Clays and Clay Minerals, VoL 31, No. 5, 397~100. 1983.

S I M U L T A N E O U S M U L T I E L E M E N T A N A L Y S I S O F C L A Y S BY I N D U C T I V E L Y
COUPLED PLASMA-ATOMIC EMISSION SPECTROSCOPY
USING SUSPENSION ASPIRATION

Key Words--Aspiration, Chemical analysis, Emission spectroscopy, Glacial till, Inductively coupled Ar
plasma, Plasma, Soil.

INTRODUCTION Alberta (Spiers, 1982). Subsamples (100 g) of soil were dis-

A variety of techniques, such as colorimetry, atomic ab- persed in distilled water for 3 min by ultrasonification at 400
sorption spectroscopy, electron microbeam analysis, X-ray W using a probe-type vibrator (Edwards and Bremner, 1967).
fluorescence, neutron activation analysis, and arc spectros- The <2-#m fraction was then separated from the suspensions
copy have been used for the chemical analysis of clay min- by a combination of gravity sedimentation and centrifugation
erals. The first two methods require dissolution of the samples techniques (McKeague, 1978). The separated clays were floc-
before analysis and are generally single-element techniques culated using 1 M CaC12 solution, washed of excess electrolyte
with significant chemical interferences present in the solution with distilled water, and freeze-dried for storage prior to sub-
matrices. The other techniques are applicable to solid phases, sequent analysis. The clay separates contained a mixture of
but require calibration standards of similar chemical com- smectites, micas, and kaolinite in approximately equal
position. Some require fairly extensive calculations to correct amounts, with lesser amounts of chlorite, and X-ray-amor-
for spectral and matrix interferences. A new analytical meth- phous aluminosilicates and iron oxides (Spiers, 1982).
od for elemental analysis of clays is inductively coupled plas- Digests of the clay separates were prepared using a modi-
ma-atomic emission spectroscopy (ICP-AES). This method fication of the HF-HCI dissolution method (Pawluk, 1967).
is suitable for most elements in the periodic table and has One-gram portions of each sample were ignited at 550~ in
low detection limits, high precision, and linear calibration porcelain crucibles for 8 hr. The samples were transferred to
over several orders of magnitude. ICP-AES enables simul- Teflon beakers and treated with 20. ml of concentrated HC1
taneous analysis of major, minor, and trace elements in a and 20 ml of concentrated HF. The mixture was evaporated
matter of seconds. to dryness on a sand bath at 85~ and the digestion procedure
The plasma excitation source is a relatively new advance repeated using 10 ml of each of the two acids. Finally, an
in emission spectrometry (Boumans, 1978). In an inductively additional 10 ml of HC1 only was added to the residue and
coupled argon plasma a stream of argon atoms is ionized by again evaporated to dryness to ensure complete volatilization
a Tesla coil. Then electrons which are passing through the of the HF. The residue was then dissolved in 25 ml of 4 N
radio-frequency (1.65 kW) induced magnetic field rotate in HC1 over the sand bath, cooled, transferred to a 50-ml vol-
an annular path. Collision between these electrons and the umetric flask and made to volume with distilled water.
argon gas causes further ionization and Ohmic heating. This Suspensions of the clays were prepared by dispersing known
argon plasma has a temperature of 6000-10,000~ in its cen- quantities of the freeze-dried samples in both distilled water
tral region. The analyte solution is then aspirated into a neb- and 2 N HC1 using a 15-sec treatment with the ultrasonic
ulizer where the flow of argon gas converts a proportion of probe. These suspensions were prepared over a concentration
the solution into an aerosol which is sprayed into the center range of 2% to 0.05% to evaluate the effect of colloidal con-
of the argon plasma. The analyte solution is thus introduced centration on the analyte nebulization rate. This was done
into a stable plasma in which the temperature is sufficient to because the most critical factor as far as analytical precision
dissociate chemical bonds and excite atoms. In such an en- is concerned with any technique requiring aspiration is to
vironment, the elements in the analyte solution emit their have the aspiration rate of both standards and analyte solu-
characteristic radiation which is focused into a conventional tions identical. The samples were agitated gently before and
air-path spectrometer where the spectral lines are diffracted during aspiration in order to maintain a uniform suspension.
from the main grating and detected by fixed photomultipliers Elemental abundances were determined with an ARL QA-
mounted along the Rowland circle of the spectrometer. 137 direct-reading spectrometer with an inductively coupled
The technique has found some usage in the earth sciences argon plasma-excitation source. Instrumental operating con-
for geochemical analyses (Burman, 1977; Floyd et al., 1980; ditions outlined in Table 1 are those recommended by the
Odegrad, 1979; Walsh and Howie, 1980), for the determi- manufacturer for obtaining optimal results for all analyzed
nation of the bulk composition of soils and sediments (Larson elements. Samples aspiration was natural with a - 5 cm head
et al., 1979; McQuacker et al., 1979; Spiers et al., 1983), and on the uptake capillary. Elemental concentration ranges of
for extractable nutrient determinations for a wide range of sets of combined liquid standards (Table 2) were chosen to
soils (Dalquist and Knoll, 1978; Jones, 1977; Soltanpour et bracket the elemental concentration ranges expected for the
aL, 1979). All of these applications, however, required that clay digests and suspensions. The elemental groupings for
samples be converted to solution form prior to analysis. standards were chosen to minimize possible interelernent ef-
Boumans (1978) indicated that ICP-AES, although pri- fects during instrument standardization. The data-processing
marily suited for the analysis of liquid samples, may be adapt- package supplied by ARL for use with the PDP 1104 mini-
ed for the direct analysis of solids. Solid or suspension analysis computer which is interfaced with the ICP-AES instrument
would eliminate the costly and time-consuming chemical dis- was used at all stages of this study.
solution for geological materials. The objective of the current
study was therefore to determine whether ICP-AES is, indeed, RESULTS AND DISCUSSION
suitable for direct analyses of clay suspensions without sample Preliminary experiments indicated that suspensions at the
dissolution. higher concentrations gave a 25% or greater reduction in as-
piration rate as compared to the dissolved samples. No ap-
MATERIALS AND METHODS parent difference was observed between suspensions in HC1
Clay separates (<2 #m) were obtained from 22 soils formed (2 N) made to match the standard matrix and those in distilled
on glacial till and lacustrine parent materials in northeastern water. On the basis of these findings suspensions containing

