Professional Documents
Culture Documents
Acid Digestion of Geological and Environmental Samples Using Open Vessel Focused Microwave Digestion
Acid Digestion of Geological and Environmental Samples Using Open Vessel Focused Microwave Digestion
Acid Digestion of Geological and Environmental Samples Using Open Vessel Focused Microwave Digestion
DOI 10.1007/s00216-001-1172-z
O R I G I N A L PA P E R
Received: 10 July 2001 / Revised: 10 September 2001 / Accepted: 8 October 2001 / Published online: 18 December 2001
© Springer-Verlag 2001
Abstract The application of open vessel focused mi- grounds in ICP-MS analysis, compared to fusion methods
crowave acid digestion is described for the preparation of also commonly used for solid samples. When using acid
geological and environmental samples for analysis using digestion, the resulting solution has a lower content of to-
inductively coupled plasma-mass spectrometry (ICP-MS). tal dissolved solids compared to methods using fluxing
The method is compared to conventional closed-vessel agents [1]. High purity acids are also more economical
high pressure methods which are limited in the use of HF and easier to obtain than high purity fluxing agents.
to break down silicates. Open-vessel acid digestion more Open-vessel focused microwave digestion signifi-
conveniently enables the use of HF to remove Si from cantly increases the efficiency of acid digestion and has
geological and plant samples as volatile SiF4, as well as overcome several limitations of conventional closed-ves-
evaporation-to-dryness and sequential acid addition dur- sel microwaves. Open-vessel microwave digestions of or-
ing the procedure. Rock reference materials (G-2 granite, ganic and carbonate samples, which produce large vol-
MRG-1 gabbros, SY-2 syenite, JA-1 andesite, and JB-2 umes of gases on digestion, can be carried out avoiding
and SRM-688 basalts) and plant reference materials (BCR the safety concerns of forming high pressure bombs and
and IAEA lichens, peach leaves, apple leaves, Durham sample loss caused by effervescence when opening closed-
wheat flour, and pine needles) were digested with results vessels. Large samples of organic material can be digested
comparable to conventional hotplate digestion. The mi- using the open-vessel microwave, minimizing the time-
crowave digestion method gave poor results for granitic consuming steps of drying and grinding [2]. Automated
samples containing refractory minerals, however fusion evaporation-to-dryness is not possible with closed-vessel
was the preferred method of preparation for these sam- systems, but can be accomplished in an open-vessel by at-
ples. Sample preparation time was reduced from several taching a pump to evacuate the fumes while the container
days, using conventional hotplate digestion method, to is heated.
one hour per sample using our microwave method. Acid addition can also be automated with the open-
vessel microwave, allowing the use of procedures with
Keywords Digestion · Geological samples · multiple heating and evaporation steps. Recent studies
Environmental samples · Microwave digestion · have used the open-vessel microwaves for plant digestion,
Open-vessel eliminating the need for ashing prior to digestion [3], and
enables the use of strong oxidant acids with HF [4]. A fo-
Introduction cused microwave has also been used for online HCl di-
gestion of slurries of oxide materials [5]. Open-vessel
Acid digestion of solid geological and environmental mineralization of industrial waste and sediment certified
samples is a useful technique for achieving low back- materials was compared to alkaline fusion methods and
high pressure digestion for the determination of 10 ele-
ments by inductively coupled plasma-optical emission
Electronic supplementary material to this paper can be obtained spectrometry (ICP-OES) and each method was shown to
by using the Springer Link server located at http://dx.doi.org/
10.1007/s00216-001-1172-z have advantages and disadvantages depending on the
sample matrix [6].
