United States
Environmental Protection
Agency
Risk Reduction
Engineering Laboratory
Cincinnati, OH 45268
Research and Development
EPA/600/SR-92/048 April 1992
& EPA Project Summary
Removing Radium from Water by
Plain and Treated Activated
Alumina
Deepak Garg and Dennis Clifford
This research determined the feasi-
bility of using BaSO4-impregnated acti-
vated alumina and plain activated alu-
mina for radium removal from ground-
water by fixed-bed adsorption. The ma-
jor factors influencing radium adsorp-
tion onto the two types of alumina were
identified. The radium regemerability of
the aluminas was also studied.
Good to excellent radium removals
were obtained depending on the chemi-
cal composition of the feedwater. For
example, BaSO4-impregnated alumina
treated 25,000 bed volumes (BV) and
plain alumina treated 14,000 BV of in-
fluent before radium maximum contami-
nant level (MCL) breakthrough.
The presence of sulfate ions in the
feedwater enhanced and the presence
of barium inhibited radium adsorption
on both types of alumina. The water
hardness significantly reduced radium
removal by plain alumina although it
had little effect on radium removal by
BaSO4-impregnated alumina. The
amount of BaSO4 contained in the im-
pregnated alumina correlated positively
with radium adsorption.
When acid/base regeneration was em-
ployed, BaSO4-impregnated alumina
was about 45% regenerable with re-
spect to radium and plain alumina was
about 70% to 95% regenerable. Excel-
lent performance was obtained when
using BaSO -impregnated alumina to
treat actual radium-contaminated
groundwater in Lemont, IL, where
25,000 BV could be treated before the
5 pCi/L MCL was reached. The best per-
formance with plain alumina was 12,000
BV when treating a Houston ground-
water spiked with 23 pCi/L of radium.
This Project Summary was developed
by EPA's Risk Reduction Engineering
Laboratory, Cincinnati, OH, to announce
key findings of the research project
that is fully documented in a separate
report of the same title (see Project
Report ordering information at back).
Introduction
The ability of BaSO4 precipitates to re-
move radium from solution has been
known for more than 40 yr. Copreciprtation
of RaSO4 with BaSO4 is the standard tech-
nique used to concentrate radium in analy-
sis procedures for radium isotopes. The
Dow Radium Selective Complexer (RSC)'
is an experimental adsorbent that makes
good use of the knowledge that BaSO4
scavenges radium from solution. With the
RSC, BaSO4 is impregnated into cation-
exchange beads. The work reported here
began with a similar idea, i.e., if BaSO4
could be effectively precipitated within the
pores of activated alumina, the resultant
material would serve as a good, inorganic,
radium-selective adsorbent. Activated alu-
mina is known to have a good capacity for
sulfate and, thus, was expected to be a
good host for sulfate before the precipita-
tion of BaSO . Plain activated alumina was
hot expected to be a good adsorbent for
radium but was tested for radium capacity
in control experiments. In fact, a survey of
the literature indicates that this is prob-
ably the first study done on the adsorption
of radium onto alumina of any kind.
Mention of trade names or commercial products does
not constitute endorsement or recommendation for
Printed on Recycled Paper
-------
BaSO4-impregnated alumina was pre-
pared in several ways, both in this study
and in field studies. The resultant alumina
was used to remove radium by fixed-bed
adsorption. Initial results indicated that not
only did this material remove radium from
spiked Houston tap water from a surface
source, but that plain activated alumina,
used for the control experiment, also gave
good results. Further experimentation with
BaSO,-impregnated alumina, plain alu-
mina, ana different types of feedwaters
resulted in widely variable behavior in
terms of radium removal. Very good ra-
dium removal was obtained with natural
waters, whereas essentially no removal
was obtained with synthetic waters. At
this stage, we concluded that some con-
stituent of the natural waters, not present
in the synthetic waters, was responsible
for the removal of radium and thus, initi-
ated a study of the influence of pH, chlo-
ride, sutfate, and total organic carbon
(TOO). A general objective of the lab stud-
ies with various waters was to explain
how radium is removed from natural wa-
ters by BaSO4-impregnated alumina and
plain alumina.
