fusion samples. used.49 0. K 769.941 nm The structured background due The performance is therefore to OH band emission presents no better than the “traditional” problem.215 nm Al line which suffers from an OH band structure.67 0.32 0. For very K2O 4. 2 . SY-2. the Mg/Mn ratios found in these fusion samples the effect is Standard SY-3 minimal.02 these fusion samples the effect is Na2O 4. however the concentration of the Fe is not Reference Rock samples high enough to cause an interfer- Five U.995 nm This line was chosen because it TiO2 0. however the concentrations are not significant to cause an Analytical line selection interference.152 nm This line was chosen because it Si 288. tion.204 nm The structured background due G-2. the solutions were Ca 317. The wing of the nearby Ar Table 2: Standard values used for calibration lines causes a sloping back- ground which can be minimized Results by using a higher viewing height and Dynamic background correc- The V-groove nebulizer was used because it has a tion.25 0. reduced by using a lower viewing height or by using a higher SY-3 was used as the standard for calibration plasma auxiliary flow.16 overlaps on the Mn line.03 SiO2 59. The solutions were transferred to screwtop plastic There is a weak Fe line which bottles until required for analysis. Fe 238.68 0.23 0.13 become significant. P2O5 0.158 nm This line was chosen because it has the maximum sensitivity has minimum interferences from while having the minimum other elements.02 has the maximum sensitivity while having the minimum TOTAL 98. torch with an alumina injector is ments as shown in figure 1. However for MnO 0.869 nm This line was chosen to avoid a Standard preparation Mg interference found with the SY-3 was prepared according to the sample prepa. Ti 334.933 nm The two primary Ca lines are too quantitatively transferred to 250 mL standard flasks sensitive for the levels of Ca and made to volume with distilled/deionized water.

06 Fe2O3 8.941 0. Wavelength scans of 10 ppm Al standard.43 SiO2 42.518 ± 0.11 K2O 1.91 2.005 P2O5 <0.32 ± 0.23 ± 0.158 0.29 K2O 4.995 0.02 0.77 ± 1.079 0.31 ± 0.02 CaO 2.495 0.03 ± 0.49 CaO 8.204 0.799 ± 0.18 K2O 4.017 Fe2O 3 2.70 ± 0.03 0.03 P2O5 0.0002 TOTAL 100.32 60.11 ± 0.27 8.45 ± 0.43 99.64 54.03 MgO 45.02 1.39 TOTAL 97.02 ± 2.11 P2O5 0.00002 blank and 1000 ppm Ca Na 2O 588.0003 Figure 1.67 TiO2 0.03 0.04 ± 0.37 TiO2 0.2 BCR .17 Fe2O3 13.01 0.0005 Fe 2O3 238.13 Na2O 0.06 CaO 0.49 ± 0.05 4.57 0.00 ± 1.15 CaO 7.93 99.0003 P 2O 5 177.480 ± 0.0007 CaO 317.19 0.01 0. Standard G-2 Standard PCC .15 ± 0.23 Al2O3 12.31 4.933 0.15 15.43 ± 0.30 MnO 0.54 69.07 ± 0.11 MnO 0.610 0.34 0.63 1.17 0.004 ± 0.351 ± 0.74 ± 2.01 0.06 ± 0.003 Na 2O 3.00003 standard deviation of the blank solution (the borate flux) measured 10 times with a 5 sec integration Table 3: Wavelengths used and detection limits obtained time.64 94.67 3.055 TOTAL 102.0009 The detection limit was determined as twice the TiO 2 334.70 43. The results for the measurement of four US Geologi- cal Rock standards are shown in table 4.0001 MnO 0.31 ± 0.22 Fe2O3 6.72 98.66 2.12 MgO 3.88 12.004 Na2O 4.0015 MgO 279.038 ± 0.60 0.94 ± 0. Element Wavelength Detection limit (nm) in rock (wt%) Al 2O3 396.1 Measured value Certified Value Measured value Certified Value (wt%) (wt%) (wt%) (wt%) Al 2O3 15.11 K2O 0.01 MnO 0.37 ± 0.67 ± 0.07 ± 0. MnO 257.86 13.003 Detection limit SiO 2 288.84 4.94 ± 0.02 TiO2 2.009 SiO2 63.55 Al2O3 0.69 ± 0.96 ± 0.33 41.57 6.20 ± 0.36 6.77 0.01 SiO2 71.092 ± 0.13 Na2O 3.12 0.45 ± 0.012 TiO2 <0.52 0.92 7.1 Measured value Certified Value Measured value Certified Value (wt%) (wt%) (wt%) (wt%) Al2O3 11.10 P2O5 0.00 Table 4: Results of the US Geological Rock standards analyses 3 .71 ± 0.37 4.40 ± 1.45 ± 0.05 Standard SY .40 TOTAL 100.14 ± 0.869 0.15 ± 0.42 13.40 2.09 3.1%–7% depending on the level of the element present.34 ± 0.135 ± 0.03 ± 0.152 0.0007 K 2O 769. The %RSD typically varies from 0.43 SiO2 55.048 MgO 0.27 0. The analytical wavelengths used and detection limits obtained are listed in table 3.15 0.590 ± 0.04 MgO 2.02 ± 0.

124-130 (1970). Suhr. 4 . F. 4. C. Anal. 33. Thompson. coke ash and mineral matter. 38. 7.L. (1981). Flanagan. L.J. Chem. Part 14. Australian Standard 1038. H. P. J. N. References 1. 38. 95. and Ingram. Boar. New York (1930) 2.'. Glasgow: Blackie and Son Ltd (1983). 5. C. Anal Chem.O.K. 1228 (1966). and Walsh. M.H. 81-120 (1969).Conclusion The determination of majors in geological samples by ICP-AES has been described. 6. Washington.O. Wiley. Analyst. A Handbook of Inductively Coupled Plasma Spectrometry. The values found compare very well with the certified values indicating that this methodology is totally suitable for these geological samples.N. Geochimica et Cosmochimica Acta. Ingamells.1. Analysis of coal ash. 730 (1966). The high resolution of the Liberty ICP-AES allows the user to choose the desired wavelength depend- ing on the precision and detection limits required.S. and Ingamells. 'The Chemical Analysis of Rocks. 3.