Beruflich Dokumente
Kultur Dokumente
(1) Centre for Ore Deposit Research, University of Tasmania, GPO Box 252-79, Hobart, Tasmania, 7001, Australia
(2) School of Earth Sciences, University of Tasmania, GPO Box 252-79, Hobart, Tasmania, 7001, Australia.
* Corresponding author. e-mail: Phil.Robinson@utas.edu.au
Complete dissolution is essential to obtain accurate La dissolution complète est essentielle pour obtenir
analytical results using ICP-MS. In this study, des résultats analytiques justes par ICP-MS. Dans cette
decomposition techniques (i.e. acid digestions using étude, les techniques de décomposition (i.e., attaques
Savillex Teflon vials, a high pressure digestion acides utilisant des fioles Savillex® en téflon, un système
system and microwave oven, a combined lithium d’attaque à haute pression et un four à micro-ondes,
tetraborate fusion - HF/HNO 3 acid decomposition une fusion au tétraborate de lithium combinée avec
and sodium peroxide sinter) for the total dissolution une décomposition acide HF/HNO3 et un frittage
of different types of geological reference materials avec peroxyde de sodium) pour la dissolution totale
have been investigated. Savillex Teflon vial de différents types de matériaux géologiques de
HF/HNO 3 digestion is effective for basaltic samples. référence ont été étudiées. L’attaque HF/HNO3 en
The high pressure HF/HClO 4 digestion (PicoTrace flacon téflon Savillex est efficace pour les échantillons
TC-805 digestion system, Bovenden, Germany) de basalte. L’attaque HF/HClO4 à haute pression
allows dissolution of basalts and ironstones. (système d’attaque PicoTrace® TC-805, Bovenden,
Granites and magnetite-rich samples can be Allemagne) permet la dissolution des basaltes et des
dissolved using a high pressure HF/H 2SO 4 method. roches riches en fer. Les granites et les échantillons
Geological samples cannot be effectively attacked riches en magnétite peuvent être dissous par la
by microwave acid digestion. A combined lithium méthode HF/H2SO4 à haute pression. Les échantillons
tetraborate fusion - HF/HNO 3 acid digestion géologiques ne peuvent pas être efficacement attaqués
method allows complete dissolution of many par une attaque acide en four à micro ondes. Une
different types of geological materials; however, this méthode combinée – fusion avec tétraborate de
method precludes the determination of volatile lithium et attaque acide HF/HNO3 permet la dissolution
elements due to the high fusion temperature complète de nombreux différents types de matériaux
(1000 °C). A sodium peroxide sinter method at géologiques sauf la détermination des éléments
480 °C has the potential for the rapid determination volatils due à une haute température de fusion
of Y, Sc and REE in different types of geological (1000 °C). Le frittage avec peroxyde de sodium à
materials. However, the lack of ultra-pure reagents 480 °C permet la détermination rapide de Y, Sc et
precludes the use of lithium tetraborate fusion des Terres Rares dans différents types de matériaux
and sodium peroxide sinter methods for the géologiques. Toutefois, l’absence de réactifs ultra-
measurement of geological samples with low trace purs exclut l’utilisation de la fusion avec tétraborate
element abundances. de lithium et du frittage avec peroxyde de sodium
pour l’analyse de faibles teneurs d’éléments en
Keywords: chemical decomposition, geological traces dans des échantillons géologiques.
reference materials, trace element determination,
inductively coupled plasma-mass spectrometry, Mots-clés : décomposition chimique, matériaux
comparison of techniques. géologiques de référence, détermination d’éléments
en traces, spectrométrie de masse couplée à une
torche à plasma induit, comparaison de techniques.
Sample decomposition is a fundamental and 1. Savillex screw-top Teflon vial acid digestion
critical stage in the process of geochemical sample 2. PicoTrace high pressure acid digestion
analy sis. It is often the limiting fac tor to sample 3. Microwave oven (sealed vessel, acid digestion)
throughput, which is particularly true with the applica- 4. Combined Li 2 B 4 O 7 fusion + HF/HNO 3 acid
tion of fast and modern multi-element measurement digestion
instrumentation, such as ICP-MS and ICP-AES (Chao 5. Sodium peroxide sinter
and Sanzolone 1992).
Experimental
Many digestion methods have been used for the
decomposition of geological samples. These include Instrumentation
open and closed vessel acid digestion methods
(Moselhy et al. 1978, McQuaker et al. 1979, Lechler A Finnigan MAT Element ICP-mass spectrometer
e t a l . 19 8 0 , C h u rc h 19 81, U c h i d a e t a l . 19 8 0 , (Bremen, Germany) was used in this study. This instru-
McLaren et al. 1981, Hee and Boyle 1988, Karstensen ment was equipped with a compact, double-focusing,
and Lund 1989, Jarvis 1990, Boer et al. 1993, Dulski magnetic sector of reversed Nier-Johnson geometry.