The effect of naturally occurring organ-
tea, measured as TOG, on radium adsorp-
tion was thoroughly investigated. We ex-
pected that some preferential association
of the Raa* ions with the predominantly
negatively charged natural organics would
exist Organics were found to enhance
radium removal, but not enough to ex-
plain the previously observed radium re-
moval, with natural waters. When we again
searched the literature to resolve the con-
fusion generated by the experimental re-
sults, we found that RaSO. ion pairs tend
to form in water when sultate is present
and that these uncharged species tend to
be adsorbed on all surfaces, particularly
glass.
Finally, in an effort to establish the
regenerabil'rty of the aluminas, acid/base
regenerations were attempted on radium-
saturated samples of plain and of BaSO4-
Impregnated alumina.
When the experiments were finished, a
statistical analysis of the data established
the effect of independent variables such
as pH, feed ^Ra activity, and the back-
ground concentration of sulfate, barium,
hardness, and TOO on the adsorption of
radium onto both BaSO4-impregnated and
plain alumina.
Thus, the major objectives of this study
were:
1. to determine the effectiveness of
BaSO.,-impregnated alumina in re-
moving radium from water by fixed-
bed adsorption,
2. to determine the effectiveness of
plain activated alumina in removing
radium from water by fixed-bed ad-
sorption, and to compare its radium
removal performance with that ob-
tained with BaSO4-impregnated alu-
mina, '
3. to identify the water quality factors
influencing radium adsorption onto
both BaSO4-impregnated and onto
plain alumina,
4. to evaluate the regenerability of both
BaSO4-impregnated and plain alu-
mina, using acid and base, and fi-
nally
5. to develop a set of rules for choos-
ing the proper alumina — BaSO4-
impregriated or plain — to be ap-
plied to water to remove radium.
Materials and Methods
Preparation of Barium-Sulfate-
Impregnated Alumina
Alcoa F-1 activated alumina (28 x 48
mesh, i.e., 0.$ x 0.3 mm diameter) was
used throughout the study. To utilize the
well-known radium removal capacity of
BaSO4 as wel) as the large surface area
present within 'the alumina, we attempted
to impregnate BaSO4 into the alumina. To
achieve impregnation, sulfate was first
adsorbed onto the alumina then BaSO4
was precipitated within the alumina pores
by adding excess barium in the form of
BaCI2.
The following reaction sequence is pro-
posed for impregnating barium sulfate into
the alumina pores:
Sulfation of the activated alumina
=AI-OH + H,S04
=AI-HSO
(1)
;
Precipitation! of BaSO4 in activated alu-
mina
2=AI-HS04 + Bad, - > 2 =AI-CI +
BaS04(s) + !H2S04 (2)
Before being studied in detail in these
laboratory studies, BaSO -impregnated
alumina was prepared and used in the
U.S. Environmental Protection Agency-
(EPA) funded, Lemont, IL, radium-removal
field study (Clifford et al., J. AWWA, 90:94-
104; July 1988). The BaSO4-loaded alu-
mina was prepared in a 1 -in.-diameter
glass column by slowly rinsing a 30-in.
deep bed of plain alumina with 10 BV of
0.25 N H2SO4 followed by 10 BV of 0.25 N
BaCL The column was extensively rinsed
(40 BV) with reverse osmosis (RO) prod-
uct water to eliminate BaSO4(s) fines. Ra-
dium adsorption tests with the BaSO4-
loaded alumina were carried out in the
column used to prepare the medium. A
control run with plain alumina was carried
out in Lemont in a similar 1-in.-diameter
column with a 20-in. depth of medium. A
similar impregnation procedure was used
for the laboratory experiments except that
0.5 N H28O4 and 1.0 N BaCI2 were used.
Normally, impregnations were carried out
at 22' to 25°C and 1 atm pressure. Higher
temperature (70°C), vacuum (25 mm Hg),
and higher pressure (2 atm) techniques
were, however, briefly experimented with.
Minicolumn Experiments
To reduce the volume of radium-spiked
water needed and to increase the number
of column runs that could be completed,
we used minicolumns of alumina contain-
ing approximately 5 cm3 of medium. Nu-
merous previous experiments over a pe-
riod of six yr had shown that such col-
umns could reliably be substituted for the
larger (1-in.-diameter) columns normally
used For adsorption experiments. The
minicolumns used in this research were
made of clear acrylic plastic, 10-in. (25.4
cm) long with an i.d. of 0.25 in. (6 mm).
