Full text of DOI:10.1557/JMR.2000.0182

90
Synthetic clay excels in Sr removal
Sridhar Komarneni, Tatsuya Kodama, and William J. Paulus
Materials Research Laboratory and Department of Agronomy, The Pennsylvania State University,
University Park, Pennsylvania 16802
C. Carlson
Process Technology and Environmental Management Resources, Environmental Technology Division,
Pacific Northwest National Laboratory, Richland, Washington 99352
(Received 20 August 1999; accepted 20 March 2000)
Tests with actual ground water from Hanford site, and fundamental studies of
2Na+ → Sr2+ exchange equilibria revealed that a synthetic clay is extremely selective
for 90Sr with a high capacity for uptake. Comparative studies with existing Sr
selective ion exchangers clearly revealed that the present synthetic clay exhibited
the best performance for 90Sr removal from actual ground water collected from
three different locations at Hanford. This novel Sr ion sieve is expected to be
useful for the decontamination of the environment after accidental release and
contamination with 90Sr.
Naturally occurring cation exchangers such as clays
and zeolites have been used to decontaminate and dis-
pose of the radioactive species.1–4 There has been a great
deal of effort to develop high performance synthetic cat-
ion exchangers for the uptake of Sr and Cs5–10 and their
immobilization. A need exists for high performance cat-
ion exchangers to separate Sr, Cs, and other species from
high-level alkaline tank wastes and remediation of proc-
ess and ground water at the Hanford nuclear site, the
latter to prevent contamination of the Columbia River.11
Here we report the discovery of a new synthetic clay,
Na2Si6Al2Mg6O20F4 и xH2O (nominal composition),
which excels in 90Sr removal from contaminated ground
waters. This phase will also be useful in 90Sr immobili-
zation in the interlayers by modest heating after the Sr
ion uptake.
The starting precursor for the synthesis method was
derived by a sol-gel process and is briefly described here.
A single phase or monophasic gel was prepared by dis-
solving Mg(NO3)2 и 6H2O and Al(NO3)3 и 9H2O sepa-
rately in absolute ethanol and then combining and mixing
the two solutions by stirring for 1 h before adding tetra-
ethoxysilane, Si(OC2H5)4. The stoichiometric composi-
tion of the three components on an oxide basis was
6MgO–Al2O3–6SiO2. The above mixed sol of the three
components was stirred for three hours to achieve homo-
geneity. The container was then covered with a plastic
film and placed in an oven at 60 °C to form a gel through
hydrolysis, condensation, and polymerization reactions
for three days. This monophasic gel was then dried in an
oven at 100 °C, calcined at 475 °C for 12 h, and ground
to a powder to pass through a −325 mesh screen. This gel
powder was then mixed with an equal weight of
−325 mesh NaF crystalline powder in a platinum crucible
and sintered for 18 h at 890 °C in a programmed furnace.
These sintering conditions were found to be ideal for
synthesis of this type of fluorinated clay as had been
reported previously.9 After cooling the crucible to room
temperature, the reaction products were repeatedly
washed with de-ionized water to remove all the soluble
components including NaF. The reaction products were
then washed with saturated boric acid to remove any
impurities of water insoluble fluorides. Finally, the
sample was washed with NaCl solution to completely
saturate all the exchange sites with Na, and washed with
deionized water to remove excess Na before gently dry-
ing to obtain a dry powder.
This powder was characterized by x-ray diffraction
(XRD) that revealed a phase pure swelling mica (figure not
shown) with an analyzed composition close to the expected
ideal composition of Na2Si6Al2Mg6O20F4 и xH2O.