1994, Münker 1998, Liang et al. 2000), microwave The mass range of the double-focusing magnetic
dissolution (Lamothe et al. 1986, Borman 1988, sector field analyser extended from 5 to 260 amu. Pre-
Kingston and Jassie 1988, Mathes 1988, Matusiewicz defined resolution settings (M/∆M at 10% valley defi-
and Sturgeon 1989, Kemp and Brown 1990, Nölter et nition) of 300 (low), 3000 (medium) and 7500 (high)
al . 1990, Totland et al. 1995, Yoshida et al. 1996), allowed the operating resolution to be adjusted,
Na 2O 2 sintering (Robinson et al. 1986, Longerich et depending mainly on the analytical matrices encoun-
al. 1990) and fusions with alkali fluxes such as LiBO 2, tered. In this investigation, isotopes of interest were
Li 2B 4O 7, Na 2O 2 etc. (Burman et al. 1978, Bankston et measured using electric scanning, with the magnet
al. 1979, Walsh and Howie 1980, Brenner et al. held at fixed mass. As reported in an earlier study
1980, Jarvis 1990). Totland et al. (1992) evaluated (Robinson et al. 1999), 45 Sc was measured using
results from open vessel acid digestion, LiBO2 fusion, medium resolution in order to avoid the interferences
and microwave-heated sealed-vessel acid digestion of polyatomic species such as 29 Si 16 O, 28 Si 16 O 1 H,
for thirty eight elements by ICP-MS analysis in nine 13 C 16 O etc. and the remainder of trace elements
geological reference materials of varied composition. of interest were all measured using low resolution.
Reviews of decomposition techniques in geological Instrument tuning and optimisation were performed
materials are given in Potts (1987), Heinrichs and daily using a 10 ng g -1 multi-element solution contai-
Herrmann (1990), Jarvis (1992), and Chao and ning the elements of interest. In general, instrument
Sanzolone (1992). parameters were adjusted to obtain optimum intensity
and stability. The typical instrumental settings for the
Despite this multitude of techniques, complete study are summarized in Table 1. Further details regar-
dissolution of all sample types is still a problem. In ding this instrument have been reported elsewhere
practical ICP-MS analysis, care must be taken when (Moens et al. 1994, Townsend et al. 1998, Robinson et
adapting methods for multi-element determination, al. 1999, Yu et al . 2000).
since many of the digestion procedures were develo-
ped for measurement of only a few elements. The solu- Digestion apparatus
tion chemistry involved in many dissolution procedures
is poorly understood and refinements and understan- PicoTrace high pressure digestion system: The
ding of these techniques have not kept pace with the acid pressure digestion system used in the study was a
rapid development in modern analytical instrumenta- P i c o Tr a c e TC - 8 0 5 P r e s s u r e D i g e s t i o n S y s t e m
tion (Jarvis 1992). (Bovenden, Germany). Digestion time and temperature
could be controlled automatically. Geological mate-
Therefore, in order to achieve a complete dissolu- rials were digested in 30 ml PTFE (polytetrafluoroethy-
tion and hence obtain reliable multi-element analytical lene) digestion containers using procedures in which
data using ICP-MS, a sy stematic investigation of temperatures and pressures did not exceed 180 °C
decomposition techniques for a range of basalt, granite and 2000 kPa. During pressure digestion, the PTFE
and ironstone geological reference materials is descri- containers were inserted in a PFA- (perfluoroalkoxy-
bed in this paper. The methods investigated are: ethylene) coated digestion block and tightly sealed
200
GEOSTANDARDS
NEWSLETTER
The Journal of Geostandards and Geoanalysis
Table 1.
Typical instrumental settings
with PTFE lids, which were reinforced by pressure disks quartz distillation system. Hydrofluoric acid (50% w/w
and plates. Two sixteen container blocks on the hotplate Analar, AR grade) was doubly distilled in a sub-boiling
allowed digestion of thirty two samples at a time. Teflon distillation system and ultra-pure HClO 4 (70%
During evaporation, all PTFE containers remained in w/w ARISTAR , BDH Chemicals) was also used. Ultra-
the pressure digestion unit and were covered with one pure H 2 SO 4 (double distilled sub-boiling in quartz)
large PTFE evaporation plate. Acid fumes were remo- was obtained from Seastar Chemicals Inc., Canada.
ved by a water vacuum pump and absorbed by a Ultra-pure water (≥ 16.7 MΩ) used in the study was
strong alkaline solution. No cross contamination bet- prepared by distillation with a glass distillation system
ween digestion containers was observed during the and then further purified with a MODULAB Water
evaporation stage. P u r i f i c a t i o n S y s t e m ( C o n t i n e n t a l Wa t e r S y s t e m
Corporation, Melbourne, Australia). Li 2B4O7 (AR grade,
Milestone microwave: Microwave decomposition Sigma Chemicals, WA, Australia) and Na 2 O 2 (AR
was carried out with the MLS-1200 MEGA Microwave grade, Merck) were used for the fusion and sinter
Digestion System (Milestone, Italy) in combination experiments, respectively, in this work.