V.F. Taylor · A. Toms · H.P. Longerich (✉) Conventional hotplate acid digestions are widely used
Department of Earth Sciences in geoanalysis but they nevertheless remain a major limit-
and Centre for Earth Resources Research,
Memorial University of Newfoundland,
ing factor of analytical efficiency. Microwave digestion of
St. John’s, NF A1B 3X5, Canada geological materials has not been widely reported, as
e-mail: henry@sparky2.esd.mun.ca closed-vessel procedures limit the use of HF, which is
361
used to break down silicates. Open-vessel digestion of ge- unit, was used. A vacuum pump (KNF Neuberger Inc., Model UN
ological samples allows the use of HF to remove the Si as 726.1.2FTP) was used to aid in the removal of volatile compo-
nents. Microwave sample vessels were machined in-house from
volatile SiF4, further lowering the total dissolved solids in virgin Teflon PTFE. Hotplate vessels were Savillex 15 mL Teflon
the final digest and thus minimizing matrix and memory screw cap jars. Microwave and hotplate vessels were cleaned by
effects in ICP-MS instrumentation. The removal of HF heating for 2 days in 4 mol L–1 HNO3, then 2 days in 4 mol L–1
also eliminates the need for sample introduction systems HCl, and were finally rinsed and heated in water for 12 h prior to
being dried in a flow of HEPA filtered air. Samples were weighed
that are HF resistant, or the addition of large quantities of into vessels and placed into the cavity of the microwave module,
boric acid to complex the fluoride. Digestion using HF where the lower few cm of the vessel were exposed to the micro-
can also be necessary in biological samples to break down waves, while the upper part of the vessel remained outside the mi-
siliceous structures in plants [4]. crowave cavity at ambient temperature. Microwave power could
Acid digestion of geologic samples is very successful be regulated from 20 to 200 W, with energy delivered to the sam-
ple solution through a wave guide. A condenser top was screwed
when applied to fine grained mafic samples, but digestion on to the vessel and decreased loss through evaporation; it con-
is often incomplete when applied to coarse grained granitic tained a small opening through which acids could be added or the
rocks, especially those containing significant quantities of vessel could be evacuated during drying steps [10]. During opera-
zircons (especially zircons which are large grained, young, tion, the vessel, with condenser top attached, was placed in the
cavity of the microwave module and an evaporator head was
and contain low concentrations of U). This report com- moved into place above the vessel. For refluxing steps, the evapo-
pares an open-vessel focused microwave acid digestion rator head sat about 2 cm above the top of the vessel, and for evap-
with a conventional hotplate digestion of six rock refer- oration it was manually lowered and fitted into the top of the con-
ence materials (G-2 granite, MRG-1 gabbro, SY-2 syen- denser, where it formed a seal. This evaporator head was attached
by a Teflon tube to the vacuum pump unit, which is a two piston
ite, JA-1 andesite, JB-2 basalt and NIST SRM-688 pump which rapidly draws air in through the tube attached to the
basalt), six plant reference materials (BCR CRM-482 evaporator head, and out through an exhaust tube which feeds into
lichen, IAEA-336 lichen, NIST SRM-1547 peach leaves, a fume hood. The pump lining and the tubes attached to it were
NIST SRM-1515 apple leaves, NIST SRM-8436 Durham made from Teflon, which has a high resistance to digestion fumes
wheat flour, and NIST SRM-1575 pine needles), and two containing HF. When the evaporator head was in place, samples
could be evaporated-to-dryness during heating. A programmable
“difficult to digest” granodiorite samples. Due to the pos- control unit was attached to the microwave module to operate the
sible occurrence of incomplete digestion for some sam- heating stages of the digestion procedure. Programs are stored as a
ples, it was important to monitor the completeness of di- series of steps in which the microwave power and time were spec-
gestion. For the granodiorite samples, the results obtained ified. With the addition of a 3 pump dispenser, the program could
also control the addition of reagents. The system could be com-
by ICP-MS using digestion methods were compared with pletely automated by the addition of a robotic arm to raise and
those obtained using an XRF solid sample pressed-pellet lower the condenser head between refluxing and condensing steps.
method [7], especially for the elements Rb, Sr, Y, Zr, and A sixteen vessel autosampler is also available for the microwave
Nb. unit which would allow multiple samples to be run unattended.
The procedure for open-vessel microwave digestion is
closely modeled on a current hotplate digestion method
Procedure
used in the Department of Earth Sciences at Memorial
University, but reduces the digestion time from days to an Both digestion methods used were based reported procedures [11],
hour. There is potential to automate the entire microwave except that quantities of both sample and acid were doubled in the
procedure after the samples are weighed, including acid microwave procedure (keeping sample to acid proportions the
addition and evaporation-to-dryness, and a 16 flask au- same), because of the large sample chamber where the microwaves
tosampler could be attached and programmed [8]. are focused at the bottom of the vessel. In the first step of the di-
gestion, the oxidizing acid HNO3 and HF were used. A subsequent
evaporation with H3BO3 and HNO3 then complexes and removes
any residual HF. The addition of H3BO3 aids in the prevention of
Experimental the formation of insoluble fluorides, especially with Mg and Al,
which can otherwise remove trace elements from the solution.
