United States
Environmental Protection
Agency
Municipal Environmental Research
Laboratory
Cincinnati OH 45268
Research and Development
EPA-600/S2-83-107  Feb.  1984
Project  Summary
The  Equilibrium Arsenic  Capacity
of Activated Alumina
Eric Rosenblum and Dennis Clifford
  Arsenic(V) can  be effectively  re-
moved from water by adsorption onto
activated alumina. Variables affecting
the extent of adsorption include pH,
temperature, and the presence of other
ions in solution. Adsorption isotherms
establishing a  relation between solid
and liquid phase arsenic concentrations
at equilibrium can be developed by
batch tests of 7 days or more. Despite
the fact that batch tests exposed alu-
mina to uncharacteristically high initial
arsenic concentrations, the resulting
capacities were close to those obtained
by minicolumn tests that exposed alu-
mina to low arsenic concentrations for
21 days.
  Batch tests performed with granular
Alcoa* F-1 type activated alumina (28 x
48 mesh) required as many as 7 days to
reach an equilibrium solid phase loading
of 13.5 mg As(V)/g at a liquid phase
concentration of 1.0  mg As(V)/L in an
artificial ground water containing  266
mg CIVL and 367 mg SOl'/L (pH 6,
25°C). Minicolumn  tests performed
under  similar  conditions obtained a
solid phase loading of 16.1 mg As(V)/g
ajumina at 1.0 mg As(V)/L, but they
yielded results identical to  the batch
tests at equilibrium liquid phase con-
centrations less than 0.50 mg As(V)/L.
These results verified the feasibility of
batch tests for  modeling column capa-
cities in that range.
  Minicolumn tests showed  As(V)
adsorption to be extremely dependent
on pH, as expected, since the latter
determines both the surface charge on
the alumina and the valence  of the
arsenate. Maximum adsorption oc-
curred  at pH 6. Increased temperature
resulted in  increased adsorption, as
batch tests performed at 40°C yielded
As(V) solid phase loadings 33 percent
higher (at 1.0 mg  As(V)/L)  than tests

'Mention of trade names or commercial products
does not contitute endorsement or recommendation
for use.
performed  at 25°C.  This  increased
adsorption  was attributed  to kinetic
rather than energetic considerations.
Arsenic adsorption from pH  6.0 deion-
ized water reached a maximum loading
of 25 mg As(V)/g alumina  at 1.0 mg
As(V)/L, which decreased by over 50
percent in  the presence  of 720 mg
S04 /L and 16 percent in the presence
of 532 mg CIVL. These decreases were
greater than expected on the basis of
previous experiments with fluoride
adsorption onto the same type of
alumina.

  This Project Summary was developed
by EPA's Municipal Environmental
Research 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
  Arsenic is a natural constituent of a
significant percent of the world's surface
and ground water.  In several  areas,
arsenic contamination of water supplies
has been reported at levels as high as 10
mg/L — far in excess of the 0.05 mg/L
maximum contaminant level (MCL)setby
the U.S. Environmental Protection Agen-
cy (EPA).  When  arsenic is  present in
surface water, it is usually removed by
conventional water treatment methods,
including lime softening and filtration, or
coagulation by alum or ferric sulfate. But
where communities rely on untreated
well water containing  arsenic for their
water supply, treatment for arsenic
contamination  removal must be consi-
dered. For the treatment of well water, a
packed bed process is generally preferred;
and one of the most promising column
media for arsenic removal is activated

-------
alumina, which has been used success-
fully in this manner to remove fluoride.
  This  study  of  arsenic removal by
adsorption onto activated alumina  was
designed to accomplish the  following
objectives:
  1.  To determine the capacity of activated
alumina  for As(V) by both batch  and
column methods,  and to compare the
isotherms thus obtained.
  2.  To determine the optimum pH for
adsorption of As(V) onto activated alumi-
na.
  3.  To determine the effect of competi-
tion  by chloride and sulfate ions on the
adsorption of As(V) onto activated  alu-
mina.
  The work was undertaken as  laboratory
support for future pilot-scale studies to be
performed in the University of  Houston/
EPA  Mobile Drinking Water Treatment
Research Facility.