with MDR (microwave digestion rotor) 1000/6/100/110
system and 100 ml TFM (tetrafluormethaxil) vessels. External calibration standards were prepared by
The power emission of the microwave system could gravimetric serial dilution from 100 µg ml -1 multi-
be controlled from 10 to 1000 W (in 10 W incre- element standards (QCD Analyst, USA). The standard
ments). Different digestion programmes could be cho- solutions were prepared at the beginning of each
sen for different samples. There were several steps analytical sequence with doubly distilled water and
with each digestion programme, and a different time 2% w/w HNO 3. The indium standard solution (1000 ±
could be set for each step of digestion. The maximum 3 µg ml -1 in 2% w/w HNO 3, High-Purity Standards,
pressure in the vessel could reach as high as 11000 South Carolina, USA) was diluted to working concen-
kPa. The properties of TFM include high density, high tration in 2% w/w HNO 3 before use as an internal
insulating properties and high working temperature standard for ICP-MS analysis.
( ca. 330-350 °C).
Geological reference materials
Reagents and calibration
A range of basalt, granite and ironstone geological
Concentrated acids were used exclusively in this reference materials was chosen to investigate the cha-
study. Hydrochloric acid (33% w/w BDH, AR grade) racteristics of different digestion techniques (Table2).
and HNO 3 (70% w/w BDH, AR grade) were used International geological reference materials were
after further double distillation in a sub-boiling mainly used, along with two in-house rock reference
201
GEOSTANDARDS
NEWSLETTER
The Journal of Geostandards and Geoanalysis
Table 2.
A list of geological reference materials studied
BHVO-1 Basaltic lava from Kilauea caldera, Kilauea volcano, Hawaii, USA Govindaraju (1994)
TAFAHI Basalt collected from north Tonga, used as an in-house reference material by the Eggins et al. (1997)
Research School of Earth Sciences, Australian National University
AC-E Granite from a quarry in Ailsa Craig island in the Firth of Clyde, southwest Scotland. The Potts and Holbrook (1987),
Ailsa Craig microgranite comprises alkali feldspar (low albite with small proportions of Govindaraju (1987)
orthoclase) and quartz with less 10% of amphibole, pyroxene, aenigmatite and accessory
minerals. It contains significant amounts of both light and heavy REE giving a flat chondrite-
normalized abundance distribution at about the 100 times chondrite level
GSR-1 A grey medium-grained biotite granite from Binzhou, Hunan province, PRC. Mineralisation Xie et al. (1985, 1989)
of tungsten, tin and molybdenum occurs in the contact zone between the granite and
carbonate rocks
G-2 Westerly granite from Sullivan quarry, Bradford, Rhode Island, USA, collected by Felix Chayes, Flanagan (1967)
Geophysical Laboratory, Carnegie Institution of Washington. G-2 is rich in plagioclase Govindaraju (1994)
and biotite
MA-N Albite-lepidolite granite (“quartz albititite”), from Beauvoir (Clemont-Ferrand, Massif Central), Govindaraju (1980, 1994)
France. MA-N is a special granite in the sense that transition elements and REE are at
very low concentrations whereas elements such as Ag, Be, Cd, Cs, Ga, Li, Nb, Rb, Sn, Ta
and W are at “unusually high concentrations”, and it also contains cassiterite
YG-1 Medium-grained felsic granite with occasional feldspar megacrysts up to 20 mm from the Thompson et al. (1999)
Yewangara pluton, Wyangala batholith, Lachlan granite province, New South Wales, Australia
FeR-2 Iron formation sample (magnetite rich) from Griffith Mine, Bruce Lake, Ontario, Canada Govindaraju (1994)
FeR-4 Iron formation sample (magnetite rich) from Sherman Mine, Temagami, Ontario, Canada Govindaraju (1994)
TASBAS Basalt from Tasmania, Australia (an in-house reference material at School of Earth Sciences, Robinson et al. (1986)
University of Tasmania)
TASGRAN Granite from St Helens, N.E. Tasmania, Australia (an in-house reference material at School of Robinson et al. (1986)
Earth Sciences, University of Tasmania). Contains potassium feldspar, plagioclase, biotite
and hornblende
samples (TASBAS and TASGRAN) from the School of shaken occasionally. The vial was removed from the
Earth Sciences of the University of Tasmania. hotplate twice (at the beginning and in the middle of
the digestion) and placed in an ultrasonic bath for a
Savillex Teflon vial acid digestion couple of minutes. The solution was evaporated on a
hotplate at 130-150 °C to incipient dryness. The eva-
Acid digestion in Teflon vials on a hotplate is a poration was repeated twice more after adding 1 ml
simple and convenient way to decompose geological HNO 3 each time. The residue was dissolved using 2
samples for analysis. Silicon is removed as volatile SiF 4 ml HNO3 and 3-5 ml water on the hotplate and dilu-
and no extra salts are added to the solution. ted to 100 ml with ultra-pure water in a polypropylene
bottle. All solutions were freshly analysed within 24
Digestion procedure hours of dissolution by ICP-MS.