This phase is hereafter referred to as Na-2-mica because
it contains two Na ions per unit cell. The Na-2-mica was
further characterized by scanning electron microscopy
(SEM) that revealed flakes of about 2–3 ␮m. 27Al and
29Si magic angle spinning nuclear magnetic resonance
(MASNMR) spectroscopy was used to determine the co-
ordination and nearest neighbor environments of Al and
Si. The 27Al results revealed that almost all Al was pres-
ent in the tetrahedral coordination, as expected. The 29Si
MASNMR revealed Q3 resonances of Si (3Al) at
−78.8 ppm, Si (2Al) at −82.7 ppm, Si (1Al) at
−86.7 ppm, and Si (0Al) at −93.6 ppm, and these indicate
that the negative charge is distributed nonuniformly
1254
J. Mater. Res., Vol. 15, No. 6, Jun 2000
© 2000 Materials Research Society
http://journals.cambridge.org
Downloaded: 18 Mar 2015
IP address: 138.251.14.35

Rapid Communications
on the layer. The approximate intensities of Si (3Al),
The dotted line indicates that the corrected selectivity
coefficient is equal to unity. Any points that fall above
this line indicate selectivity for Sr2+ uptake. Only one
point at high Sr concentration falls on the dotted line
indicating little or no preference. From the equilibrium
study using 0.00468N (Fig. 2), total Sr2+ exchange ca-
pacity was determined to be 231 meq/100 g, which is
Si (2Al), Si (1Al), and Si (0Al) are 21%, 30%, 32%, and
17%, respectively, which is what is expected based on
the chemical composition.
The Na-2-mica (hydrated form) was used in all the
following Sr2+ exchange experiments. A Sr2+ exchange
isotherm was determined using 25 mg (anhydrous basis)
of the Na-2-mica, and equilibrating while shaking with
25 cm3 of mixed solutions having different mol ratios of
Na+/Sr2+ at 25 °C for 4 weeks. Two different total nor-
malities (Sr + Na) of 0.00247 and 0.00468 were used in
the equilibria experiments. After equilibration, the solid
and solution phases were separated by centrifugation and
the solutions were analyzed for Sr2+ and Na+ by atomic
emission spectroscopy and also inductively coupled
plasma spectroscopy for some Sr2+ analyses. The
amounts of Sr2+ ions exchanged and Na+ ions released by
Na-2-mica were determined from the difference in the
concentration between the equilibrated solution and the
starting reference. The theoretical cation exchange ca-
pacity [247 meq/100 g on anhydrous basis] was used for
plotting the ion exchange isotherms. Selective strontium
uptake in the presence of a large excess concentration
of Na+ was also determined by equilibrating 25 mg of
Na-2-mica sample for 24 h in 25.0 cm3 of 0.50N NaCl
solution containing 0.0002N SrCl2. We also used a
0.50N NaCl solution containing 0.0004N SrCl2. These
two solutions have [Na+]/[Sr2+] equivalent ratios of
2500:1 and 1250:1, respectively. The solid and solution
phases were separated by centrifugation and solutions
were analyzed as described above. The selective uptake
of Sr from 0.5N NaCl is expressed as distribution coef-
ficient, Kd which is defined as the ratio of the amount of
strontium sorbed per gram of solid to the amount of
strontium remaining per cm3 of solution. All of the data
points in the isotherms and Kd values represent the av-
erage of triplicate measurements. The three measure-
ments for each data point agreed within 5%.
FIG. 1. Selective ion exchange isotherms for Na-2-mica with two
different total normalities of NaCl/SrCl2 solutions.
Figure 1 shows the exchange isotherms of sodium/
strontium equilibria with two different total normalities.
From these isotherms, one can see that almost all Sr2+
was taken up from solution until an equivalent fraction in
the solid of 0.6 to 0.7 was reached. These isotherms
clearly show that the Na-2-mica phase is extremely se-
lective for Sr2+ in the presence of Na+ at low concentra-
tions of Sr. These isotherms are steeper than those
reported previously for Na-4-mica,9 a different synthetic
clay, suggesting that this Na-2-mica phase is superior to
Na-4-mica in Sr2+ uptake. Figure 2 shows the Kielland
plot which is calculated from equilibrium data with
0.00468 total normality using standard procedures.12
This plot is linear within experimental errors and sug-
gests that the strontium exchange proceeds mainly on
one kind of site in Na-2-mica, i.e., the interlayer Na+ site.
At xSr < 0.9, the Kielland plots fall above the dotted line.
FIG. 2. Kielland plot for 2Na+ → Sr2+ exchange on Na-2-mica.
J. Mater. Res., Vol. 15, No. 6, Jun 2000
Downloaded: 18 Mar 2015
1255
http://journals.cambridge.org
IP address: 138.251.14.35

Rapid Communications
TABLE I. Strontium removal data from groundwater sample 199-N-106Aa (high strontium-90 concentration).