Materials and  Methods
    Alcoa F-1  type granular activated
alumina used in  the experiments  was
manually screened to a mesh size of 28 x
48 and conditioned with two cycles of 1
percent NaOH and 1 percent H2S04, with
80 bed volumes of each  solution passed
through the conditioning  column for each
cycle. Conditioned alumina was exhaus-
tively rinsed with  deionized water to pH
5.0.  The As(V) solutions were prepared
with reagent grade KH2As04.  Except for
tests of adsorption from deionized water,
all arsenic solutions contained total anion
concentrations of  15 meq/L made up of
chloride, bicarbonate, and sulfate ions at
the concentrations specified in Table 1.
Graphite furnace atomic absorption
                              spectroscopy was used to  determine
                              arsenic, with Ni(N03>2 added to prevent
                              premature volatilization. Three different
                              arsenic sources — As203, KH2As04, and
                              a purchased standard — were used for
                              standardization and cross-checking.

                              Batch  Tests
                                In  the batch tests, identical volumes
                              and concentrations of arsenic in solution
                              were exposed to different quantities of
                              activated alumina. The control contained
                              only  arsenic solution, without alumina.
                              final arsenic concentrations were deter-
                              mined,  and  the  difference between
                              batch-test and control final  concentra-
                              tions was  attributed to adsorption of
                              arsenic onto activated alumina.
                                The procedure for batch tests was as
                              follows:
                                1.  Aliquots (150-ml) of 5.0-mg As(V)/
                              L solution  were  pipetted  into 200-ml
                              glass bottles along with  measured
                              quantities (10 to 500  mg) of granular
                              activated alumina (28x48 mesh).
                                2.  Bottles were capped and placed on
                              an Eberback shaker operating at 140 5-
                              cm excursions/min.
                                3.  After  shaking," the bottles were
                              removed and their contents were decant-
                              ed and filtered through a 0.45-yum Milli-
                              pore filter.
                                4.  Each concentration of alumina was
                              tested  in duplicate.  All samples  were
                              analyzed for arsenic  concentration and
                              pH, and the results were recorded.

                              Column Tests
                                Column tests to determine the equili-
                              brium loading of As(V) onto activated
                              alumina consisted of passing an As(V)
                              solution  through glass minicolumns
Table 1.
Make-Up of Artificial Ground Water for Use in Arsenic (III) and Arsenic (V) Adsorption
Experiments
                                         pH
                                       4 to 6
                                                      pH
                                                     7 to 10
Ions
Cations:
Na*
An ions:
cr
HCOl
scfr
mg/L

345

266.2
—
367.5
meq/L

15

7.5
—
7.5
mg/L

345

177.5
305
240
meq/L

15

5
5
5
    Total Dissolved Solids
                       978.7
                                                           10675
 Total concentration of ions:  Cr = 15 meq/L = 0.015 N
 Ionic strength of pH 4 to 6 solutions: I = 0.0263 M
 Ionic strength of pH 7 to 10 solutions: I = 0.0225 M
containing  fixed amounts of activated
alumina until the effluent concentration
reached 90 percent of the  influent
concentration —  that  is,  until As(V)
removal was reduced to < 10 percent.
Control columns containing no alumina
were included for each arsenic concen-
tration and pH tested. Arsenic removed
was determined as the difference  be-
tween the arsenic in the influent and that
in the total effluent collected. The arsenic
loading on the alumina was calculated by
dividing the total weight of As(V) removed
by the weight of alumina in the column. A
diagram of the column apparatus appears
in Figure 1.
  Operation  of  the columns included
regular sampling, repeated preparation of
influent solution, and measurement and
sampling of collected effluent. In addition,
alumina in the minicolumns was periodi-
cally reclassified  (broken up) by inversion
to prevent the channeling and cementa-
tion that occurred  in the upper layers of
the column.