202
GEOSTANDARDS
NEWSLETTER
The Journal of Geostandards and Geoanalysis
Table 3.
Measured trace element concentrations (µg g -1 ), precision and reference
values for BHVO-1, TAFAHI and TASBAS by Savillex Teflon vial digestion
Digestion conditions: 2 ml HF and 0.5 ml HNO3 for 100 mg powdered rock sample. [ ] information value.
See Table 2 for sources of reference values.
203
GEOSTANDARDS
NEWSLETTER
The Journal of Geostandards and Geoanalysis
204
GEOSTANDARDS
NEWSLETTER
The Journal of Geostandards and Geoanalysis
Table 4.
Measured results (µg g -1 ) for basalt and ironstones using
3 ml HF/3 ml HClO 4 high pressure (PicoTrace) digestion
DT is digestion time. * Reference values are from Govindaraju (1994). - No value available.
[ ] Information value.
205
GEOSTANDARDS
NEWSLETTER
The Journal of Geostandards and Geoanalysis
Table 5.
A list of digestion conditions (groups A to E) using high pressure (PicoTrace) digestion
Group C 3 ml HF + 3 ml HClO4 (1) Add 5 ml HCl and evaporate to dryness. 5% w/w HCl
(2) Add 5 ml HCl again and digest at 150 °C
for 16 hr.
Group D 3 ml HF + 3 ml HClO4 (1) Add 5 ml HCl and evaporate to dryness. 2% w/w HNO3/1% w/w HCl
(2) Add 2 ml HNO3 and 1 ml HCl and digest at
150 °C for 16 hr.
Group E 3 ml HF + 3 ml HClO4 (1) Add 5 ml HCl and evaporate to dryness. 2% w/w HNO3/1% w/w HCl
(2) Add 5 ml HCl again and digest at 150 °C
for 16 hr, evaporate to dryness.
Results for the determination of thirty two elements averaging 95% (90-103%), were similar for digestion
in the three granites showed recovery rates (i.e., in methods A to E, suggesting that Nb and Ta are pre-
comparison with reference values) for digestion groups sent in minerals that are easily dissolved. Granite AC-E
A, B, C, D and E of 75%, 75%, 88%, 82% and 96% gave particularly poor results for the REE and HFSE
respectively. The best results (C and E) were obtained (except Nb and Ta) with methods A and B, which used
from a two stage pressure digestion with HF/HClO 4, a single stage HF/HClO 4/HNO 3 pressure digestion.
followed by HCl, similar to the procedure used by
Dulski (1994) for Ba and REE determinations. The only Münker (1998) reported that, even after a few
difference between C and E is that E has a final acid hours, the sample solution may lose Nb and Ta. The
concentration of 2% w/w HNO 3 and 1% w/w HCl addition of a small amount of HCl can stabilize Nb
instead of 5% w/w HCl. However, the 96% average and Ta in solution before ICP-MS analysis. In this work,
recovery for E hides the low yields of 80% to 92% for all solutions were freshly prepared and ICP-MS runs
Zr and Hf (see GSR-1 in Figure 3). Refractory zircon were always conducted within 24 hours. The digestions
was not completely dissolved. The REE, Th and U in groups A and B were identical, apart from a final
recoveries were 20-30% lower using methods A, B solution prepared in 2% w/w HNO 3 instead of 2%
and D, where the second pressure stage with HCl was w/w HNO3 and 1% w/w HCl. No obvious difference
not used. Niobium recovery was 100 ± 5% and Ta, in Nb and Ta results was observed between A and B.
HF/H 2 SO 4
206
GEOSTANDARDS
NEWSLETTER
The Journal of Geostandards and Geoanalysis
Table 6.
Measured concentrations (µg g -1 ), precisions and reference values for GSR-1, AC-E, G-2, MA-N,
YG-1, TASGRAN, FeR-2, FeR-4, TAFAHI and TASBAS by 3 ml HF/3 ml H 2 SO 4 high pressure digestion
207
GEOSTANDARDS
NEWSLETTER
The Journal of Geostandards and Geoanalysis
Table 6 (continued).
Measured concentrations (µg g -1 ), precisions and reference values for GSR-1, AC-E, G-2, MA-N,
YG-1, TASGRAN, FeR-2, FeR-4, TAFAHI and TASBAS by 3 ml HF/3 ml H 2 SO 4 high pressure digestion
See Table 2 for sources of reference values. - no value available. [ ] information value.