Total Sr
Ion-exchange
material
Liquid-solid
ratio
Kd
(ml/g)
loading
(mmol/g)
90Sr loading
(mmol/g)
DFb
4.68
IONSIV IE-
911, UOP
Sr-Treat,
Selion, Inc.
PSU Na-2-Mica
4000
4000
4000
1.62 × 104
1.13 × 104
7.32 × 104
6.56 × 10−3
5.83 × 10−3
7.62 × 10−3
9.91 × 10−10
8.80 × 10−10
1.15 × 10−9
3.63
1.89 × 101
aThe composition of 199-N-106A ground water is as follows: Li, 3; Be, <0.5; Na, 6350; Mg, 6800; K, 1960; Ca, 39400; Cr, 2.76; Mn, <0.5; Rb, 1.28; Sr,
183; Cs, <0.5; Ba, 23.2, and 90Sr., 3.92. All the values are in ng/ml except 90Sr, which is in units of pCi/ml.
bDF ס Decontamination factor is the ratio of the initial 90Sr activity in solution to that of the final activity in solution.
higher than that reported previously for Na-4-mica.9 The
extremely high preference for strontium over sodium was
also indicated by the Kd measurements, which showed
distribution coefficients of 6484 ± 458 (ml/g) and
2020 ± 134 (ml/g) with 0.0002N and 0.0004N SrCl2 con-
centrations, respectively, in the presence of 0.5N NaCl.
Further tests of this material were conducted with ac-
tual N-area ground water containing 90Sr from the Han-
ford site. Water from three different wells were treated
and compared with the performance of the best available
ion exchangers. This Na-2-mica performed better than all
the exchangers tested with the water from the three dif-
ferent wells. For example, Table I lists the distribution
coefficients, loading and decontamination factors using
ground water sample 199-N-106 A with high 90Sr con-
centration. The results clearly show that the Penn State
University (PSU) Na-2-mica gave the best 90Sr removal
as determined by Kd, highest 90Sr loading, and highest
decontamination factor (Table I). The high selectivity of
the PSU Na-2-mica phase can be explained based on the
high charge density of the layers, narrow interlayer spac-
ing and low hydration energy of Sr2+ ions. The high
charge density leads to easy dehydration of the less hy-
drated Sr2+ ions compared to hydrated Mg, Ca, Na, etc.,
ions in the ground water, and thus the narrow interlayer
spacing is ideal for the uptake of dehydrated Sr ions. All
the above results clearly show that a highly selective 90Sr
ion-sieve with a very high capacity has been developed,
which is expected to find applications in the decontami-
nation of the environment after accidental releases.
ACKNOWLEDGMENT
This work was supported by the Interfacial, Transport,
and Separation Program, and Chemical and Transport
Systems, Division of the National Science Foundation
under Grant No. CTS-9612714.
REFERENCES
1. J.M. Kerr, Bull. Am. Ceram. Soc. 38, 374 (1954).
2. W.J. Lacy, Ind. Eng. Chem. 46, 1061 (1954).
3. International Atomic Energy Agency (IAEA), Tech. Rep. Ser.
136, IAEA, Vienna (1972).
4. L.P. Hatch and W.H. Regan, Jr., Nucleonics 13, 27 (1955).
5. G.B. Amphlett, L.A. McDonald, and M.J. Redman, J. Inorg. Nucl.
Chem. 6, 220 (1958).
6. S. Komarneni and R. Roy, Nature 299, 707 (1982).
7. S. Komarneni and D.M. Roy, Science 221, 647 (1983).
8. S. Komarneni and R. Roy, Science 239, 1286 (1988).
9. W.J. Paulus, S. Komarneni, and R. Roy, Nature 357, 571 (1992).
10. R.G. Anthony, C.V. Philip, and R.G. Dosch, Waste Management
13, 503 (1993).
11. K. O’Donnell, NBC News Story, July 9 (1998).
12. J. Kielland, J. Soc. Chem. Ind. 54, 232 (1935).
1256
J. Mater. Res., Vol. 15, No. 6, Jun 2000
Downloaded: 18 Mar 2015
http://journals.cambridge.org
IP address: 138.251.14.35