Reagents and labware Two further evaporations with HNO3 complete the oxidation of the
elements in solution and dissolve most metallic elements as solu-
All water was distilled and deionised (Barnstead NanoPure II). ble nitrates [10]. In the final step, the digested sample is dissolved
Concentrated nitric acid (Fisher Brand ACS.) was diluted to with HNO3, then the HF-H3BO3 solution and oxalic acid solution
8 mol L–1 and distilled in a Teflon two-bottle sub-boiling still are added, and the sample solution is diluted by weight with water.
(Berhof America). Concentrated hydrofluoric acid (FisherBrand The low concentration of HF in the solution complexes elements
ACS 49% v/v) was also distilled in a Teflon two-bottle still [9]. as their stable fluorides, and boric acid then serves to complex ex-
Oxalic (BDH ACS) and boric (Anachemia ACS) acids were ob- cess residual fluorine, producing a solution buffered in fluoride.
tained in powdered crystalline forms and were used without further Oxalates form stable soluble complexes especially with Fe, which
purification. Three solutions were prepared: 1) 0.22 mol L–1 oxalic at high concentrations can become insoluble in solution unless HCl
acid, 2) 0.45 mol L–1 boric acid, 3) 0.45 mol L–1 boric acid solution is added.
and 0.11 mol L–1HF. Digested samples were transferred to Corning Development of the microwave method was closely modeled
Snap Seal No.1730 vials for final dilution. on the hotplate method, keeping the same reflux/evaporation cy-
cles. Refluxing was achieved at 10% delivered power (20 W) and
evaporation was achieved at 65% power (130 W), while the use of
Microwave system higher power was found to cause damage to the microwave vessel.
In the development of a method for open-vessel digestion of waste
A PROLABO MicroDigest 301 system (Rhone-Poulenc, France), samples [11], it was reported that the heating power and heating
consisting of a microwave heating/digesting unit and a control time were found to have an insignificant effect on Hg recovery,
362
whereas procedures with multiple heating steps were found to im- of 8 mol L–1 HNO3, and 2 mL of 0.45 mol L–1 H3BO3 were added
prove Hg recovery compared to a single step procedure. to the sample vessel, and the solution was evaporated at 65%
power for 5 min. To the vessel, 4 mL of 8 mol L–1 HNO3 was
added, and the solution was evaporated for 5 min at 65% power.
Hotplate procedure [12] This step was repeated once, then 4 mL of 8 mol L–1 HNO3 was
added and the solution was refluxed at 10% power for 4 min, to
Approximately 0.1 g of each sample was weighed into a clean dissolve all the sample residues. The sample solution was trans-
Teflon screw cap jar, and 2 mL of 8 mol L–1 HNO3 and 1 mL of ferred to a clean, weighed, labeled vial, and 5.4 mL of 0.22 mol L–1
concentrated HF were added to each jar. The jars were then cov- oxalic acid and 2.67 mL of the boric/HF solution were added. The
ered and heated overnight on a hotplate at 80–100 °C. If samples solution was then made up to a final weight of 120 g with water.
were not dissolved, they were heated for an additional day. The jar
covers were then removed and rinsed into the jar with 8 mol L–1
HNO3, and the solutions were then evaporated to dryness at Inductively coupled plasma-mass spectrometer (ICP-MS)
80–100 °C. Jars were cooled and another 2 mL of 8 mol L–1 HNO3
and 1 mL of HF were added to each jar, and the solutions were All the ICP-MS data acquired for this study was obtained using a
then heated for 4–5 days, then evaporated to dryness as before. To Hewlett-Packard 4500 Series inductively coupled plasma mass
each jar, 2 mL of 8 mol L–1 HNO3 and 1 mL of 0.45 mol L–1 boric spectrometer.