208
GEOSTANDARDS
NEWSLETTER
The Journal of Geostandards and Geoanalysis
3 ml HF + 0.5 ml H 2SO4 16 hours Figure 4. The recoveries of HFSE and REE for GSR-1 under various
3 ml HF + 1 ml H 2SO 4 16 hours HF/H 2SO 4 high pressure digestion conditions. All digestions were
2 ml HF + 0.5 ml H 2SO4 16 hours with100 mg rock powder at 180 °C. Digestion conditions: GSR-1a
2 ml HF + 1 ml H 2SO 4 16 and 48 hours = 16 hour digestion with 3 ml HF + 0.5 ml H 2SO 4; GSR-1b = 16
2 ml HF + 2 ml H 2SO 4 48 hours hour digestion with 3 ml HF + 1 ml H 2SO 4; GSR-1c = 16 hour
digestion with 2 ml HF + 0.5 ml H 2SO 4; GSR-1d = 16 hour
The complete results of two independent digestions digestion with 2 ml HF + 1 ml H 2SO 4; GSR-1e = 48 hour
under each HF/H 2 SO 4 ratio and digestion time are digestion with 2 ml HF + 1 ml H 2SO 4; GSR-1f = 48 hour digestion
available from the authors on request. From these with 2 ml HF + 2 ml H 2SO 4; GSR-1g = 16 hour digestion with 3
results it can be seen that: ml HF + 3 ml H 2SO 4.
209
GEOSTANDARDS
NEWSLETTER
The Journal of Geostandards and Geoanalysis
Table 7.
Comparison of power and pressure used in different microwave digestion studies
Parameter This work Jarvis et al. (1992) Totland et al. (1995) Yoshida et al. (1996)
about 97% when 3 ml HF and 3 ml H 2SO4 were used. oven and the residue was redissolved for one minute
The low recovery for Ta was due to hydrolysis. Although at each power stage. The solution was finally transfer-
fresh solutions (within 24 hours after dissolution) were red to a polypropylene bottle and diluted to 100 ml
normally analysed in this work by ICP-MS, approxima- with ultra-pure water. Two other sets of conditions
tely 15% Ta was lost. were tried, (1) as above but with 10 minutes for each
power stage instead of 5 minutes and (2) 2 ml HF +
Microwave digestion 0.5 ml HNO3 + 2 ml HClO4 + 0.2 ml H 2SO 4 with 10
minutes for each power stage.
In recent years, microwave digestion techniques
have been used to digest rock samples for ICP-MS Results
(e.g. Jarvis et al. 1992, Totland et al. 1995, Yoshida et
al. 1996). Ideally, in comparison with the high pressure An attempt was made to digest BHVO-1, GSR-1,
digestion method, the microwave digestion technique TASGRAN and FeR-2 by the microwave digestion
should be less time-consuming, with rock samples methods. Figure 5 shows the comparison between the
being digested within one hour or so. However, Jarvis measured results and reference values (Govindaraju
et al. (1992) found that microwave digestion did not 1994).
always allow the measurement of elements such as Zr
and Hf, especially when refractory mineral phases Basalt BHVO-1 recovery was similar (about 91%)
were present. In some cases, measurement of the REEs for all three sets of conditions. Average recovery was
seemed to be a problem due to incomplete dissolution 98.5% for U, better than 90% for the light REEs and
of rock samples (Yoshida et al. 1996). The work repor- most trace elements, 85% for the heavy REEs and 83%
t e d h e r e u s e d a M LS - 12 0 0 M E G A M i c r o w a v e for Zr. Yttrium showed the lowest recovery with less
Digestion System, which is a more powerful system than 80%. These recovery values are similar to those
(Table 7) than those used by Jarvis et al. (1992), for the microwave digestion of basalts reported by
Totland et al. (1995) and Yoshida et al. (1996). Yoshida et al. (1996), who found that average recove-
ries of U, Zr and Y for two Japanese basalt reference
Digestion procedure materials (JB-1 and JB-2) were 96%, 90.5% and
84.5%, respectively.
The sample (100 mg) was weighed into a TFM
(tetrafluormethaxil) vessel. After wetting with ultra-pure Recoveries for the granites TASGRAN and GSR-1
water and spiking with indium as an internal stan- were poor. TASGRAN averaged 69% recovery with little
dard, 2 ml HF and 0.5 ml HNO 3 were added to each difference between the three digestion methods. The
vessel. After thorough mixing, the vessel was introdu- recovery for GSR-1 increased from 68.6% using 5
ced into the microwave oven (Milestone MLS-1200 minutes for each power stage to 76% (10 minutes
MEGA with maximum pressure of 11000 kPa) and each stage) and 83.7% (10 minutes), when the oxidi-
digested at 250 W, 400 W, 650 W and 250 W for 5 sing acids HClO 4 and H 2SO4 were present as well as
minutes each, followed by venting for 5 min. The ves- the HF and HNO3. The average REE increased similarly
sel was then carefully removed onto a hotplate at from 61% to 72% to 77%. Recoveries of 100% were
about 150 °C and the solution evaporated to a crys- found for Nb in both granites. Zirconium and Hf reco-
talline paste. To ensure the complete removal of HF, veries ranged from 80-100% in TASGRAN, but only
two further 1 ml aliquots of HNO 3 were added and 55-80% in GSR-1 indicating differences in mineralogy.