The experimental minicolumn flow sys-
tem consisted of a feed solution contained
in a 13-gal (50 L) polyethylene carboy
pumped first through a 5-(im, in-line mem-
brane filter (with porous prefilter) and then
through the minicolumn. Although pump-
ing pressures less than 20 psig (138 kPa)
were required, a Milton Roy Laboratory
Data Control (model 396), high-pressure
metering pump was used as the feed
pump. Generally the flow rate through the
5 cm3 of granular 28 x 48 mesh alumina
was maintained at 1.5 mL/min for an empty
bed contact time (EBCT) of 3.3 min.
The radium adsorption performance of
a column run was determined by analyz-
ing the feed and effluent for radium.
The ^Ra isotope standard used in these
experiments for spiking the feedwater, was
calibrated and supplied by the EPA's En-
vironmental Monitoring Systems Labora-
tory in Las Vegas, NV.
One focus of this study was to examine
the effect of the presence of natural or-
ganics in feedwater on the removal of
radium on alumina. The organics were
isolated from Lake Houston water with the
use of RO concentration, followed by size
fractionation using uttrafiltration. The frac-
tions had cut-off limits of 10,000 (10K),
5,000 (5 K), and 1,000 (1 K), apparent
molecular weight units (AMU). The 1 to 5
K and the 5 to 10 K fractions were used
for the TOC spiking experiments because
these relatively smaller molecules were
expected to be at least partially adsorbed
by the aluminas.
-------
Batch (Isotherm) Experiments
Activated alumina used in the batch iso-
therm experiments was ground and sieved
to less than 250 u.m in diameter (100%
passed through a U.S. Standard Sieve
No. 60) to facilitate the kinetics of the
adsorption process. The alumina was then
transferred to five different bottles for pH
adjustment with 1 N NaOH or 1 N HCI.
This was done over a period of 2 days
with continuous shaking and monitoring of
pH. The water was decanted and the alu-
mina dried in an oven overnight. The alu-
mina was then exposed to the atmosphere
for 6 hr to ensure that its weight was
stable. pH adjustment of the water was
extremely difficult because of the lack of
buffer capacity in the deionized water; this
unsufficient buffering is why the equilib-
rium pHs were quite different from the
initial pHs.
After 1-L polyethylene bottles were filled
with the radium-spiked water and the cho-
sen weight of activated alumina, they were
placed in a heavy duty tumbler for 7 days
for equilibration at 20' to 25°C. One set of
bottles for determining one isotherm in-
cluded a blank with no alumina, and five
other bottles contained different chosen
weights of pH-adjusted alumina. After the
7-day contact period, the suspension was
filtered through a 0.45-n.m membrane fil-
ter before determining the equilibrium pH
activity of the filtrate.
Regeneration Experiments
The spent alumina columns were re-
generated with 0.5 N HCI followed by 0.25
N NaOH solutions. The runs had a total
acid contact time of 45 min and a total
base contact time of 60 min. Each regen-
eration effluent bed volume was collected
separately with the use of a fraction col-
lector. Eventually, the samples were ana-
lyzed for pH, ^Ra, and TOG.
Chemical Analysis
All ^Ra activity measurements in this
study were done by the EPA Method 900.1
(gross radium alpha screening procedure
for drinking water) described in "Prescribed
Procedures for the Measurements of Ra-
dioactivity in Drinking Water" (EPA-600/
4-80/032, August 1980).
Because we know this method can pro-
duce falsely high ^Ra results when un-
wanted short-lived alpha emitters are
present, we avoided the problem by wait-
ing typically 21 days to count the samples.
With proper allowances for the decay of
short-lived radionuclides, the precision of
the Method 900.1 analyses for ^Ra was
±6% relative standard deviation.
A Perkin-Elmer Model 5500 Inductively
Coupled Plasma Spectrometer was used
to analyze for barium, calcium, and mag-
nesium.
A Dionex ion chromatograph measured
sulfate concentration and a Dohrmann
(Model DC-80) ultra-low-level TOC ana-
lyzer measured TOC.
The BaSO4 loading on the BaSO4-im-
pregnated alumina was analyzed with the
use of a hot alkaline tetra sodium
ethylenediammine tetraacetate (EOTA)
extraction procedure followed by measure-
ment of barium in the extract with the use
of the ICP. This procedure was developed
specifically for this research because of
the failure of the usual acid digestion pro-
cedures to dissolve the BaSO4-impreg-
nated alumina.
The City of Houston, Water Quality Con-
trol Branch, Laboratory Section provided
complete chemical analysis of the Hous-
ton surface water used in these experi-
ments.