Results and Discussion

Equilibrium
  Figure 2 compares As(V) isotherms
developed from  batch tests conducted
under similar conditions for different
lengths  of  time. As  can be seen, the
isotherm representing results of the 14-
day  batch  tests  indicates significantly
greater adsorption  at each liquid phase
concentration than does that for either
the 1- or 2-day tests. On the other hand,
isotherms obtained by  7- and 14-day
tests appear to be in  close agreement,
indicating that most of the various con-
centrations of alumina had  reached
equilibrium  with  the  liquid  phase  ar-
senic within 7 days at 40°C.
  Furthermore, a comparison of  equili-
brium isotherms derived under similar
conditions from batch  tests and column
tests (Figure 3) indicates that the results
obained by these  two methods agree very
well at liquid phase concentrations of less
than  0.50  mg  As(V)/L.  Above that
concentration,  however, the column
isotherm predicts  solid  phase  loadings
progressively higher than those obtained
by the 7-day batch tests.
  One explanation  for this divergence is
suggested by Figure 2, which indicates
greater deviation between the 7- and 14-
day  isotherms at  higher liquid  phase
arsenic  concentrations.  This  result
implies that the higher loadings achieved
by the column method might be obtained
by batch tests performed for some period of
time longer than 7 days. Nevertheless,
the batch method seems adequate  for

-------
                                          —20-L Nalgene Carboy
"T'-Connection
Nalgene Tubing -
Gum Rubber
Connection
                                                                                  15.0
1.0 g Alumina _
(5 cm depth)
Glass Wool Plug
Peristaltic Pumps -
"T"-Connection
w/Gum Rubber
Hose and
Hoseclamp
                                          'Minicolumns
                                           (Duplicates + Blank)

                                          - Blank (no alumina)

                                          • Glass Wood Plug
                                           "T"-Connection
                                           w/Gum Rubber Hose
                                           (shown open for sampling)
                                                      5-gallon Polycarbonate
                                                      Bottles
Figure  1.
Minicolumn apparatus for arsenic(V) adsorption onto alumina by the column
method.
 predicting  column  capacity in the low
 range « 0.50 mg/L) of arsenic concen-
 trations likely  to  be encountered  in
 contaminated ground water.
  Data  from the column tests used  to
 determine  the  equilibrium  isotherm  in
 Figure 3 are plotted on a log-log scale in
 Figure 4 according to the linearized form
 of  the Freundlich  equation with a
 correlation coefficient (r2) of 0.9982,
 Constants  derived from this plot  by the
 slope-intercept method yield a Freundlich
 equation such that
            „   1 c o r* *0 43)
            qe = ID.£ Ce
 where qe and Ce are the solid (mg/g) and
                              liquid phase (mg/L equilibrium As(V)
                              concentrations, respectively.
                                The Freundlich model, which fits the
                              data  better than the Langmuir model,
                              allows for both  heterogeneous adsorp-
                              tion site energies and multilayer adsorp-
                              tion. Although the isotherms obtained at
                              pH 8  (Figure 2) have  a shape more
                              characteristic of the BET isotherm model,
                              saturation concentration was not ob-
                              tained at pH  6 in the column test results
                              shown  in Figure 3. Determination of
                              such  a saturation  concentration at
                              optimum pH may be considered an area
                              for further study.
                                                                           O—o '«>
-------
    0.05
C,(mg/L)
0.2     0.5
               log C, (mg/L)
Figure 4.   Linearized Freundlich isotherm
           for adsorption ofarsenic(V) onto
           activated alumina by column
           method at 25 °C.

            6.0,
the alumina becomes progressively less
positively charged; at pH <6.0, the anions
from the acids added  to lower the pH
compete with the arsenate anion for
adsorption sites.