evaporated to near dryness again. Finally, 2 ml HNO 3
and 5 to 10 ml ultra-pure water were added to the Ironstone FeR-2 gave a similar low average recovery
vessel. The vessel was replaced in the microwave (85.6, 85.8 and 87.8%) for the three microwave
210
GEOSTANDARDS
NEWSLETTER
The Journal of Geostandards and Geoanalysis
ICP-MS/reference
Digestion procedure
ICP-MS/reference
Results
211
GEOSTANDARDS
NEWSLETTER
The Journal of Geostandards and Geoanalysis
Reference values
ICP-MS/chondrite
Measured
212
GEOSTANDARDS
NEWSLETTER
The Journal of Geostandards and Geoanalysis
ICP-MS/chondrite
Reference values
The geological reference materials used in this Table 8 shows the recovery (in comparison with
study were GSR-1, G-2, AC-E, TASGRAN, FeR-2, FeR-4, reference values) of HFSE, Rb, Sr, Ba, Mo, Sb, Cs, Tl, Pb
BHVO-1, TAFAHI and TASBAS. The variation (accuracy) and Bi for the reference materials digested between
between measured results and reference values was two and four times each. Strontium and Th results
normally within ± 10%. Excluding data for Y, where show almost complete (100%) recovery, although the
reference values are high (Robinson et al. 1999), the blank levels of Sr (2-5 µg g -1) and Th (0.1 µg g -1) in
average bias for granites (AC-E and TASGRAN) is our Merck Na 2 O 2 preclude determination of these
-1.3% and +3.6% with a spread of 2 to 3%. Granites elements at low levels. Niobium had a reasonable
(GSR-1 and G-2) were about 6% low with a spread average recovery of 91% but the range of 76 to
of 3 to 5%. Basalts (BHVO-1, average bias -4.9% 100% is not consistent enough to be useful. Barium
and TASBAS, +0.1%) had a spread of 2 to 3% and and Bi also gave reasonable average recovery yields
the low level basalt (TAFAHI, +3.6%) had an expec- of 86% and 78% respectively, and again the recovery
ted wider spread of 6.9%, compared to the reference was variable (63-101%). Zirconium, Hf, Ta and Pb
values. Excluding data for Tm in FeR-4 (where the showed 69 to 80% average yields and were affected
reference value appears low), the ironstones FeR-2 by variable reagent (Na 2 O 2 ) contamination. Due to
and FeR-3 had an average bias of -3% and +2% solubility in a basic environment, Sn, Sb, Tl and U
respectively, although the spread was 6 to 7%, as gave only 48 to 66% average yields and Rb, Mo and
elements in these reference materials are low in Cs were almost completely lost (4-7% yields) during
concentration and not so well characterized. Precision the digestion.
(1 s ) for the REE, Sc and Y for the reference materials,
which were digested between three and six times, Compared to other digestion techniques, Na 2O 2
was generally better than 5% with the lowest average sinter is less time consuming for the determination of
for all elements being 1.8% for TASGRAN and the Sc, Y and REE in geological samples. A batch of
highest average 4.1% for G-2, except for TAFAHI samples may be digested within 4 to 5 hours.
(7.8%) which has lower trace element levels. Good Therefore, it is possible to obtain REE results in the
correlation (R 2 = 0.9979) between the measured same day by this digestion method in conjunction with
results and reference values can be seen in Figure 8. ICP-MS. However, Na 2O 2 sinter may not be suitable
213
GEOSTANDARDS
NEWSLETTER
The Journal of Geostandards and Geoanalysis
Table 8.
Recovery (% yield) for nine selected geological reference materials using the Na 2 O 2 sinter method
Missing data indicates elements either too low in concentration or preferred values unavailable. n = number of digestions.
Table 9.
Summary of elements measurable by different digestion methods
Basalt REE, Sc, Rb, Sr, Y, Zr, Nb, REE, Sc, Rb, Sr, Y, Zr, Nb, REE, Sc, Rb, (Sr), Y, Zr, Nb, REE, Sc, Rb, Sr, Y, Zr, Nb, REE, Sc, Sr, Y and Th
Mo, Sn, Sb, Cs, Ba, Hf, Mo, Sn, Sb, Cs, Ba, Hf, Mo, Sn, Sb, Cs, (Ba), Hf, Mo, Cs, Ba, Hf, Ta, Th
Ta, Tl, Pb, Bi, Th and U Ta, Tl, Pb, Bi, Th and U Ta, Tl, (Pb), Bi, Th and U and U
Granite - - REE, Sc, Rb, (Sr), Y, Zr, REE, Sc, Rb, Sr, Y, Zr, Nb, REE, Sc, Sr, Y and Th
(Nb), Mo, Sb, Cs, (Ba), Hf, Mo, Cs, Ba, Hf, Ta, Th
(Ta), Tl, (Pb), Bi, Th and U and U
Ironstone - REE, Sc, Rb, Sr, Y, Zr, Nb, REE, Sc, Rb, (Sr), Y, Zr, Nb, REE, Sc, Rb, Sr, Y, Zr, Nb, REE, Sc, Sr, Y and Th
Mo, Sn, Sb, Cs, Ba, Hf, Mo, Sn, Sb, Cs, (Ba), Hf, Mo, Cs, Ba, Hf, Ta,
Ta, Tl, Pb, Bi, Th and U Ta, Tl, (Pb), Bi, Th and U Th and U
Elements in brackets at high concentration cannot be measured due to the formation of insoluble sulfates or hydrolysis in solution. Some low
abundance ultramafic rocks such as dunites, peridotites and serpentines can also be decomposed by the HF/HClO 4 method (Yu et al. 2000).