Results and Discussion
Lemont Field Study Results
The performance of plain alumina used
to treat the Lemont water was unexpect-
edly good (Figure 1). Based on published
information regarding the poor adsorption
of alkaline-earth cations at pH 7.2 (the
Lemont water pH), we did not expect any
significant radium removal beyond a few
hundred bed volumes. Nevertheless, we
found that about 3400 BV could be treated
before radium reached its MCL (which
corresponded to 3.25 pCi/L ^'Ra and 1.75
pCi/L ^Ra for a total of 5 pCi/L). By way
of explanation, we can only state that no
previous work has been published on ra-
dium adsorption onto activated alumina
and apparently published work on mg/L
levels of barium adsorption cannot be ex-
trapolated to pg/L levels of radium ad-
sorption.
Although plain alumina performed bet-
ter than expected, BaSO4-impregnated alu-
mina was outstanding by comparison —
27,500 BV to the radium MCL (Figure 1).
Also apparent in Figure 1 is improved
radium removal performance following pe-
riods of flow interruption (15 days and 7
days). This suggests that intermittent op-
eration of the alumina column would yield
higher radium loading and lower radium
concentration in the effluent. This is be-
cause stopping the flow to the column
allows relaxation of the radium concentra-
tion gradient in the solid phase, which,
upon restarting the flow, results in a greater
liquid-solid concentration gradient and,
consequently, a higher radium flux into
the alumina granules.
Minicolumn Results with
Radium-Spiked Waters
The minicolumn runs in the laboratory
with three different feedwaters provided
some hard-to-interpret results. Unlike the
Lemont results, 226Ra-spiked Houston
groundwater produced similar perfor-
mances for both plain and BaSO4-impreg-
nated aluminas. The plain alumina (12,000
BV) performed somewhat better than did
the BaSO4-loaded variety (10,000 BV) (Fig-
ure 2). The difference in alumina perfor-
mance between the Lemont and Houston
groundwaters is probably due to pH. Plain
alumina, as expected, performed better
when the pH was increased to 7.6 to 8.2
compared with the 7.2 pH of the Lemont
water. This agrees with the general obser-
vation that adsorption of cations onto alu-
mina increases as pH increases.
The lower sulfate concentration (17 mg/
L) of the Houston groundwater, compared
with that of the Lemont water (80 mg/L),
may also have contributed to the poorer
performance of the BaSO4-impregnated
alumina. For, as we found later, sulfate
concentration is a major factor in radium
adsorption onto both types of alumina.
The performance of both types of alu-
mina on the 226Ra-spiked synthetic water
was dramatically different than the perfor-
mance on the Lemont and Houston
groundwaters. We found nearly immedi-
ate breakthrough (500 BV) of radium with
the pH 7.5 to 8.0 synthetic water contain-
ing a relatively high concentration of barium
(0.9 mg/L) and no suifate. The poor per-
formance of the aluminas was attributed
mainly to the high level of barium, an
alkaline earth cation similar to radium in
chemical behavior.
The performance of the aluminas on
^Ra-spiked Houston surface water was,
again, different from that on the
groundwaters and the synthetic water. The
plain and BaSO4-impregnated aluminas
behaved similarly: the MCL was reached
at about 2,000 BV when the feedwater
contained 50 pCi/L of ^Ra. The relatively
good performance of the plain alumina
was thought to be due to the surface
water's 5 mg/L TOC, which was suspected
of complexing the radium before being
adsorbed by the alumina. For this reason,
additional minicolumn tests were per-
formed in which natural organic matter
TOC was added to resolve the issue.
When 4 mg/L TOC was added to Lemont
groundwater during a minicolumn test, no
improvement in radium removal was ob-
-------
O BaSO4 Alumina - Run E18
A Plain Alumina • Run E19
Ui
0.0
0 10000 20000 \ 30000 40000 50000 60000
Throughput, Bed Volumes
i , ' , , , .. • '.• . . : • . . • . • •:,
Flgun 1. Radium removal from Lemont, IL groundwatsr by plain (O.k L) and pressure-impregnated (0.23 L) BaSO.alumina columns. Average influent
SO/-- 80 mg/L; TOO=0,7 mg/L; Ha2* « 0.2 mg/L; total radium - 15pCi/L; and EBCT=* 3 min. 20,000 BV corresponds to a 42 • day tun length
ataminEBCT. !