acid were added, and the solutions were evaporated to dryness on
the hotplate. Two more cycles of adding 2 mL of 8 mol L–1 HNO3
followed by evaporation-to-dryness were completed. Finally, 2 mL Trace procedure
of 8 mol L–1 HNO3 were added to each sample and the jars were
covered and warmed until all the residue had completely dis- Samples were analyzed by two routine methods in use at the
solved. Samples were then transferred to clean, weighed, Corning Memorial University. The “Trace” package determined the iso-
Snap Seal No.1730 vials, and jars and covers were rinsed into the topes 7Li, 85Rb, 86Sr, 89Y, 90Zr, 93Nb, 133Cs, 137Ba, 139La, 140Ce,
vials with water. To each vial, 1.3 mL of 0.22 mol L–1 oxalic acid, 141Pr, 145Nd, 147Sm, 151Eu, 160Gd, 159Tb, 163Dy, 165Ho, 167Er, 169Tm,
0.67 mL of 0.45 mol L–1 boric/0.11 mol L–1 HF solution, and wa- 173Yb, 175Lu, 177Hf, 181Ta, 205Tl, 206Pb, 207Pb, 208Pb, 209Bi, 232Th, and
ter were added to afford solutions with a final weight of approxi- 238U. The Trace package uses a combination of external calibra-
mately 60 g. tion, surrogate calibration, standard addition, and an internal stan-
dard for data reduction. The standard addition spike contained Li,
Sr, Zr, Cs, Ba, La, Pr, Nd, Sm, Er, Tm, Tl, Bi, and U. Into each
Microwave procedure “spiked” tube was weighed 0.5 g of sample solution, 5.0 g of spike
and 4.5 g of 0.2 mol L–1 HNO3. Into each “unspiked” tube was
Approximately 0.2 g of powdered sample was weighed into a di- weighed 0.5 g sample solution and 9.5 g of 0.2 mol L–1 HNO3. An
gestion vessel, a condenser head was then attached and the vessel internal standard solution of indium (In) was mixed on line using
was placed in the digester (Fig. 1). During heating stages, the up- the sample peristaltic pump using flow rates of 0.3 g min–1 sample
per portion of the flask, which was outside of the actual digester, and 0.015 g min–1 In spike. This produced a solution which
remained at ambient temperature, so that it acted as a reflux con- reached the nebulizer having a concentration of approximately
denser to minimize sample loss. The heating and evaporation steps 0.08 g sample per kg solution. The samples and standards were
of the procedure were automated, and acids were added manually first background-corrected by subtracting the mean of all calibra-
by pipette, although a pump dispensing system is available which tion blanks in the run (0.2 mol L–1 HNO3). Interference corrections
can be attached to the microwave unit and also be programmed to for the oxides of Ba, Pr, Nd, Sm, Eu, Gd, Tb, Dy, and Ho were ap-
dispense specified volumes of up to three different solutions. In plied to the analytes Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu, and Hf.
this work, the suction head was lowered manually for the evapora- The elemental ion interference of 160Dy on 160Gd was also cor-
tion steps and then raised for refluxing. To more fully automate the rected. Each of these interference corrections were determined
digestion process a robotic arm could to be added to the system to from the external standard solutions used and then normalized to
raise and lower the suction head. the UO+/U+ ratio in the calibration standards relative to that of the
To the sample vessel were added 4 mL of 8 mol L–1 HNO3 and spiked samples. Two external standard solutions were used for the
2 mL of concentrated HF, and the solution was refluxed for 15 min calibration of all elements, with the exception of Nb and Ta, which
at 10% microwave power (20 W), with the evaporator head raised. were surrogated by Zr or Hf. A matrix correction factor was calcu-
The evaporator head was then lowered, and the solution was lated from the ratio of the sensitivity determined by external cali-
heated for 5 min at 65% power (130 W). The acid addition, reflux- bration to the sensitivity determined by using internal standardiza-
ing, and evaporation steps were then repeated once. Another 4 mL tion from each element in the spike. For elements not contained in
the spike, the matrix was corrected using mass interpolation of the
determined matrix factors. Drift was corrected using the signal
from the In internal standard.
Environmental/exploration procedure