for the determination of elements at low abundance in (1) Savillex Teflon vial acid digestion can be used
rock samples due to the lack of ultra-pure Na2O2. for the decomposition of basalts. The optimum ratio of
HF to HNO 3 for 100 mg powdered rock sample is 2
Conclusions ml to 0.5 ml. In order to achieve a complete dissolu-
tion, rock samples should be digested on a hotplate at
Results of five different sample digestion proce- about 130-150 °C for at least 48 hours. Granites can-
dures show that geological samples should be diges- not be completely decomposed by Savillex vial
ted using different digestion techniques according to digestion.
individual sample compositions. Table 9 gives a brief
summary of the measurable elements by Savillex , (2) Basalts and iron formation samples can be
PicoTrace , fusion and sinter digestion methods. The completely digested by the HF/HClO 4 high pressure
advantages and disadvantages of the five different digestion method at 180 °C for 48 hours. However,
digestion methods investigated in this study are sum- some low abundance ultramafic rocks such as dunites,
marized as follows: peridotites and serpentinites can be decomposed by
214
GEOSTANDARDS
NEWSLETTER
The Journal of Geostandards and Geoanalysis
HF/HClO 4 at about 180 °C for 16 hours (Yu et al. normally be decomposed and analysed in the same
2000). Generally, granites cannot be digested by the day. HFSE could not be measured (in this study) due
HF/HClO 4 high pressure digestion method unless a to variable yields and the high levels of HFSE in
complicated two-stage digestion procedure is applied. procedural blank solutions. Again, this method is not
The two-stage procedure can give recoveries of 85 to recommended for the measurement of elements at low
100% for more than thirty trace elements investigated abundance in samples because of the high blank
in granites. levels.
Basalts, granites (with accessory mineral phases This work also highlights the necessity of carefully
such as zircon, magnetite, fluorides and monazite) characterising decomposition procedures with respect
and ironstones can be completely decomposed by the to specific geological matrices of interest before any
3 ml HF/3 ml H 2SO 4 high pressure digestion method analysis is undertaken. It should be reiterated that no
at 180 °C for 16 hours. Granites containing cassiterite single decomposition protocol will allow the measure-
cannot be completely decomposed using this method. ment of all trace elements in all types of geological
In order to achieve complete decomposition, the samples by ICP-MS.
volumes of either HF or H2SO 4 for 100 mg powdered
samples should be controlled at 3 ml and any reduc- Acknowledgements
tion in the volumes of HF and H 2SO 4 may result in the
incomplete dissolution of rock samples. The main T h i s re s e a rc h p r o j e c t w a s s u p p o r t e d b y t h e
disadvantage of the HF/H 2SO 4 high pressure diges- Au stra lia n Re se a rch C o u ncil’ s Re se a rch C e ntres
tion method is that the evaporation time is very long, Programme. The Centre of Ore Deposit Research, the
taking, at present, 4 days to evaporate the mixture to School of Earth Sciences and the Central Science
dryness. Work is under way to improve the evapora- Laboratory (CSL), University of Tasmania, are gratefully
tion process by controlled heating of the evaporation acknowledged for the use of their digestion and ICP-
cover plate. MS facilities. Ashley Townsend of the CSL is particularly
thanked for his assistance with the ICP-MS analysis.
(3) Microwave digestion methods used during this Carsten Münker and Alfred Schneider are thanked for
study are not suitable for geological samples. The their assistance with the PicoTrace digestion apparatus.
recovery of thirty two trace elements in BHVO-1 (basalt) Ross Large and Tony Crawford at the Centre for Ore
was only 80-100% in comparison with reference values. Deposit Research, University of Tasmania, are thanked
Granites TASGRAN and GSR-1 gave average recoveries for their support.
of 69-84% and ironstone FeR-2 an average recovery
of 87%. References
(4) Li 2B4O7 fusion + HF/HNO 3 method can gene- Bankston D.C., Humphris S.E. and
rally be used to decompose many different geological Thompson G. (1979)
Major and minor oxide and trace element determination
samples such as basalts, granites and ironstones. in silicate rocks by direct current plasma optical emission
However, this method precludes the measurement of echelle spectrometry. Analytical Chemistry, 51, 1218-1225.
volatile elements such as Sn, Sb, Tl and Pb due to the
Boer R.H., Beukes G.J., Meyer F.M. and
high decomposition temperature (1000 °C). This Smith C.B. (1993)
method increases the level of total dissolved solid in Fluoride precipitates in silicate wet-chemistry: Implications
on REE fractionation. Chemical Geology, 104, 93-98.
solution and hence strongly degrades instrumental
intensity during an analytical run. In addition, this Borman S.A. (1988)
decomposition method may not be suitable for the Microwave dissolution. Analytical Chemistry, 60,
715A-716A.
decomposition of elements at low abundance in rock
samples, due to the lack of commercially available Brenner I.B., Watson A.E., Russell G.M. and
ultra-pure Li2B4O7. Goncalves M. (1980)
A new approach to the determination of the major and
minor constituents in silicate and phosphate rock.