0.5
0.4
A Plain Alumina-Continuous Flow-Run £15
O SaSQj Alumina*Cont!nuous Flow-Run E16
• BaSO4 Alumina-Intermittent Flow-Run E17
3000 6000\ 9000
Throughput, Bed Volumes
12000
15000
Flgvr* 2, Radium removal from Ra - 226 - spiked Houston groundw'ater by vacuum impregnated BaSO4 alumina — continuous and intermittent flow.
Average Influant SO* -17mg/L; 7OC= 0.6 mg/L; Ba**£ 0.02 mg/L; average influent Ra-226=23 pd/L; feed and effluent pH * 7.6 to 8.2; BV
•• 6.2 mL; and EBCT= 3.3 min. 12,000 BV corresponds to a 28-day run length at 3.3 min EBCT.
-------
served. Thus, the influence of TOC was
not as significant as suspected.
Statistical Evaluation of Results
Because of the number of variables in-
volved, the results obtained with column
experiments were statistically analyzed.
The objective was to establish the effects
of the independent variables, including pH,
feed 22*Ra activity, sulfate, barium, hard-
ness, and TOC concentrations, on the ad-
sorption of radium onto both BaSO4-im-
pregnated and plain alumina.
The analysis was done by means of a
correlation matrix in which an attempt was
made to correlate the dependent variable,
pCi/g radium loading, with the indepen-
dent variables just mentioned.
Table 1 summarizes the effects of the
various independent variables on the ra-
dium loadings obtained onto both BaSO4-
impregnated alumina and plain alumina.
Batch Isotherm Studies
The radium isotherm studies'were done
after the column experiments in an at-
tempt to obtain a better understanding of
the effects of pH and sulfate concentra-
tion on radium removal by plain alumina.
The initial isotherm tests with synthetic
water of variable radium activity were con-
ducted in glass bottles to determine the
effect of pH on radium removal. In these
tests, we observed significant radium ad-
sorption onto the glass. After correcting
for the influence of radium adsorption onto
the glass, it was clear that, as expected,
the higher the pH the greater the radium
adsorption. For example, as pH was in-
creased from 6.8 to 7.9, the radium load-
ing on plain alumina in equilibrium with 5
pCi/L^Ra increased from 100 to 350 pCi/
9-
The results of the second set of iso-
therms designed to determine the influ-
ence of sulfate versus chloride as back-
ground anions are shown in Figure 3.
Polyethylene bottles were used in these
tests to overcome the problems caused
by radium adsorption onto glass. Obvi-
ously, the presence of sulfate dramatically
increased the adsorption of radium onto
plain alumina. At 5 pCi/L in the water, the
increase in solid phase radium loading is
approximately four-fold. This verifies the
conclusion from the column studies re-
garding the importance of sulfate.
The mechanism of improved radium up-
take is thought to be one of enhanced
adsorption of neutral RaSOS(J) ion pairs
— the mechanism presumed to be re-
sponsible for enhanced radium adsorption
onto glass.
Barium in the Column Effluent
There was concern about barium leak-
ing from the BaSO4-impregnated alumina
because a drinking water MCL of 2 mg/L
exists for barium. Thus, the column efflu-
ents from the Lemont field study and the
lab study were checked for increased
barium. In both the field and lab study, an
increased effluent barium level of approxi-
mately 0.2 mg/L was found. This was not
considered serious because it is so far
below the existing 2 mg/L and proposed 5
mg/L MCLs.
Acid/Base Regeneration of
Alumina
Figure 4 presents the data from the
regeneration of the plain and BaSO4-im-
pregnated alumina columns, which had
adsorbed approximately equal amounts of
radium. Clearly, the plain alumina is better
regenerated in terms of radium elution
Table 1. Summary of the Effects* of Various Independent Variables on the Adsorption of Radium
onto BaSO, - Impregnated and Plain Alumina
Variable
BaSO4 - Impregnated Alumina Plain Alumina
Feed hardness content
Feed sulfate concentration
Feed barium concentration
Amount BaSo4 loaded
Feed ^Ra activity
Feed TOC concentration
Feed pH
0
++
N/A
0
N/A
The effects of the independent variables on ^Ra loading were judged as
significant when R-squared was > 0.1, and F - statistc was > 1.0.
Causes very significant increase in radium adsorption.
Causes significant increase in radium adsorption.
Causes very significant decrease in radium adsorption.
Not significant.