Copyright 9 1983, The Clay Minerals Society 397

398 Spiers, Dudas, and Hodgins Clays and Clay Minerals

Table 1. Operating procedures for routine analysis using the Table 3. Analysis of A.P.I. Garfield nontronite (H 33a) il-
Applied Research Laboratories QA-137 ICP-AES instru- lustrating agreement between analyses by electron microprobe
ment. and by suspension aspiration with inductively coupled plas-
ma-atomic emission spectroscopy.
Incident power 1650 W
Reflected power 10 W Electron ICP-AES l
Element microprobe (%)
Entrance slit width 12 ~m
Argon flow rates: Si 24.571 24.674 • 0.005
Coolant 10.5 liter/min A1 3.865 3.742 • 0.006
Sample carrier gas 1 liter/rain K 0.00 100 ppm
Plasma 1.5 liter/min Mg 0.398 0.381 • 0.003
Sample aspiration rate 2.4 ml/min Ca 1.358 1.365 • 0.004
Observation height 0.63 coil height Fe 26.153 26.217 • 0.002
Nebulizer Meinhart concentric
Operating temperature 24~ i Based on 10 analyses.

from unity probably arose from differences in viscosity of the

0.1-0.2 g of clay in 100 ml of distilled water were used in suspensions and solutions. The increased viscosity of the clay
subsequent analyses to produce minimal aspiration-rate ef- suspensions was shown to induce an aspiration rate 7-9%
fects while maintaining levels of several minor and trace ele- slower than that of both digest and standard solutions. Such
ments above instrumental detection limits. Clay suspensions a rate would cause a decrease in the number of atoms reaching
at this concentration gave a 7-9% decrease in aspiration rate. the plasma region of the instrument. This decrease in aspi-
The results of analyses of the clay suspensions were com- ration rate appears to be directly responsible for the analytical
pared with those obtained following routine chemical disso- variation observed and must be corrected by use o f a factor
lution procedures (Figure 1). The high coefficients of deter- of 1.08.
mination (r2) for all elements demonstrate clearly the potential The analyses described above indicate that it is possible to
of suspension aspiration for rapid and comparable analysis obtain data using suspension aspiration which are comparable
of the bulk chemistry of clay separates. This close agreement to those obtained by analysis of dissolved clay separates, but
in analytical results indicates that ICP-AES is able to disso- they do not indicate the degree of accuracy of the data. For
ciate the structural bonds of the clay species in the analyte this purpose five API reference clay minerals, previously ana-
suspensions to produce an atomic vapor in which the con- lyzed for all major elements using an electron microprobe
stituent atoms are excited to emit their characteristic spectra. technique (Smith and Cavell, 1980), were used to create clay
The b values (slopes) were not significantly different among suspension standards which would not show the viscosity
the 13 elements determined. All b values were between 1.07 differences noted for chemical standards. Suspensions of other
and 1.09 indicating higher recorded concentrations in the API clays for which microprobe analyses were available were
solutions than the suspensions. The deviation of the slopes then analyzed to determine the utility of the clay standard