(5) Finally, most geological samples can be com- Chemical Geology, 28, 321-330.
pletely decomposed by the sodium peroxide sinter
method. This method can provide fast analysis times
for Sc, Y and REEs, and geological samples can
215
GEOSTANDARDS
NEWSLETTER
The Journal of Geostandards and Geoanalysis
216
GEOSTANDARDS
NEWSLETTER
The Journal of Geostandards and Geoanalysis
McLaren J.W., Berman S.S., Boyke Y.J. and Totland M., Jarvis I. and Jarvis K.E. (1992)
Russell D.S. (1981) An assessment of dissolution techniques for the analysis
Simultaneous determination of major, minor, and trace of geological samples by plasma spectrometry. Chemical
elements in marine sediments by inductively coupled Geology, 95, 35-62.
plasma-atomic emission spectrometry. Analytical
Chemistry, 51, 1802-1806. Totland M., Jarvis I. and Jarvis K.E. (1995)
Microwave digestion and alkali fusion procedures for the
McQuaker N.R., Brown D.F. and determination of the platinum-group elements and gold
Kluckner P.D. (1979) in geological materials by ICP-MS. Chemical Geology,
Digestion of environmental materials for analysis of 124, 21-36.
inductively coupled plasma-atomic emission spectrometry.
Analytical Chemistry, 51, 1082-1084. Townsend A.T., Yu Z., McGoldrick P. and
Hutton J.A. (1998)
Moens L., Verrept P., Dams R., Greb, U., Jung G. Precise lead isotope ratios in Australian galena samples
and Laser B. (1994) by high resolution inductively coupled plasma-mass
New high-resolution inductively coupled plasma spectrometry. Journal of Analytical Atomic Spectrometry,
spectrometry technology applied for the determination of 13, 809-813.
V, Fe, Cu, Zn and Ag in human serum. Journal of
Analytical Atomic Spectrometry, 9, 1075-1078. Uchida H., Uchida T. and Iida C. (1980)
Determination of major and trace elements in silicate
Moselhy M.M., Boomer D.W., Bishop J.N., rocks by inductively coupled plasma-emission
Diosady P.l. and Howlett A.D. (1978) spectrometry. Analytica Chimica Acta, 116, 433-437.
Multielemental analysis of environmental materials by
inductively argon plasma excitation and direct-reading Walsh J.N. and Howie R.A. (1980)
spectrometry. Canadian Journal of Spectroscopy, 33, An evaluation of the performance of an inductively
186-195. coupled plasma source spectrometer for the determination
of the major and trace constituents of silicate rocks and
Münker C. (1998) minerals. Mineralogical Magazine, 43, 967-974.
Nb/Ta fractionation in a Cambrian arc/back arc system,
New Zealand, source constraints and application of Xie X., Yan M., Li L. and Shen H. (1985)
refined ICP-MS techniques. Chemical Geology, 144, Useable values for Chinese standard reference samples
23-45. of stream sediments, soils, and rocks, GSD 9-12, GSS 1-8
and GSR 1-6. Geostandards Newsletter, 9,
Nölter T., Maisenbacher P. and Puchelt H. (1990) 277-280.
Microwave and digestion of geological and biological
standard reference materials for trace element analysis Xie X., Yan M., Wang C., Li L. and Shen H. (1989)
by inductively coupled plasma-mass spectrometry. Geochemical standard reference samples GSD 9-12,
Spectroscopy, 5, 49-53. GSS 1-8 and GSR 1-6. Geostandards Newsletter, 13,
83-179.
Potts P.J. (1987)
A handbook of silicate rock analysis. Blackie (London), Yoshida S., Muramatsu Y., Tagami T. and
622pp. Uchida S. (1996)
Determination of major and trace elements in Japanese
Potts P.J. and Holbrook J.R. (1987) rock reference samples by ICP-MS. International Journal
Ailsa Craig - the history of a reference material. of Analytical Chemistry, 63, 195-206.
Geostandards Newsletter, 11, 257-260.
Yu Z., Robinson P., Townsend A.T., Münker C. and
Robinson P., Higgins N.C. and Jenner G.A. (1986) Crawford A.J. (2000)
Determination of rare-earth elements, yttrium and Determination of high field strength elements, Rb, Sr, Mo,
scandium in rocks by an ion exchange-X-ray fluorescence Sb, Cs, Tl and Bi at ng g-1 levels in geological reference
technique. Chemical Geology, 55, 121-137. materials by magnetic sector ICP-MS after HF/HClO4
high pressure digestion. Geostandards Newsletter: The
Journal of Geostandards and Geoanalysis, 24, 39-50.
217