Not applicable.
than is BaSO4-impregnated alumina, prob-
ably because of the difference in the
mechanism of adsorption onto the two
types of alumina. The hypothesis is that
the major mechanisms of removal of ^Ra
onto plain alumina include the adsorption
of RaSOS( J) ion pairs, Ra2+-organic li-
gand complexes, and Ra2* ions. The re-
moval of radium by BaSO-impregnated
alumina is mainly because Ra2* ions are
incorporated into the BaSO4crystals. While
BaSO4 is being impregnated into the alu-
mina, a fraction of the plain alumina ad-
sorption sites presumably remains unaf-
fected. Therefore, removal of radium onto
BaSO -impregnated alumina also takes
place by all the mechanisms mentioned
above for plain alumina. The removal of
radium by plain alumina is a surface phe-
nomenon, whereas the incorporation of
radium into the matrix of BaSO4 crystals
can proceed beneath the surface. There-
fore, one can expect radium to be
desorbable, either by displacement or by
dissolution of the various radium species,
during the acid/base regeneration of plain
alumina. In the case of the BaSO4-im-
pregnated alumina, complete radium des-
orption can only occur when the BaSO4
dissolves, which is difficult to achieve be-
cause of its low solubility. Removal of
radium can also occur by the slow diffu-
sion of radium out of the Ba(Ra)SO4 ma-
trix.
Another noticeable difference between
the regeneration of plain and BaSO4-im-
pregnated alumina is that, whereas all of
the 226Ra comes off during the acid regen-
eration in the case of the plain alumina, a
significant amount of ^Ra also comes off
during the base regeneration in the case
of the BaSO^-impregnated alumina. This
can be explained on the basis of slower
desorption of radium that has been incor-
porated into the BaSO4 crystals. Mass bal-
ances on radium indicate about 95%
regenerability of plain alumina and about
45% regenerability of BaSO4-impregnated
alumina under the rather exhaustive
regenerant conditions tested.
Conclusions
1. Adsorption onto both BaSO4-im-
pregnated alumina and plain alu-
mina are technologically attractive
means to remove radium from
small-community water supplies, as
demonstrated by the long column
runs and the high radium loadings
obtained in field and laboratory stud-
ies. Radium loadings of 380' pCi/
gm for the BaSO4-impregnated alu-
mina and 280 pCi/gm for plain alu-
mina were obtained with typical
-------
2000
1500
t
a"
1000
500
O Control, 0.001 M
A Chloride Background, 0.01 M
m Sulfate Background, 0.005 M
40 60 80
Ce, pCi Ra-226/L Water
100
120
Figure 3. Isotherms forthe adsorption of Ra-226 onto activated alumina—ionic composition study. Polyethylene bottles were used Equilibrium time was
7 days and the pH was 6.77.
300000
Plain Alumina - Run R2
, BaSQd Alumina - Run R1
5 10 15
Throughput, Bed Volumes ofRegenerant
Figure 4, Regeneration of plain and BaSO4 alumina columns exhausted with Ra-226 spiked Houston surface water. Acid regeneration EBCT= 4.5 min;
bass regeneration EBCT = 6 min; and regenerant = 11 BVo.SNHCI followed by9BV 0.25 N NaOH.
-------
groundwaters containing about 12
pCi/L of radium.
2. For the Lemont groundwater with a
total radium activity of 18 pCi/L,
radium MCL breakthrough was
reached at approximately 25,000
BV for the fixed-bed process of ad-
sorption with BaSO4-impregnated
alumina. With plain alumina, radium
MCL breakthrough was reached
with a run length of 12,000 BVwith
a 23 pCi/L radium-spiked Houston
groundwater.
3. The presence of hardness inhibits
radium removal onto plain alumina
through direct competition for ad-
sorption sites by calcium and mag-
nesium. Hardness, in the range
found in groundwaters, however, did
not greatly influence radium removal
by BaSO4-impregnated alumina be-
cause the major mechanism of ra-
dium adsorption onto BaSCyim-
pregnated alumina seemed to be
the selective exchange of Ra2* for
Ba2* ions.
4. The presence of sulfate in waters
enhances radium removal onto both
types of alumina probably through
the formation of RaSO° ion pairs.
In batch experiments with plain alu-
mina, for an equilibrium ^Ra activ-
ity of 5 pCi/L, approximately 450
pCi/gm 22
nated alumina is a direct function
of the amount of BaSO4 loaded into
the alumina matrix.