curves for analysis of unknown clay separates. An example
of both the accuracy and reproducibility of these analyses is
shown in Table 3. This method of analysis has the added
Table 2. Spectral lines, detection limits, and standard groups advantage of providing accurate data for Si, an element which
used for instrumental calibration. cannot be determined by the dissolution method used in this
study as it is volatilized as silicon tetrafluoride.
Detection Standard This study shows that ICP-AES is capable of decomposing
Spectral line limi0 Group concentration
Element (nm) (ppm) number range (ppm) crystalline clay materials to provide accurate elemental anal-
yses of clay suspensions. Furthermore, matrix effects resulting
Cu 3247.5 1.0 1 0-10 from viscosity differences between chemical standards and
Co 2388.9 1.0 1 0-10 sample clay suspensions can be eliminated by using reference
B 2497.7 0.5 1 0-10 clay minerals as suspension standards. For routine analyses
Cr 2677.2 0.5 1 0-10 of composition of clay separates, the main advantages of this
As 1937.7 5.0 1 0-10 technique are its simplicity, speed, and accuracy. An operator
Pb 2203.5 4.0 1 0-10 can analyze 20-25 samples per hour for a wide range of ele-
Cd 2265.0 0.5 1 0-10 ments with no tedious sample digestion being required.
Zn 2025.5 0.1 2 0-100 The technique described here indicates that it is possible
Mn 2576.1 0.1 2 0-I 00 to sample a clay suspension of unknown composition, gravi-
Mo 2020.3 0.1 2 0-100 metrically determine its concentration, and proceed with
V 3102.3 0.5 2 0-100 analysis using ICP-AES following dilution of a subsample to
Sr 2136.2 0.1 2 0-100 the required concentration. Saturation of subsamples of the
P 4077.7 4.5 2 0-100 original suspensions with different index cations and anions
Ca 3158.9 2.5 3 0-1000 would enable the calculation of both anion- and cation-ex-
Na 5895.9 5.0 3 0-1000 change capacities of the separates, with the added advantage
Mg 2790.8 3.0 4 0-1000 of producing replicate analyses for all other compositional
K 7664.9 8.0 4 0-1000 elements. The remaining portion of these suspensions could
Fe 2599.4 1.5 5 0-2000 then be filtered onto ceramic disks or smeared onto glass slides
A1 3082.2 5.0 6 0-2000 for X-ray powder diffraction analysis for determination of the
mineralogical composition. In essence, a rather complete
Extrapolated to indicate minimum concentrations in clay chemical and mineralogical characterization could be ob-
materials using dissolution technique described in this paper. tained on a single suspension sample.
Vol. 31, No. 5, 1983 Clay analyses by atomic emission spectroscopy 399

~176

11

I i i I I I i I
' (~ ' I'I ' I'3 5 ; -~ 8
AI~ Fe % K %

1.0 200C 300

b : 1.07
I I i i I i
o'o'.5'1'.o'1'.5 0 I000 2000 3000 0 150 3 0 4 0

Mg % N o ,ug/g B ,ug/g

I-..
200 "/ 200 1200

b : 1.09 B -- 1.08 b : 1.08

"',4
i i i i i I i i I i l
100 200 300 100 200 300 0 ' '
600 ' 1~oo' 1~oo

Cr ~ag/g Mn pg/g P ,ug/g

200 400 80

b : 1.07
i i 1 i i i i i ,
o ,oo 2;0 3;0 0 200 400 600 ' 4'0 . . 80. . 120

Sr pg/g V ,ug/g Zn jug/g

SUSPENSION ASPIRA T/ON

Figure 1. Comparative plots showing total elemental abundances obtained by direct aspiration of suspensions and by solution
analysis of dissolved clay separates.
400 Spiers, Dudas, and Hodgins Clays and Clay Minerals