Arsenic Adsorption and
 Temperature
  Isotherms for adsorption of As(V) onto
activated alumina at two different
temperatures are presented in Figure 6.
As can be seen arsenic  adsorption
increased when the temperature of the
batch tests was raised  from 25° to 40°C.
This result appears to contradict the
prevailing assumption that adsorption is
an exothermic reaction, since an increase
in temperature would then tend to drive
the reaction in the direction of desorption
from the solid phase.
  One explanation for  this apparent
anomaly,  however, is that a higher
temperature could increase the rate of
diffusion into the solid phase and  thus
 increase the  solid phase  loading  at  a
given pH and time to  equilibrium. This
 hypothesis  agrees with the findings of
 both batch and minicolumn  tests —that
 the adsorption  reaction  was largely
 controlled by kinetic constraints.
                                        75.0
                                        70.0

                                         5.0
                         0.0
                                                       -1.200
                                                                        .134
                                                                         &

                                                                        .06V
                              5.0   6.0   7.0
                                        pH
                                               8.0   9.0
                         Figure 5.    Effect of pH on adsorption of
                                    arsenic(V) onto activated alu-
                                    mina by column method.

                                    (Ca = 1 mg As(V)/L; TDS = 979
                                    \pH 5.6], and 1067 mg/L  \pH
                                    7-9])
                            0.0  1.0  2.0  3.0   4.0   5.0
                                        C. (mg/L)

                         Figure 6.    Effect or temperature on adsorp-
                                    tion of arsenic(V) onto activated
                                    alumina by batch method.

                                    (Co = 5mgA S(V)/L; pH = 6; TDS
                                    = 979 mg/L: Cl" = 7.5 meq/L;
                                    SOl" =7.5 meq/L; *.„ = 7 days)
Arsenic Adsorption and Ion
Competition
  Figure 7 compares equilibrium  iso-
therms  for  adsorption  of As(V) onto
activated alumina with those of solutions
containing different concentrations of
competing  anions.  It indicates that
arsenic adsorption is significantly reduced
in the presence of chloride and sulfate. At
an equilibrium concentration of 1.0 mg
As(V)/L in the liquid phase, solid phase
loading was reduced by 16 percent in the
presence of 532 mg CIYL (15 meq/L),
and  by  more  than 50 percent  in the
presence of 720 mg SO?~/L (15 meq/L),
as compared with adsorption from
deionized water. Similarly, in an artificial
ground water  composed  of 7.5 meq/L
each of chloride and sulfate, adsorption
was reduced by more than 40 percent.
  These results agree with the proposed
selectivity sequence for competitive
adsorption onto activated alumina, which
predicts a greater reduction from sulfate
competition than from chloride competi-
tion. However,  the sensitivity of As(V)
adsorption to competition from chloride
ions was greater than that expected on
the basis of previous work with fluoride
adsorption onto the same type of alumina.
  The full  report was  submitted in
fulfillment of Cooperative Agreement No.
CR-807939-02 by the  University of
Houston  under the sponsorship  of the
U.S.  Environmental Protection Agency.

-------
         35.0
                                     ' O
                           •'        .>•£•
                                   Deionized Water

                                   15meq/LNaCI
                           D~~Q
          1.0
                                    7.5meq/LNaCI
                                    7.5 meq/L Na^SOt
                                                               .467
                                                               .400
                                                               .334
                                                               .267
                                                               .200
                                                               .734
                                                               .067
             .10      1.00      2.00      3.00      4.00      5.00

                                  Ct(mg/L)

Figure 7.  Effect of competition by chloride and sulfate ions on adsorption of arsenic(V) onto
         activated alumina.

         (T = 25°C, C0 = Smg AS(V)/L; pH = 6; f.q = 7 days)
  Eric Rosenblum and Dennis Clifford are with the University of Houston, Houston,
    TX 77004.
  Tom Sorg is the EPA Project Officer (see below).
  The complete report, entitled "The Equilibrium Arsenic  Capacity of Activated
    Alumina," (Order No. PB8411O 527; Cost: $11.50, subject to change) will 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:
          Municipal Environmental Research Laboratory
          U.S. Environmental Protection Agency
          Cincinnati, OH 45268

-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300
                     0000329
                                                                                         U.S. GOVERNMENT PRINTING OFFICE: 1984-759-102/855

-------