9. BaSO4-impregnated alumina used
to adsorb radium is only about 45%
to 50% regenerable by acid/base
regeneration. This is because
BaSO4-impregnated alumina was
mainly regenerated by extracting
radium from the highly insoluble
BaSO., crystals; however, plain alu-
mina was regenerated by displac-
ing radium from the alumina sur-
face.
10. Plain alumina used to adsorb ra-
dium is about 70% to 95% regen-
erable through acid/base regenera-
tion. The regeneration efficiencies
of plain alumina depended mainly
on the characteristics (barium and
TOO concentration) of the
feedwater to which the alumina had
been exposed. The presence of
barium in the feedwater caused a
decrease in regeneration efficien-
cies because of the coprecipitation
of radium with BaSO4. The pres-
ence of TOC caused part of the
226Ra to come off during the base
regeneration.
11. The performance of BaSO4-impreg-
nated alumina or the plain alumina
depends on the water chemistry;
which is the better to use can be
decided after an in-depth evalua-
tion of the water chemistry of the
water supply in question. Clearly,
the BaSO4-impregnated alumina
performed better with high-hardness
waters fe 250 mg/L as CaCO3),
whereas plain alumina performed
equally well or better with low-hard-
ness waters (< 100 mg/L).
12. In light of the above conclusions,
BaSO4-impregnated alumina seems
to be a good candidate for the ra-
dium decontamination of ion-ex-
change regeneration brine solutions
or other brackish waters. (More
work, however, needs to be done
on the effects of calcium and barium
on radium adsorption in brines.)
Plain alumina is probably a better
choice to remove radium from low-
hardness well waters, containing >
40 mg/L sulfate, particularly if the
regenerability of the alumina is re-
quired.
Recommendations
More work is needed to optimize the
preparation of BaSO -impregnated alu-
mina. The radium loading capacity of the
BaSO4-impregnated alumina is directly re-
lated to the amount of BaSO4 loaded into
its matrix. Industrial-scale manufacture of
this material will be possible only after the
impregnation process is optimized on a
laboratory-scale.
Laboratory and field studies should be
done to determine the feasibility of using
the BaSO4-impregnated alumina to decon-
taminate spent ion-exchange brine before
its reuse in processes that would selec-
tively remove radium without softening.
Such processes would (a) not introduce
sodium into the product water, (b) not use
any salt for regenerant, and (c) not pro-
duce a brine to be disposed of into the
sanitary sewer. They would, however, pro-
duce a spent adsorbent media for ulti-
mate disposal.
More work is needed to clearly identify
the role of metal-ligand complexation on
the adsorption of metal ions onto iron-
oxides as it pertains to the radium-alu-
mina system. The role of natural organics
is especially important, and a clear under-
standing of their influence on radium ad-
sorption as related to the molecular size
and component constitution is necessary.
The regeneration of both BaSO4-impreg-
nated and plain alumina should be further
studied using acid and base regenera-
tions. The primary objectives should be to
determine (a) the various desorption
mechanisms involved, (b) the usefulness
of the regenerated media, (c) the maxi-
mum possible concentrations of radium in
the spent regenerants, and (d) the pos-
sible means of final disposal of radium-
contaminated spent regenerant wastewa-
ters.
The full report was submitted in fulfill-
ment of Cooperative Agreement No. CR-
813148 by the University of Houston un-
der the sponsorship of the U.S. Environ-
mental Protection Agency.
&U.S. GOVERNMENT PRINTING OFFICE: 1992 - 648-080/40241
-------
Deepak (?arg and Dennis Clifford are with the University of Houston, Houston, TX
77204-4791. • '<
Thomas J. Sorg is the EPA Project Officer (see below).
TTie complete report, entitled "Removing Radium from Water by Barium Sulfate -
Impregnated Activated Alumina.11 (Order No. PB92-164 V89/AS; Cost: $19.00;
subject to change) wilt be available only from: <
National Technical Information Service
5285 Port Royal Road '•
Springfield, VA 22161 ;
Telephone: 703-487-4650
The EPA Project Officer can be contacted at: I
Risk Reduction Engineering Laboratory
U.S. Environmental Protection Agency <
Cincinnati, OH 45268 \
United States
Environmental Protection
Agency
Center for Environmental
Research Information
Cincinnati, OH 4J5268
BULK RATE
POSTAGE & FEES PAID
EPA
PERMIT NO. G-35
Official Business
Penalty for Private Use $300
EPA/600/SR-92/048
------- |