Department o f Soil Science G . A . SPIERS McKeague, J. A., ed. (1978) Manual on Soil Sampling and
University o f Alberta M . J . DUDAS Methods o f the Canada Soil Survey Committee: Can. Soc.
Edmonton, Alberta T6G 2E3 Soil Sci., 212 pp.
Canada McQuaker, N. R., Klucker, P. D., and Gok, N. G. (1979)
Calibration of an inductively coupled plasma emission
Alberta Soil and Feed Testing Laboratory L.W. HODGINS
spectrometer for the analysis of environmental materials:
Alberta Agriculture
AnaL Chem. 51, 888-895.
Edmonton, Alberta T6G 2E3
Odegard, M. (1979) Determination of major elements in
Canada
geologic materials by ICP-AES: Jarrell-Ash Plasma News-
REFERENCES letter 2, 4-7.
Pawluk, S. (1967) Soil analyses by atomic absorption spec-
Boumans, P. W. J. M. (1978) Inductively coupled plasma-
trometry: Atomic Abs. Newsletter 6, 53-56.
atomic emission spectroscopy: its present and future po-
Smith, D. G. W. and Cavell, P. A. (1980) Electron micro-
sition in analytical chemistry: Optica Pure Y Appicada 11,
probe analysis of clay minerals: in Electron Probe Micro-
143-171.
analysis in Mineralogy, Proc. 1 lth General Meeting Int.
Burman, J. (1977) Analysis ofphosphorus in geological sam-
Min. Assoc., Novosibirsk, September 1978, 90-100.
pies: ICP Information Newsletter 3, 33-36.
Soltanpour, P. N., Workman, S. M., and Schwab, A.P. (1979)
Dalquist, R. L. and Knoll, J. W. (1978) Inductively coupled
Use of inductively coupled plasma spectrometry for the
plasma-atomic emission spectrometry: analysis of biolog-
simultaneous determination of macro and micronutrients
ical materials and soils for major, trace and ultra-trace ele-
in NH4HCO3-DTPA extracts of soils: Soil Soc. Amer. J.
ments: Appl. Spectrosc. 32, 1-29.
43, 75-78.
Edwards, A. P. and Bremner, J. M. (1967) Dispersion of
soil particles by sonic vibration: J. Soil Sci. 18, 47-63. Spiers, G.A. (1982) Mineralogy and geochemistry of parent
materials of the Athabasca Tar Sands Region: M.Sc. thesis,
Floyd, M. A., Fassel, V. A., and D'Silva, A. P. (1980) Com-
University of Alberta, Edmonton, Alberta, 256 pp.
puter-controlled scanning monochromator for the deter-
Spiers, G. A., Dudas, M. J., :and Hodgins, L. W. (1983)
mination of 50 elements in geochemical and, environmental
Instrumental conditions and procedure for multielement
samples by inductively coupled plasma-atomic emission
analysis of soils and plant tissue by ICP-AES: Comm. Soil
spectrometry: Anal, Chem. 52, 2168-2173.
Sci. Plant AnaL 14, (in press).
Jones, J. B. (1977) Elemental analysis of soil extracts and
Walsh, J. N. and Howie, R.A. (1980) An evaluation of the
plant tissue ash by plasma emission spectroscopy: Comm.
performance of an inductively coupled plasma source spec-
Soil Sci. Plant Anal. 8, 349-365. trometer for trace constituents of silicate rocks and min-
Larson, G. F., Morrow, R. W., and White, L. E. (1979) erals. Min. Mag. 43, 967-974.
Application of an inductively coupled plasma/direct read-
ing polychromator to the multielement analysis of stream
sediment extracts: 1CP Inform. Newsletter 4, p. 461. (Received 17 January 1983; accepted 12 March 1983)